1. 脑机接口技术体系与分类现状
1.1 技术类型与核心框架
脑机接口（Brain-Computer Interface, BCI）是一种颠覆性的技术，旨在实现人脑与外部设备之间的直接通信和控制，从而绕过传统的神经肌肉通路12。BCI技术的核心在于捕捉、分析大脑活动产生的神经信号，并将其转化为可执行的命令，以实现对外部设备（如机械臂、假肢、计算机光标等）的控制或进行信息交流3。根据神经信号采集方式的侵入性程度，BCI技术主要可分为侵入式（Invasive）、非侵入式（Non-invasive）和部分侵入式（Partially Invasive）三大类，每种类型都有其独特的技术边界、信号采集原理及适用场景。
侵入式BCI：这类技术需要通过手术将电极植入大脑皮层内部，直接记录神经元的电活动。其主要优势在于能够提供高空间分辨率和高信噪比的神经信号，从而实现更精确、更稳定的控制4。例如，微电极阵列可以直接感知单个或少量神经元的放电活动，适用于需要精细运动控制的场景。然而，侵入式BCI的缺点也显而易见，包括手术风险、潜在的感染、组织损伤以及电极的生物相容性问题，可能导致信号质量随时间下降15。目前，侵入式BCI主要用于严重瘫痪患者的功能重建，如帮助肌萎缩侧索硬化症（ALS）患者恢复交流能力或控制机械假肢36。
非侵入式BCI：与侵入式技术不同，非侵入式BCI无需手术，通过放置在头皮上的传感器来检测大脑活动。脑电图（Electroencephalography, EEG）是目前最常用且最具代表性的非侵入式BCI技术78。EEG通过测量头皮表面的电位变化来反映大脑皮层的神经活动，具有成本低、便携性高、对人体无创的优点19。其他非侵入式技术还包括功能性近红外光谱（fNIRS），它通过测量大脑皮层血氧水平变化来间接反映神经活动210。非侵入式BCI易于推广和大规模应用，但也面临信号易受伪影干扰、空间分辨率较低以及信号穿透头皮和颅骨时的衰减等挑战1112。尽管如此，非侵入式BCI在神经康复、辅助技术、娱乐和认知增强等领域展现出巨大的潜力813。
部分侵入式BCI：这类技术介于侵入式和非侵入式之间，通常指将电极放置在硬脑膜下（如硬膜外电图, ECoG）或血管内。部分侵入式BCI在提供相对较好信号质量的同时，降低了直接植入脑组织所带来的风险。ECoG能够提供比EEG更高的信噪比和空间分辨率，且无需穿透脑实质，被认为是兼顾安全性和性能的一种折衷方案。
从神经信号记录到外部设备控制的完整技术链路通常包括以下几个关键环节：
- 信号采集（Signal Acquisition）：通过上述不同类型的传感器获取大脑产生的电生理或血流动力学信号。
- 信号预处理（Signal Preprocessing）：对原始信号进行去噪、滤波和伪影去除，以提高信号质量，例如，滤除眼电、肌电等非脑源性干扰12。
- 特征提取（Feature Extraction）：从预处理后的信号中提取与用户意图相关的神经特征，如EEG中的节律活动（如运动想象引起的SMR节律）或事件相关电位（如P300波、SSVEP）81114。
- 特征选择（Feature Selection）：从提取的特征中选择最具判别力的子集，以优化解码效率和准确性15。
- 信号解码与分类（Signal Decoding & Classification）：利用机器学习算法（如深度学习、支持向量机等）对提取的神经特征进行实时解码，将其映射为特定的控制指令或意图，从而实现对外部设备的控制716。
- 设备控制与反馈（Device Control & Feedback）：解码后的指令驱动外部设备执行相应动作，同时通过视觉、听觉或触觉等方式向用户提供反馈，形成闭环系统，使用户能够调整其大脑活动以优化控制效果2。
BCI技术的不断发展，正逐步从实验室走向临床和日常应用，为残障人士带来新的希望，也为人类理解和利用大脑开辟了前所未有的途径1。
1.2 非侵入式BCI的普及化趋势
非侵入式脑机接口（BCI）因其安全性、便携性和相对较低的成本，正迅速成为推动BCI技术普及化的核心力量717。这类技术避免了侵入式手术的风险和伦理考量，使得它们能够在更大规模的研究中应用，并逐步融入日常场景，极大地拓展了BCI的应用边界718。
低成本与高便携性技术的发展
近年来，非侵入式BCI在硬件方面取得了显著进步，尤其是在低成本和高便携性设备的研发上。传统的EEG设备通常笨重且需要专业操作，而现在市场涌现出大量消费级EEG头戴设备，它们体积小巧、佩戴方便，部分甚至可无线连接。这些设备降低了BCI的使用门槛，使得普通用户也能接触和体验脑控技术。此外，功能性近红外光谱（fNIRS）等新兴非侵入式技术也因其便携性、对运动伪影的耐受性以及直接测量血氧动力学变化的能力而备受关注，尽管其信号采样率和空间分辨率相较于EEG仍有提升空间1920。这些硬件创新为非侵入式BCI在日常环境中的应用奠定了基础。
应用场景的拓展：从简单任务到复杂操作
早期非侵入式BCI主要聚焦于相对简单的控制任务。例如，基于EEG的系统常用于光标控制，通过识别用户想象的左右手运动所产生的特定脑电波模式（如运动想象SMR节律）来移动屏幕上的光标21。这种基础性的控制能力虽然有限，但对于严重瘫痪患者而言，已是重要的沟通和环境控制手段132223。
随着技术的发展，特别是机器学习和深度学习算法的引入，非侵入式BCI的能力得到了显著提升，开始能够处理更为复杂和多自由度的控制任务724。当前的研究和应用趋势体现在：
- 智能设备控制：非侵入式BCI已能实现对智能家居设备的控制，例如通过脑电信号调节灯光、开关电视或控制智能机器人2526。这些应用使得残疾人能够更独立地管理自己的生活环境27。
- 机器人操作与机械臂控制：最新的研究表明，非侵入式BCI能够实现对复杂机器人设备的精细控制。例如，多模态BCI系统（结合运动想象和稳态视觉诱发电位SSVEP）已经被用于控制无人机在三维物理空间中执行复杂的飞行任务，准确率高达86.5%28。更有研究通过深度学习解码用户脑电信号，实现了对机械臂进行连续抓取和放置物体的操作，这比传统的离散指令控制更加复杂和接近真实世界的交互需求29。这标志着非侵入式BCI已从简单的二维光标移动，升级到更具挑战性的三维空间操作和多关节机械臂的连续控制。
- 连续追踪与多维度控制：传统BCI的离散控制限制了其在实际应用中的灵活性。得益于深度学习算法在处理连续神经信号方面的优势，非侵入式BCI现在能够实现连续追踪（Continuous Pursuit）等复杂任务，要求用户在二维空间中追踪一个移动物体24。这种连续控制能力的提升，使得BCI在游戏、康复训练和辅助驾驶等领域具有更大的应用潜力。
尽管非侵入式BCI在信号分辨率和信噪比方面仍面临挑战，但其易用性、安全性和不断增强的解码能力，使其成为推动BCI技术走出实验室、服务于更广泛人群的关键驱动力。未来，随着硬件成本的进一步降低和解码算法的持续优化，非侵入式BCI有望在日常生活中扮演更加重要的角色。
2. 核心技术突破：从信号采集到智能解码
2.1 硬件创新与生物相容性提升
脑机接口技术的发展离不开硬件的革新，尤其是信号采集设备在生物相容性、长期稳定性及信号质量方面的持续改进。当前硬件创新的重点在于开发柔性电极、可穿戴传感器和形状自适应设备，以解决传统植入式设备面临的组织炎症反应、信号衰减以及机械不匹配等关键挑战30。
柔性电极的研发进展：
传统的刚性电极在植入大脑后，由于机械性质与脑组织（弹性模量约0.5-50 kPa）的巨大差异，会引起微运动和慢性炎症反应，导致神经元损伤和胶质瘢痕形成，进而影响信号记录的长期稳定性30。柔性电极，如基于单壁碳纳米管（SWCNT）/聚乙烯薄膜31、导电聚合物基纳米材料32以及聚乙烯醇（PVA）/卡拉胶/碳纳米管（CNTs）水凝胶复合材料33等，正成为解决这一问题的关键。这些材料能够更好地匹配脑组织的机械特性，减少电极与组织界面的微运动，从而降低急性插入损伤和慢性炎症反应。
- 碳纳米管基柔性电极：研究表明，嫁接到中密度聚乙烯（MD-PE）薄膜上的单壁碳纳米管（SWCNT）阵列，能够实现对硬脑膜下皮层电图（ECoG）长达8周的高质量和非常稳定的记录。组织学和免疫组织化学分析显示，SWCNT阵列具有良好的生物相容性，减少了炎症反应，并且其柔性和可拉伸性使其能很好地贴合皮层表面的不规则形貌，降低了电极-组织阻抗31。
- 导电聚合物和水凝胶复合材料：导电聚合物和水凝胶因其良好的生物相容性、机械柔韧性和电导率，被广泛应用于柔性电极的开发32。例如，一种基于PVA/卡拉胶/CNTs的水凝胶复合电极展现出优异的机械性能（拉伸强度633 kPa，弹性模量243 kPa），并在脑机接口测试中实现了与湿电极相当的信号质量和信息传输率，且具有良好的柔性，有望用于可穿戴传感设备33。
- 润滑型非免疫原性神经探针：为最大限度地减少急性插入创伤并提高长期信号记录性能，研究人员开发了表面润滑的神经探针。这种探针在插入过程中能显著降低摩擦力，减少86%对脑组织的冲击，从而立即将可测量信号电极的数量提高102%。此外，通过抑制胶质增生，信号可测量周期从8周延长至16周，显著延长了植入式BCI的使用寿命34。
可穿戴传感器与形状自适应设备：
非侵入式BCI领域，可穿戴传感器和形状自适应设备的进步也十分显著。高分辨率、多通道的无线EEG采集系统，例如基于Cortex-M4微控制器并采用Wi-Fi传输的64通道系统，支持24位高分辨率和高采样率，兼容湿式、Ag/AgCl和干式传感器，极大地方便了EEG信号的长期和移动监测35。此外，耳戴式EEG传感器也代表了脑活动监测的一个新前沿，其隐蔽性高且靠近大脑，有望成为新一代非侵入式BCI的理想选择。针对耳戴式EEG，开发了验证工具包，包括桌面应用程序和适应幻影评估概念，以评估不同耳EEG外形和传感器配置的性能，并通过优化电极参考配置，增加了听觉稳态响应（ASSR）功率36。
生物相容性与信号稳定性的优化：
为了解决长期植入中的生物相容性问题，除了材料本身的柔韧性，研究也关注于电极表面的功能化修饰。例如，采用生物活性涂层可以促进神经元生长，减少星形胶质细胞和小胶质细胞的激活，从而降低炎症反应并维持电极与神经组织的良好界面30。微针阵列（microneedles, MNs）作为一种新兴的生物医学工程技术，也为BCI提供了新的可能，其多功能性使其在神经科学和临床实践中展现出巨大的潜力，有望应用于BCI领域37。这些创新共同推动了BCI硬件向更安全、更稳定、更高效的方向发展，为实现长期可靠的神经信号采集奠定了基础。
2.2 智能解码算法的演进
脑机接口（BCI）系统性能的关键在于其能够准确、实时地将复杂的神经信号转化为可操作的指令。近年来，随着机器学习，特别是深度学习和强化学习技术的飞速发展，神经信号的特征提取和意图解码算法取得了突破性进展，极大地提升了BCI的解码精度和应用范围。
机器学习在神经信号解码中的应用
机器学习算法在BCI中的核心作用是识别神经信号中与特定意图或行为相关的模式。传统的解码方法如线性判别分析（LDA）、支持向量机（SVM）等在某些任务中表现良好，但面对神经信号的非线性、高维和时变特性时，其性能往往受限。现代机器学习方法，尤其是深度学习，能够自动从原始神经信号中学习更抽象、更具判别力的特征，从而实现更强大的解码能力3839。
深度学习：
深度神经网络（DNNs）以其多层结构和非线性映射能力，在神经解码领域展现出巨大潜力。卷积神经网络（CNNs）擅长处理具有局部结构的时间序列数据，如脑电信号的特定节律或事件相关电位，能够有效地进行特征提取。循环神经网络（RNNs），特别是长短期记忆网络（LSTMs），则适用于处理具有时间依赖性的神经信号，捕捉大脑活动的动态变化40。例如，一种新颖的Transformer架构被开发用于分析单神经元尖峰活动，能够提前230毫秒预测正确的运动方向，并在运动抑制任务中预测运动的发生或抑制，这展示了深度学习在解码复杂运动控制意图方面的强大能力41。此外，深度学习模型还被用于从脑成像数据中进行群体水平的解码，通过主题嵌入（subject embedding）技术来学习和利用主题间变异性的结构，从而提高解码性能，尤其是在数据量更大的情况下，群体模型有望超越个体模型38。
深度学习还被应用于解决EEG信号中的伪影去除问题。例如，一种两阶段深度学习方法，结合了修改后的IC-UNet架构用于伪影去除和修改后的VGGNet用于伪影类型识别，在去除眼动伪影方面表现出色，测试集准确率达到99.35%，这对于可穿戴EEG设备的可靠应用至关重要42。强化学习：
强化学习通过试错学习优化策略，使BCI系统能够根据用户反馈自适应地调整解码模型，从而提高控制的直观性和效率。在复杂的、连续的控制任务中，如机器人臂操作，强化学习能够帮助系统学习最佳的神经到行为映射，适应个体用户的神经活动模式，并随着用户训练的进行而不断改进。
实时解码算法优化
实时性是BCI投入实际应用的关键。针对复杂运动和触觉反馈等任务，解码算法的优化主要集中在提高处理速度和精度：
- 时间变异线性判别分析 (TVLDA)：在侵入式BCI中，高伽马波段的功率特征和LDA常用于分类。然而，传统的LDA是静态的，难以捕捉高伽马特征中通常存在的瞬态信息。TVLDA通过将LDA扩展到时间变异特征空间，并结合时域白化（time-domain whitening）阶段，显著提高了手势和手指运动的解码精度。例如，TVLDA在三类高水平手势识别中达到了99%的平均准确率，在单个手指解码中达到96%的平均准确率，这些精度是目前BCI领域报道的最高水平之一43。
- 连续二维轨迹解码：研究表明，通过低频脑信号解码尝试性运动（attempted movement）的连续二维轨迹是可行的。这对于脊髓损伤患者等运动功能受限的个体恢复上肢运动具有重要意义。在健全人中的实验显示，连续运动解码的平均相关性远高于随机水平；在一名脊髓损伤患者中也观察到了高于随机水平的解码性能，这表明尝试性运动解码有望成为健全人可行性研究与运动障碍患者实际应用之间的桥梁44。
多模态信号融合的优势
单一模态的神经信号（如EEG或EMG）可能因其固有的局限性（如EEG的空间分辨率低、EMG易受疲劳影响）而限制BCI的性能。将不同类型的生理信号融合，可以提供更丰富、更全面的信息，从而显著提升解码的鲁棒性和准确性45。
- EEG与EMG结合：脑电图（EEG）反映大脑的意图，而肌电图（EMG）反映肌肉的残余活动。将这两种信号融合，可以构建所谓的“混合BCI”（Hybrid-BCI）系统。例如，一项研究将EEG与EMG活动融合，发现在肌肉疲劳情况下，混合BCI的性能优于单独使用EMG，即使EMG单独表现良好，混合系统也能提供更可靠的控制，并在信号衰减时表现出优雅的性能下降45。在假肢控制中，通过深度学习方法融合EEG和EMG信号的二维图像特征，显著提高了截肢腕手运动的识别率。与单一模态EEG信号模型相比，融合方法在健康对照组中分类准确率提高了4.18-4.35%，在患者组中提高了2.51-3.00%46。这表明多模态融合能够有效克服单一信号的局限性，提供更鲁棒和准确的运动意图解码，对于残肢患者的假肢控制尤为重要。
- 多尺度信号处理：结合多尺度信号处理（如熵优化小波分析）和深度学习模型，可以更精确地分析神经传导速度（NCV）。一种先进的机器学习框架将多尺度信号处理与生理约束的深度学习相结合，通过对1842名患者的数据验证，在运动NCV准确性方面提高了23.4%，在感觉纤维方面提高了28.7%。该框架保持了生理可解释性，并通过小波优化分辨率尺度、温度补偿和概率性进展追踪等方式实现了卓越性能47。
这些智能解码算法的进步，正在将BCI从简单的概念验证阶段推向实际应用，为运动障碍患者、神经疾病患者以及认知增强领域带来革命性的解决方案。
3.1 运动功能重建与神经康复
脑机接口（BCI）技术在神经康复领域，特别是运动功能重建方面，展现出巨大的潜力。对于脑卒中和脊髓损伤等导致运动功能障碍的患者，BCI结合其他康复手段，能够显著促进神经可塑性，改善运动功能。
脑卒中患者的BCI辅助康复
脑卒中是全球范围内导致长期残疾的主要原因之一。许多患者在脑卒中后出现上肢或下肢运动功能障碍。BCI通过将患者的运动意图（通常是运动想象产生的脑电信号）转化为对外部设备的控制信号，帮助患者重新激活受损的神经通路，促进功能恢复。
- 上肢运动康复：多项临床研究证实了BCI在脑卒中上肢康复中的有效性。例如，一项针对缺血性脑卒中患者的随机对照试验显示，接受BCI康复训练的患者，其上肢富美尔评定量表（Fugl-Meyer Assessment upper extremity, FMA-UE）评分的改善优于传统康复组。具体而言，BCI组患者在1个月后的FMA-UE评分平均提高13.17分，而对照组为9.83分，组间差异为3.35分（95% CI, 1.05-5.65; p = 0.0045）48。另一项Meta分析也汇总了21项随机对照试验（涉及886名患者）的结果，表明BCI训练能显著改善脑卒中患者的FMA-UE评分（平均差异MD = 3.69，95% CI 2.41-4.96，P < 0.00001），并对Wolf运动功能测试（WMFT）和动作研究手臂测试（ARAT）等其他运动功能评估指标也有显著提升49。这些研究强调了BCI在亚急性期和慢性期脑卒中患者中均有效，尤其是在结合功能性电刺激（FES）或机器人辅助时效果更佳49。
- BCI与功能性电刺激（FES）结合：BCI与FES的结合是一种强大的协同康复策略。FES通过电刺激瘫痪肌肉，诱导肌肉收缩，提供运动反馈。当BCI实时解码患者的运动意图，并触发相应的FES时，这种同步的“思想-行动”偶联能够增强运动相关皮层区域的神经可塑性，形成更有效的神经通路。研究发现，BCI结合FES的疗法能够显著且持久地促进慢性脑卒中患者的运动恢复，其效果优于假FES对照组。这种恢复与功能性神经可塑性的量化指标相关，表现为受影响半球运动区域间功能连接的增强，且这种增强与功能改善显著相关50。此外，有研究表明，基于运动想象的BCI结合FES的康复训练，对于改善慢性脑卒中患者的腕伸肌力量（MRC-WE）和主动腕伸展活动度（AROM-WE）更有效，并且能增强受影响半球的功能连接51。
- BCI与机器人外骨骼结合：机器人外骨骼为脑卒中患者提供了机械辅助运动，结合BCI控制，可以实现患者意图驱动的精确运动。BCI控制的外骨骼不仅作为辅助设备，更能在重复使用数周后触发运动恢复，即使是慢性瘫痪患者也能受益52。机器人辅助步态训练（RAGT）结合BCI被证实能够改善脑卒中患者的运动功能、步态、平衡和神经可塑性，表现为皮层激活增加和功能连接改善53。
脊髓损伤患者的BCI辅助康复
脊髓损伤（SCI）常导致运动和感觉功能的严重丧失。BCI技术为脊髓损伤患者提供了恢复运动功能的新途径。
- 上肢功能重建：一项概念验证研究展示了高位颈段脊髓损伤患者通过侵入式BCI（iBCI）控制的植入式FES系统，协调瘫痪手臂和手进行抓握和伸取动作的能力。该患者成功地用意念控制自己的手臂完成了诸如喝咖啡和自行进食等功能性任务，这代表了BCI结合FES在恢复四肢瘫痪患者伸取和抓握运动方面的重大进展54。
- 下肢运动与步态康复：在下肢康复方面，BCI同样显示出前景。一项针对完全性截瘫患者的随机试点研究发现，结合BCI和触觉反馈的辅助步态训练，能够增强感觉和运动功能的恢复，优于单独的步态训练55。此外，BCI结合机器人外骨骼辅助的步态训练，特别是当与虚拟现实（VR）反馈和神经调控相结合时，能够促进神经可塑性，改善下肢运动功能53。
- 神经可塑性促进机制：BCI辅助康复的核心机制之一是促进神经可塑性。当BCI与外部反馈（如FES或机器人辅助）相结合时，它形成一个闭环系统，将患者的运动意图与实际运动结果紧密偶联。这种偶联行为驱动的神经刺激（brain-controlled neuromuscular stimulation）被认为能够诱导神经元网络的重组和功能性连接的增强56。例如，运动想象训练结合BCI可以增强中风患者的上肢功能和注意力的同时，改变大脑皮层激活模式，促进神经可塑性5758。这些改变包括受损半球运动区域功能连接的增加，以及对侧前中央回ZALFF（低频振幅）和同侧楔叶ZReHo（局部一致性）的改善，这些神经影像学指标与运动功能的改善呈正相关5759。
总的来说，BCI技术通过提供意图驱动的、实时的反馈和辅助，与机器人外骨骼、功能性电刺激等康复手段协同作用，在脑卒中和脊髓损伤患者的运动功能重建与神经康复中发挥着越来越重要的作用，为患者带来了新的康复希望。
3.2 神经与精神疾病治疗
脑机接口（BCI）技术在神经和精神疾病的治疗中正发挥着越来越重要的作用，为帕金森病、癫痫和意识障碍等疾病提供了新的干预策略，并在抑郁症、焦虑症等精神疾病的认知调节中展现出探索性应用。
帕金森病：自适应深部脑刺激（aDBS）
深部脑刺激（DBS）已是治疗晚期帕金森病（PD）的一种成熟方法，但传统的DBS通常以连续、固定的参数进行刺激，可能导致不必要的能量消耗和副作用 60。近年来，BCI技术的发展催生了自适应深部脑刺激（adaptive Deep Brain Stimulation, aDBS），通过实时监测患者大脑活动中的病理性生物标志物（如局部场电位LFP），根据大脑状态动态调整刺激参数，以达到更优的治疗效果。
- 工作原理：aDBS系统利用植入电极记录的神经信号（通常是丘脑底核STN或运动皮层的局部场电位），作为反馈信号来控制刺激的开启、关闭或参数调整 6162。例如，当检测到与帕金森病症状相关的特定脑电节律（如beta波段振荡）增强时，系统会施加刺激以抑制这些病理活动；当症状缓解或病理活动减弱时，刺激强度则可降低或暂停。这种闭环控制机制避免了持续刺激，减少了能量消耗，并可能降低副作用 63。
- 临床疗效：多项研究已证实aDBS相较于传统连续DBS（cDBS）的优势。一项早期研究在8名晚期PD患者中测试了BCI控制的aDBS，结果显示其运动评分改善了66%（非盲评估）和50%（盲法评估），分别比cDBS高出29%（p = 0.03）和27%（p = 0.005）61。更重要的是，实现这些改善的同时，刺激时间减少了56%，显著降低了能量需求（p < 0.001）61。另一项盲法随机交叉试点试验也表明，aDBS能够识别与高/低多巴胺能状态及其残余运动症状相关的刺激诱发伽马振荡，并实现了运动症状和生活质量的改善，为个性化aDBS的临床应用提供了基础 64。此外，有研究利用完全植入式神经假体，通过监测皮层窄带伽马（60-90 Hz）振荡来控制刺激电压，在维持治疗效果的同时，实现了38%-45%的能量节省 65。
- 长期稳定性与展望：完全植入式双向神经接口的出现，使得帕金森病患者能够进行长期多位点脑电记录，这为aDBS系统的优化提供了宝贵的平台 66。未来aDBS有望通过更精细的生物标志物识别和更智能的控制算法，实现帕金森病症状的更精准、个性化管理 62。
癫痫：闭环神经调控
癫痫是一种常见的神经系统疾病，其特征是反复发作的异常脑电活动。BCI技术在癫痫治疗中的应用主要集中在开发闭环刺激系统，以实时检测癫痫发作或其前兆，并及时施加干预以阻止或减轻发作。
- 非侵入式闭环声学BCI：一项开创性研究开发了一种新型非侵入式闭环声学脑机接口（aBCI），通过监测脑电图（EEG）信号来解码癫痫发作的起始，并触发迷走神经的超声刺激以终止癫痫发作 67。在癫痫大鼠模型中，该aBCI系统选择性激活了结节神经节中的机械敏感神经元，同时抑制了海马和杏仁核的神经元兴奋性，并通过迷走神经的超声刺激迅速阻止了癫痫发作 67。与传统迷走神经电刺激相比，aBCI显示出显著的抗癫痫效果，为按需治疗异常神经放电提供了一种新颖、安全且有效的方法 67。
- 侵入式癫痫监测与干预：侵入性电极（如立体定向脑电图sEEG）在癫痫灶定位和监测中已被广泛应用，其高时空分辨率的信号为BCI解码癫痫活动提供了优质数据 6869。结合实时信号处理和刺激技术，可以开发出能够预测并预防癫痫发作的闭环系统，例如响应性神经刺激（RNS）系统，它可以在检测到异常电活动时立即提供短脉冲电刺激。
意识障碍：诊断、沟通与评估
意识障碍（Disorders of Consciousness, DOC）患者，如植物状态（Unresponsive Wakefulness Syndrome, UWS）和微意识状态（Minimally Conscious State, MCS），其意识水平的准确评估和沟通能力的恢复是临床上面临的巨大挑战。BCI技术为这些患者提供了新的诊断、沟通和评估工具，尤其是对于那些无法通过行为响应表达意识的患者 70717273。
- 意识检测与评估：BCI可以通过分析患者对特定刺激（如听觉、视觉指令）的脑电响应，来探测是否存在潜在的意识活动或认知功能。例如，要求患者想象移动肢体或执行心算任务，并监测其EEG信号中是否出现与任务相关的特定模式（如P300事件相关电位或SMR节律的变化）。这些方法有助于识别出那些有意识但无法通过行为表现出来的患者（即认知运动分离）74。
- 沟通接口：对于少数保留意识的DOC患者，BCI可以作为一种非语言沟通手段。通过训练患者利用特定的思维活动（如“是”与“否”的思考对应不同的脑电模式）来选择屏幕上的字母或词语，从而实现简单的信息交流 6。尽管目前仍处于早期阶段，但这种沟通能力对于改善患者的生活质量具有深远意义。
- 挑战与未来方向：DOC患者脑损伤的异质性以及神经信号的复杂性给BCI的应用带来了挑战 75。未来的研究需要开发更鲁棒的BCI系统，能够适应不同患者的神经损伤模式，并能进行长期、稳定的意识评估和沟通。此外，还需要更标准化和可靠的评估方法，以促进BCI在DOC患者临床实践中的广泛应用 75。
抑郁症与焦虑症：认知调节
在精神疾病领域，BCI，特别是结合神经反馈（neurofeedback）的方法，正被探索用于调节认知功能和情绪状态，以缓解抑郁症和焦虑症等症状。
- 神经反馈训练：通过实时向患者反馈其自身脑电活动的信息，神经反馈训练旨在帮助患者学会自主调节特定脑区的活动或特定的脑电节律。例如，抑郁症患者可能通过训练提高与积极情绪相关的脑电活动（如左侧额叶α波功率），而焦虑症患者可能通过训练降低与焦虑相关的脑电活动。这种非药物、非侵入性的干预方式，有望作为辅助治疗手段，改善患者的情绪和认知控制能力。
- 闭环神经调控：除了神经反馈，更先进的闭环BCI系统也可以用于精神疾病的治疗。例如，通过监测患者的脑电活动来识别与情绪障碍相关的异常模式，并实时提供靶向性刺激（如经颅磁刺激TMS或经颅电刺激tES），以纠正这些异常。尽管这方面研究仍处于早期阶段，但其个性化和精确定向的潜力，为治疗难治性精神疾病提供了新的思路。
4. 新兴趋势：个性化与交叉融合
4.1 个性化BCI的开发与应用
随着脑机接口（BCI）技术从实验室走向更广泛的应用，个性化（Personalization）已成为提升用户体验、系统性能和临床转化效率的关键趋势。个性化BCI的开发旨在根据用户的个体差异，如脑结构、认知模式、生理状态甚至心理状态，量身定制BCI系统，从而提高用户适配性、降低训练成本并优化长期使用效果。
基于用户生理差异的个性化系统设计
每个人的大脑结构和功能都存在独特性，导致神经信号的特征以及对BCI指令的反应模式各不相同。传统的“一刀切”式BCI模型往往难以适应所有用户，导致部分用户在使用BCI时表现不佳，甚至出现“BCI文盲”（BCI Illiteracy）现象，即15%-30%的用户无法有效控制BCI系统 76。个性化BCI致力于解决这一问题，其设计理念主要体现在以下几个方面：
- 神经特征的个体化识别：不同用户在执行相同思维任务（如运动想象）时，其产生的脑电信号特征可能存在差异。个性化系统会通过学习识别每个用户特有的神经信号模式，而非依赖于预设的通用模式。例如，针对稳态视觉诱发电位（SSVEP）BCI，可以通过推荐用户无关（UI）模型作为代表，然后通过在线适应和基于用户相关（UD）数据的迁移学习技术进行增强，从而减少新用户的校准工作量并提高模型的泛化能力 77。这种方法在在线实验中实现了243.49 bits/min的平均信息传输率（ITR），是目前无校准设置下报告的最高ITR 77。
- 脑结构与功能连接的整合：结合个体的脑部核磁共振（MRI）数据，可以更精确地定位电极位置或解读皮层活动源。例如，利用功能性神经影像技术（如fMRI），可以识别与特定认知任务相关的个体化脑区激活模式，从而指导BCI解码算法的优化。
- 认知模式和心理状态的考量：用户在操作BCI时的心理状态，如疲劳、沮丧和注意力水平，会显著影响BCI的性能 78。个性化BCI系统可以整合心理生理指标（如通过EEG信号预测用户的心理状态 78 79），动态调整算法参数或提供适应性反馈，以应对这些短期变化。例如，有研究通过回归算法预测用户报告的疲劳、沮丧和注意力水平变化，并基于此设计了自适应BCI，在部分用户中显著提高了分类准确率 78。
提升用户适配性、降低训练成本
个性化BCI的主要作用之一是提高系统的用户适配性，使BCI更易于学习和使用，并减少长时间的校准和训练过程。
- 减少校准时间：传统BCI通常需要一个较长的校准阶段来收集数据以训练解码器。个性化方法，特别是基于迁移学习和自适应校准的技术，能够显著缩短这一过程。例如，通过从多源数据集推荐代表性用户无关模型，并结合在线自适应和迁移学习，可以为新用户减少大约160次的校准尝试，使得BCI在几乎无校准的情况下也能保持高预测能力 77。这种方法使得SSVEP-BCI的平均信息传输率达到243.49 bits/min，为实现即插即用、高性能的BCI奠定了基础 77。
- 自适应训练与共适应学习：个性化BCI鼓励用户和机器之间的共适应（co-adaptation）。在共适应训练中，用户学习如何调节他们的脑活动以生成清晰的信号，同时BCI系统也适应用户的独特信号模式。这种双向学习过程可以通过自适应算法实现，例如，系统可以从一个与主题无关的分类器开始，然后在一个会话内通过监督和无监督学习逐步演变为针对该用户的优化分类器。这种方法使得即使是“BCI文盲”的用户也能在短时间内（例如60分钟内）获得显著的控制能力，且无需离线校准 76。
- 跨用户泛化能力：为了实现更好的跨用户泛化，研究人员提出了自适应二进制多目标和声搜索（ABMOHS）算法，用于选择最优的EEG通道子集，并结合新的自适应跨用户泛化模型（ACGM）。该方法在保持或提升分类准确率的同时，显著减少了所需的通道数量，并显示出优于传统方法的跨用户泛化能力，从而减少了训练阶段的校准时间，提高了运动想象BCI的实用性 80。
个性化校准技术实践效果
个性化校准技术，如自适应参数调整，已被证明能够有效提升BCI的性能。
- 光标控制的个性化：一项针对低成本无线EEG光标控制BCI的可用性研究发现，用户可视化能力与光标精神控制水平之间存在正相关性。这项工作为设计个性化的以用户为中心的辅助BCI提供了有益的见解 81。通过分析训练数据和问卷，可以评估用户体验和相关混杂变量，从而为BCI的个性化设计提供依据。
- 鲁棒性与信号噪声抑制：为了应对长期使用中可能出现的信号中断和通道失效，研究人员开发了自适应统计过程控制（SPC）方法来自动识别受损通道，并利用掩蔽（masking）方法从神经网络解码器中自动移除这些通道，同时通过迁移学习和无监督学习更新模型权重。这种框架能够在不要求用户收集额外校准数据的情况下，保持高解码性能并最大限度地减少计算时间和数据存储需求，从而显著提高了BCI的鲁棒性和可用性 82。
- 闭环自适应BCI：最新进展是基于EEG的自适应双向闭环BCI，它能够根据用户的脑活动动态调整，从而增强神经康复中的响应性和有效性 83。这些系统通过机器学习算法优化用户互动，并通过活动依赖性神经可塑性促进恢复成果 83。
通过这些个性化策略，BCI技术正在变得更加智能、用户友好和高效，从而加速其在医疗康复、辅助生活和人机交互等领域的广泛应用。
4.2 多技术交叉融合场景
脑机接口（BCI）技术与多种新兴技术的交叉融合，正在开启人机交互、神经康复和认知增强的新范式。特别是与虚拟现实（VR）、增强现实（AR）的协同应用，以及与神经调控技术（如经颅磁刺激TMS）的结合，展现了巨大的潜力，有望在沉浸式康复训练、认知增强和触觉反馈复现等方面实现突破。
BCI与虚拟现实（VR）/增强现实（AR）的协同应用
VR和AR技术能够创建沉浸式或增强型的环境，为BCI的应用提供了丰富且可控的交互平台。这种融合不仅提升了用户体验，更在康复、训练和认知增强方面带来了新的可能性。
- 沉浸式康复训练：VR/AR与BCI结合，能够为神经康复，特别是运动功能障碍患者提供高度沉浸且个性化的训练环境。患者可以在虚拟世界中执行特定的运动任务，而BCI则负责解码其运动意图并驱动虚拟肢体或物理外骨骼进行相应的动作。结合视觉反馈，可以增强其运动学习和神经可塑性。VR/AR提供的实时、多感官反馈（如视觉、听觉和触觉反馈），能够显著提高患者的参与度和积极性，使康复过程更具吸引力和效率。此外，有研究表明，结合VR/AR的BCI系统在无人机控制等任务中，无论是在VR还是AR环境中，其性能和用户体验方面均无显著差异，这表明P300 BCI范式在不同沉浸式环境中具有良好的可靠性84。体验的顺序在一定程度上会影响表现，但BCI的性能在适应经验后能很好地泛化85。
- 认知增强与技能训练：VR/AR结合BCI还可以用于认知增强和技能训练。例如，通过BCI监测用户的大脑状态（如注意力、认知负荷），并结合VR/AR场景进行自适应调整，可以优化学习效果或提升认知表现。例如，在虚拟环境中进行模拟训练，BCI可以实时评估学习者的专注度，并在必要时提供辅助或调整任务难度，从而提高训练效率。
- 精神健康干预：VR/AR与BCI的结合在精神健康领域也展现出潜力。例如，在精神分裂症患者的听觉言语幻觉（AVH）治疗中，通过VR/AR构建“人-虚拟形象共生”系统，结合BCI来理解和干预患者的脑活动，有望提供非药物的创新治疗方案86。这种沉浸式体验与脑电反馈的结合，也被应用于通过数字辅助正念训练来培养自我调节技能，从而促进可持续的心理健康87。
BCI与神经调控技术的结合
神经调控技术（如经颅磁刺激TMS）能够非侵入性地调节大脑活动，与BCI结合可以形成一个强大的闭环系统，进一步增强BCI的性能或促进神经功能恢复。
- 提升BCI性能：神经调控可以通过改变特定脑区的兴奋性，优化BCI信号的生成或解码。例如，通过5 Hz重复经颅磁刺激（rTMS）靶向初级视觉皮层（V1），可以增强稳态视觉诱发电位（SSVEP）BCI的性能。研究发现，这种调控通过抑制背景活动来提高神经网络的信噪比，并增强视觉注意力，使得中频（18-22 Hz）和高频（28-32 Hz）SSVEP命令的判别能力分别提高了7.53%和11.4%88。类似地，经颅磁刺激（TMS）已被证明能够改善运动想象（MI）BCI的分类性能89。针对右侧背外侧前额叶皮层（RDLPFC）的间歇性Theta脉冲刺激（iTBS），能够通过诱导神经可塑性改变，显著提高MI-BCI的训练性能，特别是针对那些MI-BCI控制效率低下的个体，通过调节额顶注意网络、前运动-顶叶网络和辅助运动区（SMA）的激活，解决“BCI低效问题”90。
- 促进神经可塑性与康复：BCI与神经调控的结合在神经康复中也具有巨大潜力。例如，在脑卒中患者的康复中，结合运动想象BCI和rTMS被认为是一种有前景的联合疗法。研究发现，根据患者的皮层激活模式进行个性化的rTMS治疗，可以显著提高MI解码性能和Fugl-Meyer评估分数，这强调了根据个体神经重组模式定制神经调控策略的重要性91。
- 认知增强：通过神经调控技术对特定脑区进行刺激，结合BCI实时监测认知状态，有望实现精确的认知增强。例如，在学习或注意力任务中，BCI可以监测用户的认知负荷，并在必要时触发TMS刺激，以提高信息处理效率或维持专注力。
触觉反馈复现：假肢感知与体感神经假体
恢复感官反馈，尤其是触觉和本体感觉，对于提高假肢的功能性和用户的本体感至关重要。BCI与触觉反馈技术的融合，正在改变假肢用户与世界的互动方式。
- 体感神经假体：体感神经假体旨在通过直接刺激神经系统（如体感皮层或周围神经）来恢复触觉和本体感觉9293。结合BCI，用户可以通过意念控制机械臂执行抓取动作，同时，机械臂上的传感器捕获的触觉信息可以通过神经刺激的方式传递回大脑，从而产生真实的触觉感知。例如，通过双向BCI，在体感皮层进行颅内微刺激（ICMS）来诱发触觉感知，显著改善了四肢瘫痪患者对机械臂的控制性能，使抓取任务的时间缩短了一半94。这种集成视觉和触觉反馈的系统，使得假肢操作更接近于健全人的能力，提升了抓握物品的效率。
- 增强型假肢感知：除了触觉，本体感觉的恢复也至关重要。通过BCI控制的假肢，结合提供关节位置和运动信息的神经反馈，可以帮助用户更好地感知假肢的姿态和运动。这种多模态感官反馈的集成，使得BCI系统不仅能输出控制指令，还能接收并解读来自外部世界的感官信息，从而实现更自然的“人-机”闭环交互。
综上所述，BCI与其他前沿技术的交叉融合，正在推动其在康复、增强和交互领域的边界。这种多技术协同的趋势，不仅为解决现有挑战提供了新思路，也为未来BCI的广泛应用描绘了激动人心的前景。
5. 挑战与未来方向
5.1 技术与临床转化挑战
脑机接口（BCI）技术虽然取得了显著进展，但其从实验室到临床和实际应用的转化过程中仍面临诸多严峻挑战。这些挑战主要体现在技术瓶颈和临床转化两个方面，包括信号噪声抑制、长期植入稳定性、跨用户泛化性，以及临床标准化协议的缺失和大规模临床试验的局限性等。
技术瓶颈
信号噪声抑制：
神经信号，特别是脑电图（EEG）信号，具有固有的低信噪比（SNR）和非平稳性，极易受到各种生理和非生理伪影的干扰。例如，眼电（EOG）、肌电（EMG）以及电源线干扰等都可能与脑电信号混叠，严重影响信号的质量和解码精度 95。即使是侵入式BCI，如皮层脑电图（ECoG），在长时间记录中也可能面临噪声和漂移问题。有效的噪声抑制是提高BCI系统鲁棒性和可靠性的前提，但目前尚未有普适性的解决方案能彻底解决所有类型的噪声问题。尽管深度学习方法在伪影去除方面展现出潜力，如结合修改后的IC-UNet和VGGNet的两阶段深度学习模型在眼动伪影去除上达到了99.35%的准确率，但仍需进一步研究以应对更复杂的噪声环境。长期植入稳定性：
对于侵入式和部分侵入式BCI而言，电极的长期植入稳定性是一个核心挑战。植入物与脑组织之间的生物相容性问题，如慢性炎症反应、胶质瘢痕形成以及电极的机械磨损和腐蚀，会导致信号质量随时间推移而下降，甚至完全失效 9697。这要求用户需要进行频繁的重新校准或更新解码器，从而降低了系统的实用性。
例如，皮层内微电极阵列虽然能提供高精度的神经信号，但其长期稳定性受限于组织反应 96。即便是在一项研究中ECoG植入物实现了稳定的信号监测 98，并通过点击检测器实现长时间（90天）的高性能拼写功能而无需重新训练 99，但其仍可能面临信号调制功率的瞬态下降问题 99。血管内BCI作为一种微创替代方案，也存在血栓形成风险、长期电极稳定性和解剖变异性等挑战 100。尽管柔性电极和新型生物相容材料的研发（如嫁接到MD-PE薄膜上的SWCNT阵列）旨在减少炎症反应和提高长期记录稳定性，但实现数年甚至数十年的稳定性和功能性仍是当前面临的巨大挑战 97。跨用户泛化性与个体差异：
神经信号具有显著的个体差异性，不同用户在执行相同任务时，其大脑活动模式可能存在很大差异。这意味着为一个用户训练的BCI解码器往往不能直接应用于其他用户，导致系统缺乏泛化能力 95101。这种“主题间变异性”是BCI领域的一个主要障碍，使得每次新用户使用BCI时都需要耗费大量时间进行校准和模型训练。这不仅增加了计算成本，也限制了BCI技术的广泛应用。尽管研究者正在探索动态卷积 101、迁移学习和自适应算法 101 来提高模型的跨用户泛化能力，但彻底解决这一问题仍需深入研究。
临床转化挑战
临床标准化协议缺失：
当前BCI的临床研究缺乏统一的标准化协议，尤其是在疗效评估指标方面存在不统一的现象。例如，在运动康复领域，虽然FMA-UE评分是常用指标，但不同研究在BCI训练方案、干预时长、患者选择标准以及结果报告方式上差异较大，这使得不同研究之间的结果难以比较和整合，也阻碍了临床指南的制定。这种标准化程度的不足影响了BCI技术的科学验证和推广。大规模临床试验的局限性：
BCI技术的临床试验通常面临样本量小、随访时间短的局限性。尤其对于侵入式BCI，由于手术风险和伦理考量，招募大量患者进行长期跟踪研究面临困难。小样本量可能导致研究结果的统计学效力不足，难以得出具有普遍性的结论。同时，短期的随访时间无法充分评估BCI设备的长期有效性、安全性和患者生活质量的改善程度。例如，血管内BCI的临床数据目前仍然稀缺，仅有少数患者参与了临床试验，且需要更多研究来优化信号处理、增强电极生物相容性并改进血管内操作流程以实现更广泛的临床应用 100。这种局限性使得BCI技术在获得监管部门批准和进入主流临床实践方面面临重重障碍。
综上所述，BCI技术要实现大规模临床应用，需要在硬件、算法和临床实践层面进行多维度、深层次的突破，以克服现有技术与临床转化中的挑战。
5.2 伦理与安全考量
脑机接口（BCI）技术作为一项颠覆性创新，在带来巨大医疗和功能恢复潜力的同时，也引发了深刻的伦理、社会和法律挑战。随着BCI技术日益成熟并逐渐走向商业化应用，对神经数据隐私、技术滥用、人机边界以及社会公正等方面的考量变得尤为紧迫。
神经数据隐私与安全
BCI系统直接记录和解码大脑活动，这些“神经数据”（neurodata）包含了个体最私密的信息，如思想、意图、情绪和认知状态 102。神经数据泄露的风险是BCI技术面临的最主要伦理问题之一 103。
- 隐私泄露风险：大脑活动信号可以被解码以推断用户的私人信息，例如政治倾向、性取向、记忆、情绪状态，甚至“内言”（inner speech） 104105。一旦这些高度敏感的个人信息被未经授权地获取、存储或滥用，将对个人隐私构成严重威胁 102106。这可能导致个体在就业、保险、法律程序中面临歧视，甚至可能被用于商业营销或政治操纵 102。现有的数据保护法规，如《通用数据保护条例》（GDPR），可能不足以全面覆盖和保护这些新型神经数据，因为它们通常关注行为数据而非直接来源于大脑的心理数据 102103。
- 神经安全威胁：BCI系统可能成为网络攻击的目标，导致神经数据被窃取、篡改或滥用。恶意攻击者甚至可能通过BCI对用户的大脑功能进行未经授权的干预或控制，例如干扰设备操作、植入虚假信息或诱发不良情绪状态 105。这不仅威胁到用户的个人安全，也可能引发社会层面的不稳定。
- 监管空白与倡议：目前，全球范围内对神经技术的监管框架存在空白，允许神经数据的无限制解码和商业化 102。为应对这一挑战，有学者呼吁将所有源自大脑的数据归类为敏感健康数据，并将其纳入现有医疗法规的保护范围 102。同时，应实施数据加密、差分隐私和联邦学习等技术措施，以确保神经数据的隐私保护 102。建立多学科的伦理框架和监管机制，包括立法者、科学家、伦理学家、用户和公众的参与，是确保BCI技术负责任发展和使用的关键 107.
技术滥用与意识操控
BCI技术的强大功能也伴随着潜在的滥用风险，其中最令人担忧的是对个体意识和自主性的影响。
- 意识操控与自主性受损：BCI设备，特别是那些能够提供实时反馈或预测用户意图的“智能BCI”，可能在无意中甚至故意地影响用户的决策和行为 108109。例如，BCI系统若能预测用户的行动意图并提前执行，可能会让用户感到失去了对自身行为的完全控制，从而侵蚀其自主性和主体感 104108。一些研究已指出，BCI可能导致用户产生“失去控制”或“身份认同破裂”的强烈不适感 108。
- 心理完整性威胁：BCI直接干预大脑活动，可能会威胁到个体的心理完整性（mental integrity） 109。如果BCI技术可以绕过一个人的推理能力，或导致他们疏远自己的精神状态，这将被视为对心理完整性的威胁 109。这在军事或安全领域尤其值得关注，BCI的双重用途性质（既可用于医疗，又可用于军事）使得对相关神经技术的研究、开发和部署需要高度警惕和严格监管 110。
- 社会公平与可及性：BCI技术的高成本可能导致其仅限于少数富裕人群使用，从而加剧社会不平等 107。如何确保BCI技术能够公平可及，避免形成“神经鸿沟”，是未来社会公正的重要考量 103107。
人机边界与“治疗-增强”界限
随着BCI技术日益深入地整合人脑与机器，传统上关于“人”与“机器”、“治疗”与“增强”的界限变得模糊，引发了关于人类本质和身份认同的哲学和社会思考。
- 增强型BCI的伦理困境：BCI最初主要用于治疗神经疾病和恢复受损功能，但其在认知增强、技能学习和娱乐方面的潜力，使得“治疗性BCI”与“增强性BCI”之间的界限日益模糊 103111。例如，提高记忆力、注意力或感知能力的BCI是否应被视为治疗，还是仅仅为了超越人类自然能力的增强？这种模糊性引发了公平竞争、社会压力和身份认同等伦理问题。
- 自我认同与主体性：长期使用BCI，尤其是侵入式BCI，可能改变用户对自身身体和思维的感知，影响其自我认同。一些BCI用户报告称，他们感受到设备成为了自己身体的一部分，甚至是自我延伸 112。然而，当BCI失效或引发不良后果时，这种融合感可能导致严重的心理困扰，产生“异化”或“背叛”的感觉 108。
- 责任与问责：在BCI辅助决策或行动的场景中，当机器出错时，责任应由谁承担？是BCI用户、开发商、医生还是AI算法本身？这种责任归属问题需要明确的法律和伦理框架来界定 103105。
建立多学科伦理框架与监管机制
鉴于BCI技术所涉及的复杂伦理问题，仅仅依靠生物伦理学的传统原则可能不足以应对当前的挑战。
- 多学科合作：需要汇集神经科学家、工程师、伦理学家、法学家、政策制定者、患者代表以及公众，共同参与BCI伦理规范的制定 107113。
- 风险创新方法：应采纳“风险创新”（risk innovation）方法，在技术开发的早期阶段就识别和评估潜在的伦理风险，并通过前瞻性的设计和政策制定来规避或减轻这些风险 111。
- 用户中心设计：在BCI的设计和开发过程中，应始终坚持以用户为中心，充分考虑用户体验、价值观和需求，并提供足够的控制权和透明度，以保护用户的自主性 104113。
- 透明度与公众参与：提高BCI技术研发过程的透明度，并鼓励公众对BCI伦理问题的讨论和参与，有助于形成广泛的社会共识，从而引导BCI技术的负责任发展 106112。
总而言之，BCI技术的伦理与安全考量是一个多层面、动态演进的议题，需要持续的跨学科对话和制度创新，以确保这项强大技术能够造福人类，而不是带来不可预见的负面后果。
内容由 AI 生成，仅供参考，请仔细甄别
参考文献
1Review on brain-computer interface technologies in healthcare.PubMed
Evelyn Karikari, Konstantin A Koshechkin
Biophys Rev. 2023 Sep 14;15(5):1351-1358. doi: 10.1007/s12551-023-01138-6. eCollection 2023 Oct.
Brain-computer interface (BCI) technologies have developed as a game changer, altering how humans interact with computers and opening up new avenues for understanding and utilizing the power of the human brain. The goal of this research study is to assess recent breakthroughs in BCI technologies and their future prospects. The paper starts with an outline of the fundamental concepts and principles that underpin BCI technologies. It examines the many forms of BCIs, including as invasive, partially invasive, and non-invasive interfaces, emphasizing their advantages and disadvantages. The progress of BCI hardware and signal processing techniques is investigated, with a focus on the shift from bulky and invasive systems to more portable and user-friendly options. Following that, the article delves into the important advances in BCI applications across several fields. It investigates the use of BCIs in healthcare, particularly in neurorehabilitation, assistive technology, and cognitive enhancement. BCIs' potential for boosting human capacities such as communication, motor control, and sensory perception is being thoroughly researched. Furthermore, the article investigates developing BCI applications in gaming, entertainment, and virtual reality, demonstrating how BCI technologies are growing outside medical and therapeutic settings. The study also gives light on the problems and limits that prevent BCIs from being widely adopted. Ethical concerns about privacy, data security, and informed permission are addressed, highlighting the importance of strong legislative frameworks to enable responsible and ethical usage of BCI technologies. Furthermore, the study delves into technological issues such as increasing signal resolution and precision, increasing system reliability, and enabling smooth connection with existing technology. Finally, this study paper gives an in-depth examination of the advances and future possibilities of BCI technologies. It emphasizes the transformative influence of BCIs on human-computer interaction and their potential to alter healthcare, gaming, and other industries. This research intends to stimulate further innovation and progress in the field of brain-computer interfaces by addressing problems and imagining future possibilities.
2Brain-computer interfaces: Definitions and principles.PubMed
Jonathan R Wolpaw, José Del R Millán, Nick F Ramsey
Handb Clin Neurol. 2020;168:15-23. doi: 10.1016/B978-0-444-63934-9.00002-0.
Throughout life, the central nervous system (CNS) interacts with the world and with the body by activating muscles and excreting hormones. In contrast, brain-computer interfaces (BCIs) quantify CNS activity and translate it into new artificial outputs that replace, restore, enhance, supplement, or improve the natural CNS outputs. BCIs thereby modify the interactions between the CNS and the environment. Unlike the natural CNS outputs that come from spinal and brainstem motoneurons, BCI outputs come from brain signals that represent activity in other CNS areas, such as the sensorimotor cortex. If BCIs are to be useful for important communication and control tasks in real life, the CNS must control these brain signals nearly as reliably and accurately as it controls spinal motoneurons. To do this, they might, for example, need to incorporate software that mimics the function of the subcortical and spinal mechanisms that participate in normal movement control. The realization of high reliability and accuracy is perhaps the most difficult and critical challenge now facing BCI research and development. The ongoing adaptive modifications that maintain effective natural CNS outputs take place primarily in the CNS. The adaptive modifications that maintain effective BCI outputs can also take place in the BCI. This means that the BCI operation depends on the effective collaboration of two adaptive controllers, the CNS and the BCI. Realization of this second adaptive controller, the BCI, and management of its interactions with concurrent adaptations in the CNS comprise another complex and critical challenge for BCI development. BCIs can use different kinds of brain signals recorded in different ways from different brain areas. Decisions about which signals recorded in which ways from which brain areas should be selected for which applications are empirical questions that can only be properly answered by experiments. BCIs, like other communication and control technologies, often face artifacts that contaminate or imitate their chosen signals. Noninvasive BCIs (e.g., EEG- or fNIRS-based) need to take special care to avoid interpreting nonbrain signals (e.g., cranial EMG) as brain signals. This typically requires comprehensive topographical and spectral evaluations. In theory, the outputs of BCIs can select a goal or control a process. In the future, the most effective BCIs will probably be those that combine goal selection and process control so as to distribute control between the BCI and the application in a fashion suited to the current action. Through such distribution, BCIs may most effectively imitate natural CNS operation. The primary measure of BCI development is the extent to which BCI systems benefit people with neuromuscular disorders. Thus, BCI clinical evaluation, validation, and dissemination is a key step. It is at the same time a complex and difficult process that depends on multidisciplinary collaboration and management of the demanding requirements of clinical studies. Twenty-five years ago, BCI research was an esoteric endeavor pursued in only a few isolated laboratories. It is now a steadily growing field that engages many hundreds of scientists, engineers, and clinicians throughout the world in an increasingly interconnected community that is addressing the key issues and pursuing the high potential of BCI technology.
3Brain-computer interfaces for amyotrophic lateral sclerosis.PubMed
Dennis J McFarland
Muscle Nerve. 2020 Jun;61(6):702-707. doi: 10.1002/mus.26828.
A brain-computer interface (BCI) is a device that detects signals from the brain and transforms them into useful commands. Researchers have developed BCIs that utilize different kinds of brain signals. These different BCI systems have differing characteristics, such as the amount of training required and the degree to which they are or are not invasive. Much of the research on BCIs to date has involved healthy individuals and evaluation of classification algorithms. Some BCIs have been shown to have potential benefit for users with minimal muscular function as a result of amyotrophic lateral sclerosis. However, there are still several challenges that need to be successfully addressed before BCIs can be clinically useful.
4Neurobionics and the brain-computer interface: current applications and future horizons.PubMed
Jeffrey V Rosenfeld, Yan Tat Wong
Med J Aust. 2017 May 1;206(8):363-368. doi: 10.5694/mja16.01011.
The brain-computer interface (BCI) is an exciting advance in neuroscience and engineering. In a motor BCI, electrical recordings from the motor cortex of paralysed humans are decoded by a computer and used to drive robotic arms or to restore movement in a paralysed hand by stimulating the muscles in the forearm. Simultaneously integrating a BCI with the sensory cortex will further enhance dexterity and fine control. BCIs are also being developed to: provide ambulation for paraplegic patients through controlling robotic exoskeletons; restore vision in people with acquired blindness; detect and control epileptic seizures; and improve control of movement disorders and memory enhancement. High-fidelity connectivity with small groups of neurons requires microelectrode placement in the cerebral cortex. Electrodes placed on the cortical surface are less invasive but produce inferior fidelity. Scalp surface recording using electroencephalography is much less precise. BCI technology is still in an early phase of development and awaits further technical improvements and larger multicentre clinical trials before wider clinical application and impact on the care of people with disabilities. There are also many ethical challenges to explore as this technology evolves.
5Flexible Electrodes for Brain-Computer Interface System.PubMed
Junjie Wang, Tengjiao Wang, Haoyan Liu, et al.
Adv Mater. 2023 Nov;35(47):e2211012. doi: 10.1002/adma.202211012. Epub 2023 Oct 15.
Brain-computer interface (BCI) has been the subject of extensive research recently. Governments and companies have substantially invested in relevant research and applications. The restoration of communication and motor function, the treatment of psychological disorders, gaming, and other daily and therapeutic applications all benefit from BCI. The electrodes hold the key to the essential, fundamental BCI precondition of electrical brain activity detection and delivery. However, the traditional rigid electrodes are limited due to their mismatch in Young's modulus, potential damages to the human body, and a decline in signal quality with time. These factors make the development of flexible electrodes vital and urgent. Flexible electrodes made of soft materials have grown in popularity in recent years as an alternative to conventional rigid electrodes because they offer greater conformance, the potential for higher signal-to-noise ratio (SNR) signals, and a wider range of applications. Therefore, the latest classifications and future developmental directions of fabricating these flexible electrodes are explored in this paper to further encourage the speedy advent of flexible electrodes for BCI. In summary, the perspectives and future outlook for this developing discipline are provided.
6"Your Thoughts are (were) Free!": Brain-Computer-Interfaces, Neurofeedback, Detection of Deception, and the Future of Mind-Reading.PubMed
Niels Birbaumer
Appl Psychophysiol Biofeedback. 2025 Dec;50(4):551-568. doi: 10.1007/s10484-024-09648-z. Epub 2024 Jun 14.
This review describes the historical developement and rationale of clinically relevant research on neurophysiological "mind reading" paradims: Brain- Computer-Interfaces, detection of deception, brain stimulation and neurofeedback and the clinical applications in drug resistant epilepsy, chronic stroke, and communication with paralyzed locked-in persons. The emphasis lies on completely locked-in patients with amyotrophic lateral sclerosis using non-invasive and invasive brain computer interfaces and neurofeedback to restore verbal communication with the social environment. In the second part of the article we argue that success and failure of neurophysiological "mind reading" paradigms may be explained with a motor theory of thinking and emotion in combination with learning theory. The ethical implications of brain computer interface and neurofeedback approaches, particularly for severe chronic paralysis and loss of communication diseases and decisions on hastened death and euthanasia are discussed.
7Non-Invasive Brain-Computer Interfaces: State of the Art and Trends.PubMed
Bradley J Edelman, Shuailei Zhang, Gerwin Schalk, et al.
IEEE Rev Biomed Eng. 2025;18:26-49. doi: 10.1109/RBME.2024.3449790. Epub 2025 Jan 28.
Brain-computer interface (BCI) is a rapidly evolving technology that has the potential to widely influence research, clinical and recreational use. Non-invasive BCI approaches are particularly common as they can impact a large number of participants safely and at a relatively low cost. Where traditional non-invasive BCIs were used for simple computer cursor tasks, it is now increasingly common for these systems to control robotic devices for complex tasks that may be useful in daily life. In this review, we provide an overview of the general BCI framework as well as the various methods that can be used to record neural activity, extract signals of interest, and decode brain states. In this context, we summarize the current state-of-the-art of non-invasive BCI research, focusing on trends in both the application of BCIs for controlling external devices and algorithm development to optimize their use. We also discuss various open-source BCI toolboxes and software, and describe their impact on the field at large.
8Decoding brain signals: A comprehensive review of EEG-Based BCI paradigms, signal processing and applications.PubMed
Hitesh Yadav, Surita Maini
Comput Biol Med. 2025 Sep;196(Pt C):110937. doi: 10.1016/j.compbiomed.2025.110937. Epub 2025 Aug 18.
Brain-computer interface (BCI) based on electroencephalography (EEG) is a fast-developing field with a wide range of applications such as assistive technology, neurorehabilitation, entertainment, cognitive enhancement, etc. Since EEG is a non-invasive technique that captures brain activity in real time, it is ideally suited for developing interfaces that enable direct brain-to-device communication. The different paradigms utilised in EEG-based BCIs, such as Motor Imagery (MI), Steady-State Visual Evoked Potentials (SSVEP), P300 Event-related Potentials (ERP), and Hybrid paradigms that integrate several strategies for enhanced performance, are the main emphasis of this systematic review. This paper also explores the signal processing techniques, feature extraction strategies, and classification algorithms necessary for handling low-amplitude and noisy EEG recordings. The applications of BCI in different fields, as well as the challenges and possible solutions of EEG-based BCIs, are also covered in this article. Overall, the state-of-the-art in EEG-based BCIs is thoroughly reviewed in this comprehensive review article, which also identifies important areas for further study and technological advancement.
9Virtual Electroencephalogram Acquisition: A Review on Electroencephalogram Generative Methods.PubMed
Zhishui You, Yuzhu Guo, Xiulei Zhang, et al.
Sensors (Basel). 2025 May 18;25(10):3178. doi: 10.3390/s25103178.
Driven by the remarkable capabilities of machine learning, brain-computer interfaces (BCIs) are carving out an ever-expanding range of applications across a multitude of diverse fields. Notably, electroencephalogram (EEG) signals have risen to prominence as the most prevalently utilized signals within BCIs, owing to their non-invasive essence, exceptional portability, cost-effectiveness, and high temporal resolution. However, despite the significant strides made, the paucity of EEG data has emerged as the main bottleneck, preventing generalization of decoding algorithms. Taking inspiration from the resounding success of generative models in computer vision and natural language processing arenas, the generation of synthetic EEG data from limited recorded samples has recently garnered burgeoning attention. This paper undertakes a comprehensive and thorough review of the techniques and methodologies underpinning the generative models of the general EEG, namely the variational autoencoder (VAE), the generative adversarial network (GAN), and the diffusion model. Special emphasis is placed on their practical utility in augmenting EEG data. The structural designs and performance metrics of the different generative approaches in various application domains have been meticulously dissected and discussed. A comparative analysis of the strengths and weaknesses of each existing model has been carried out, and prospective avenues for future enhancement and refinement have been put forward.
10The evolution of neuromodulation for chronic stroke: From neuroplasticity mechanisms to brain-computer interfaces.PubMed
Brian F Saway, Charles Palmer, Christopher Hughes, et al.
Neurotherapeutics. 2024 Apr;21(3):e00337. doi: 10.1016/j.neurot.2024.e00337. Epub 2024 Feb 19.
Stroke is one of the most common and debilitating neurological conditions worldwide. Those who survive experience motor, sensory, speech, vision, and/or cognitive deficits that severely limit remaining quality of life. While rehabilitation programs can help improve patients' symptoms, recovery is often limited, and patients frequently continue to experience impairments in functional status. In this review, invasive neuromodulation techniques to augment the effects of conventional rehabilitation methods are described, including vagus nerve stimulation (VNS), deep brain stimulation (DBS) and brain-computer interfaces (BCIs). In addition, the evidence base for each of these techniques, pivotal trials, and future directions are explored. Finally, emerging technologies such as functional near-infrared spectroscopy (fNIRS) and the shift to artificial intelligence-enabled implants and wearables are examined. While the field of implantable devices for chronic stroke recovery is still in a nascent stage, the data reviewed are suggestive of immense potential for reducing the impact and impairment from this globally prevalent disorder.
11EEG-Based Brain-Computer Interfaces Using Motor-Imagery: Techniques and Challenges.PubMed
Natasha Padfield, Jaime Zabalza, Huimin Zhao, et al.
Sensors (Basel). 2019 Mar 22;19(6):1423. doi: 10.3390/s19061423.
Electroencephalography (EEG)-based brain-computer interfaces (BCIs), particularly those using motor-imagery (MI) data, have the potential to become groundbreaking technologies in both clinical and entertainment settings. MI data is generated when a subject imagines the movement of a limb. This paper reviews state-of-the-art signal processing techniques for MI EEG-based BCIs, with a particular focus on the feature extraction, feature selection and classification techniques used. It also summarizes the main applications of EEG-based BCIs, particularly those based on MI data, and finally presents a detailed discussion of the most prevalent challenges impeding the development and commercialization of EEG-based BCIs.
12Electroencephalography Signal Processing: A Comprehensive Review and Analysis of Methods and Techniques.PubMed
Ahmad Chaddad, Yihang Wu, Reem Kateb, et al.
Sensors (Basel). 2023 Jul 16;23(14):6434. doi: 10.3390/s23146434.
The electroencephalography (EEG) signal is a noninvasive and complex signal that has numerous applications in biomedical fields, including sleep and the brain-computer interface. Given its complexity, researchers have proposed several advanced preprocessing and feature extraction methods to analyze EEG signals. In this study, we analyze a comprehensive review of numerous articles related to EEG signal processing. We searched the major scientific and engineering databases and summarized the results of our findings. Our survey encompassed the entire process of EEG signal processing, from acquisition and pretreatment (denoising) to feature extraction, classification, and application. We present a detailed discussion and comparison of various methods and techniques used for EEG signal processing. Additionally, we identify the current limitations of these techniques and analyze their future development trends. We conclude by offering some suggestions for future research in the field of EEG signal processing.
13Brain-computer interfaces in neurological rehabilitation.PubMed
Janis J Daly, Jonathan R Wolpaw
Lancet Neurol. 2008 Nov;7(11):1032-43. doi: 10.1016/S1474-4422(08)70223-0. Epub 2008 Oct 2.
Recent advances in analysis of brain signals, training patients to control these signals, and improved computing capabilities have enabled people with severe motor disabilities to use their brain signals for communication and control of objects in their environment, thereby bypassing their impaired neuromuscular system. Non-invasive, electroencephalogram (EEG)-based brain-computer interface (BCI) technologies can be used to control a computer cursor or a limb orthosis, for word processing and accessing the internet, and for other functions such as environmental control or entertainment. By re-establishing some independence, BCI technologies can substantially improve the lives of people with devastating neurological disorders such as advanced amyotrophic lateral sclerosis. BCI technology might also restore more effective motor control to people after stroke or other traumatic brain disorders by helping to guide activity-dependent brain plasticity by use of EEG brain signals to indicate to the patient the current state of brain activity and to enable the user to subsequently lower abnormal activity. Alternatively, by use of brain signals to supplement impaired muscle control, BCIs might increase the efficacy of a rehabilitation protocol and thus improve muscle control for the patient.
14Motor-Imagery-Based Brain-Computer Interface Using Signal Derivation and Aggregation Functions.PubMed
Javier Fumanal-Idocin, Yu-Kai Wang, Chin-Teng Lin, et al.
IEEE Trans Cybern. 2022 Aug;52(8):7944-7955. doi: 10.1109/TCYB.2021.3073210. Epub 2022 Jul 19.
Brain-computer interface (BCI) technologies are popular methods of communication between the human brain and external devices. One of the most popular approaches to BCI is motor imagery (MI). In BCI applications, the electroencephalography (EEG) is a very popular measurement for brain dynamics because of its noninvasive nature. Although there is a high interest in the BCI topic, the performance of existing systems is still far from ideal, due to the difficulty of performing pattern recognition tasks in EEG signals. This difficulty lies in the selection of the correct EEG channels, the signal-to-noise ratio of these signals, and how to discern the redundant information among them. BCI systems are composed of a wide range of components that perform signal preprocessing, feature extraction, and decision making. In this article, we define a new BCI framework, called enhanced fusion framework, where we propose three different ideas to improve the existing MI-based BCI frameworks. First, we include an additional preprocessing step of the signal: a differentiation of the EEG signal that makes it time invariant. Second, we add an additional frequency band as a feature for the system: the sensorimotor rhythm band, and we show its effect on the performance of the system. Finally, we make a profound study of how to make the final decision in the system. We propose the usage of both up to six types of different classifiers and a wide range of aggregation functions (including classical aggregations, Choquet and Sugeno integrals, and their extensions and overlap functions) to fuse the information given by the considered classifiers. We have tested this new system on a dataset of 20 volunteers performing MI-based brain-computer interface experiments. On this dataset, the new system achieved 88.80% accuracy. We also propose an optimized version of our system that is able to obtain up to 90.76%. Furthermore, we find that the pair Choquet/Sugeno integrals and overlap functions are the ones providing the best results.
15[A review on electroencephalogram based channel selection].PubMed
Xiangzhe Li, Dan Wang, Baiwen Zhang, et al.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2024 Apr 25;41(2):398-405. doi: 10.7507/1001-5515.202308034.
The electroencephalogram (EEG) signal is the key signal carrier of the brain-computer interface (BCI) system. The EEG data collected by the whole-brain electrode arrangement is conducive to obtaining higher information representation. Personalized electrode layout, while ensuring the accuracy of EEG signal decoding, can also shorten the calibration time of BCI and has become an important research direction. This paper reviews the EEG signal channel selection methods in recent years, conducts a comparative analysis of the combined effects of different channel selection methods and different classification algorithms, obtains the commonly used channel combinations in motor imagery, P300 and other paradigms in BCI, and explains the application scenarios of the channel selection method in different paradigms are discussed, in order to provide stronger support for a more accurate and portable BCI system.
16Encoder-decoder optimization for brain-computer interfaces.PubMed
Josh Merel, Donald M Pianto, John P Cunningham, et al.
PLoS Comput Biol. 2015 Jun 1;11(6):e1004288. doi: 10.1371/journal.pcbi.1004288. eCollection 2015 Jun.
Neuroprosthetic brain-computer interfaces are systems that decode neural activity into useful control signals for effectors, such as a cursor on a computer screen. It has long been recognized that both the user and decoding system can adapt to increase the accuracy of the end effector. Co-adaptation is the process whereby a user learns to control the system in conjunction with the decoder adapting to learn the user's neural patterns. We provide a mathematical framework for co-adaptation and relate co-adaptation to the joint optimization of the user's control scheme ("encoding model") and the decoding algorithm's parameters. When the assumptions of that framework are respected, co-adaptation cannot yield better performance than that obtainable by an optimal initial choice of fixed decoder, coupled with optimal user learning. For a specific case, we provide numerical methods to obtain such an optimized decoder. We demonstrate our approach in a model brain-computer interface system using an online prosthesis simulator, a simple human-in-the-loop pyschophysics setup which provides a non-invasive simulation of the BCI setting. These experiments support two claims: that users can learn encoders matched to fixed, optimal decoders and that, once learned, our approach yields expected performance advantages.
17Ensemble Wavelet Decomposition-Based Detection of Mental States Using Electroencephalography Signals.PubMed
Smith K Khare, Varun Bajaj, Nikhil B Gaikwad, et al.
Sensors (Basel). 2023 Sep 13;23(18):7860. doi: 10.3390/s23187860.
Technological advancements in healthcare, production, automobile, and aviation industries have shifted working styles from manual to automatic. This automation requires smart, intellectual, and safe machinery to develop an accurate and efficient brain-computer interface (BCI) system. However, developing such BCI systems requires effective processing and analysis of human physiology. Electroencephalography (EEG) is one such technique that provides a low-cost, portable, non-invasive, and safe solution for BCI systems. However, the non-stationary and nonlinear nature of EEG signals makes it difficult for experts to perform accurate subjective analyses. Hence, there is an urgent need for the development of automatic mental state detection. This paper presents the classification of three mental states using an ensemble of the tunable Q wavelet transform, the multilevel discrete wavelet transform, and the flexible analytic wavelet transform. Various features are extracted from the subbands of EEG signals during focused, unfocused, and drowsy states. Separate and fused features from ensemble decomposition are classified using an optimized ensemble classifier. Our analysis shows that the fusion of features results in a dimensionality reduction. The proposed model obtained the highest accuracies of 92.45% and 97.8% with ten-fold cross-validation and the iterative majority voting technique. The proposed method is suitable for real-time mental state detection to improve BCI systems.
18Electroencephalography-Based Brain-Computer Interfaces in Rehabilitation: A Bibliometric Analysis (2013-2023).PubMed
Ana Sophia Angulo Medina, Maria Isabel Aguilar Bonilla, Ingrid Daniela Rodríguez Giraldo, et al.
Sensors (Basel). 2024 Nov 6;24(22):7125. doi: 10.3390/s24227125.
EEG-based Brain-Computer Interfaces (BCIs) have gained significant attention in rehabilitation due to their non-invasive, accessible ability to capture brain activity and restore neurological functions in patients with conditions such as stroke and spinal cord injuries. This study offers a comprehensive bibliometric analysis of global EEG-based BCI research in rehabilitation from 2013 to 2023. It focuses on primary research and review articles addressing technological innovations, effectiveness, and system advancements in clinical rehabilitation. Data were sourced from databases like Web of Science, and bibliometric tools (bibliometrix R) were used to analyze publication trends, geographic distribution, keyword co-occurrences, and collaboration networks. The results reveal a rapid increase in EEG-BCI research, peaking in 2022, with a primary focus on motor and sensory rehabilitation. EEG remains the most commonly used method, with significant contributions from Asia, Europe, and North America. Additionally, there is growing interest in applying BCIs to mental health, as well as integrating artificial intelligence (AI), particularly machine learning, to enhance system accuracy and adaptability. However, challenges remain, such as system inefficiencies and slow learning curves. These could be addressed by incorporating multi-modal approaches and advanced neuroimaging technologies. Further research is needed to validate the applicability of EEG-BCI advancements in both cognitive and motor rehabilitation, especially considering the high global prevalence of cerebrovascular diseases. To advance the field, expanding global participation, particularly in underrepresented regions like Latin America, is essential. Improving system efficiency through multi-modal approaches and AI integration is also critical. Ethical considerations, including data privacy, transparency, and equitable access to BCI technologies, must be prioritized to ensure the inclusive development and use of these technologies across diverse socioeconomic groups.
19Decoding subjective preference from single-trial near-infrared spectroscopy signals.PubMed
Sheena Luu, Tom Chau
J Neural Eng. 2009 Feb;6(1):016003. doi: 10.1088/1741-2560/6/1/016003. Epub 2008 Dec 22.
Near-infrared spectroscopy (NIRS) has recently been identified as a safe, portable and relatively low-cost signal acquisition tool for non-invasive brain-computer interface (BCI) development. The ultimate goal of BCI research is for the user to be able to communicate functional intent directly through thoughts. In this paper we propose an NIRS-BCI paradigm based on directly decoding neural correlates of decision making, specifically subjective preference evaluation. Nine subjects were asked to mentally evaluate two possible drinks and decide which they preferred. Frequency domain near-infrared spectroscopy was used to image each subject's prefrontal cortex during the task. Using mean signal amplitudes as features and linear discriminant analysis, we were able to decode which drink was preferred on a single-trial basis with an average accuracy of 80%.
20Brain-computer interfaces for neurorehabilitation.PubMed
Sujesh Sreedharan, Ranganatha Sitaram, Joseph S Paul, et al.
Crit Rev Biomed Eng. 2013;41(3):269-79. doi: 10.1615/critrevbiomedeng.2014010697.
Brain-computer interfaces (BCIs) enable control of computers and other assistive devices, such as neuro-prostheses, which are used for communication, movement restoration, neuro-modulation, and muscle stimulation, by using only signals measured directly from the brain. A BCI creates a new output channel for the brain to a computer or a device. This requires retrieval of signals of interest from the brain, and its use for neuro-rehabilitation by means of interfacing the signals to a computerized device. Brain signals such as action potentials from single neurons or nerve fibers, extracellular local field potentials (LFPs), electrocorticograms, electroencephalogram and its components such as the event-related brain potentials, real-time functional magnetic resonance imaging, near-infrared spectroscopy, and magneto-encephalogram have been used. BCIs are envisaged to be useful for communication, control and self-regulation of brain function. BCIs employ neurofeedback to enable operant conditioning to allow the user to learn using it. Paralytic conditions arising from stroke or other diseases are being targeted for BCI application. Neurofeedback strategies ranging from sensory feedback to direct brain stimulation are being employed. Existing BCIs are limited in their throughput in terms of letters per minute or commands per minute, and need extensive training to use the BCI. Further, they can cause rapid fatigue due to use and have limited adaptability to changes in the patient's brain state. The challenge before BCI technology for neuro-rehabilitation today is to enable effective clinical use of BCIs with minimal effort to set up and operate.
21A comprehensive review of EEG-based brain-computer interface paradigms.PubMed
Reza Abiri, Soheil Borhani, Eric W Sellers, et al.
J Neural Eng. 2019 Feb;16(1):011001. doi: 10.1088/1741-2552/aaf12e. Epub 2018 Nov 15.
Advances in brain science and computer technology in the past decade have led to exciting developments in brain-computer interface (BCI), thereby making BCI a top research area in applied science. The renaissance of BCI opens new methods of neurorehabilitation for physically disabled people (e.g. paralyzed patients and amputees) and patients with brain injuries (e.g. stroke patients). Recent technological advances such as wireless recording, machine learning analysis, and real-time temporal resolution have increased interest in electroencephalographic (EEG) based BCI approaches. Many BCI studies have focused on decoding EEG signals associated with whole-body kinematics/kinetics, motor imagery, and various senses. Thus, there is a need to understand the various experimental paradigms used in EEG-based BCI systems. Moreover, given that there are many available options, it is essential to choose the most appropriate BCI application to properly manipulate a neuroprosthetic or neurorehabilitation device. The current review evaluates EEG-based BCI paradigms regarding their advantages and disadvantages from a variety of perspectives. For each paradigm, various EEG decoding algorithms and classification methods are evaluated. The applications of these paradigms with targeted patients are summarized. Finally, potential problems with EEG-based BCI systems are discussed, and possible solutions are proposed.
22Brain-computer interfaces for communication and rehabilitation.PubMed
Ujwal Chaudhary, Niels Birbaumer, Ander Ramos-Murguialday
Nat Rev Neurol. 2016 Sep;12(9):513-25. doi: 10.1038/nrneurol.2016.113. Epub 2016 Aug 19.
Brain-computer interfaces (BCIs) use brain activity to control external devices, thereby enabling severely disabled patients to interact with the environment. A variety of invasive and noninvasive techniques for controlling BCIs have been explored, most notably EEG, and more recently, near-infrared spectroscopy. Assistive BCIs are designed to enable paralyzed patients to communicate or control external robotic devices, such as prosthetics; rehabilitative BCIs are designed to facilitate recovery of neural function. In this Review, we provide an overview of the development of BCIs and the current technology available before discussing experimental and clinical studies of BCIs. We first consider the use of BCIs for communication in patients who are paralyzed, particularly those with locked-in syndrome or complete locked-in syndrome as a result of amyotrophic lateral sclerosis. We then discuss the use of BCIs for motor rehabilitation after severe stroke and spinal cord injury. We also describe the possible neurophysiological and learning mechanisms that underlie the clinical efficacy of BCIs.
23Single-paradigm and hybrid brain computing interfaces and their use by disabled patients.PubMed
M de Neeling, M M Van Hulle
J Neural Eng. 2019 Oct 23;16(6):061001. doi: 10.1088/1741-2552/ab2706.
Brain computer interfacing (BCI) has enjoyed increasing interest not only from research communities such as engineering and neuroscience but also from visionaries that predict it will change the way we will interact with technology. Since BCIs establish an alternative communication channel between the brain and the outside world, they have been hailed to provide solutions for patients suffering from severe motor- and/or communication disabilities such as fully paralyzed locked-in syndrome patients. However, despite single-case successes, which sometimes reach a broad audience, BCIs are actually not routinely used to support patients in their daily life activities. This review focusses on non-invasive BCIs, introduces the main paradigms and applications, and shows how the technology has improved over recent years. We identify patient groups that potentially can benefit from BCIs by referring to disability levels and etiology. We list the requirements, indicate how BCIs can tap into their spared competences, and discuss performance issues also in view of other assistive communication technologies. We discuss hybrid BCIs, a more recent development that combines paradigms and signals, possibly also of non-brain origin, to increase performance in terms of accuracy and/or communication speed, also as a way to counter the low performance with a given paradigm by involving another, more suitable one (BCI illiteracy). Finally, we list a few hybrid BCI solutions for patients and note that demonstrations with the ones based entirely on brain activity are still scarce.
24Continuous Tracking using Deep Learning-based Decoding for Non-invasive Brain-Computer Interface.PubMed
Dylan Forenzo, Hao Zhu, Jenn Shanahan, et al.
bioRxiv. 2024 Apr 25:2023.10.12.562084. doi: 10.1101/2023.10.12.562084.
UNLABELLED: Brain-computer interfaces (BCI) using electroencephalography (EEG) provide a non-invasive method for users to interact with external devices without the need for muscle activation. While noninvasive BCIs have the potential to improve the quality of lives of healthy and motor impaired individuals, they currently have limited applications due to inconsistent performance and low degrees of freedom. In this study, we use deep learning (DL)-based decoders for online Continuous Pursuit (CP), a complex BCI task requiring the user to track an object in two-dimensional space. We developed a labeling system to use CP data for supervised learning, trained DL-based decoders based on two architectures, including a newly proposed adaptation of the PointNet architecture, and evaluated the performance over several online sessions. We rigorously evaluated the DL-based decoders in a total of 28 human participants, and found that the DL-based models improved throughout the sessions as more training data became available and significantly outperformed a traditional BCI decoder by the last session. We also performed additional experiments to test an implementation of transfer learning by pre-training models on data from other subjects, and mid-session training to reduce inter-session variability. The results from these experiments showed that pre-training did not significantly improve performance, but updating the models mid-session may have some benefit. Overall, these findings support the use of DL-based decoders for improving BCI performance in complex tasks like CP, which can expand the potential applications of BCI devices and help improve the quality of lives of healthy and motor-impaired individuals. SIGNIFICANCE STATEMENT: Brain-computer Interfaces (BCI) have the potential to replace or restore motor functions for patients and can benefit the general population by providing a direct link of the brain with robotics or other devices. In this work, we developed a paradigm using deep learning (DL)-based decoders for continuous control of a BCI system and demonstrated its capabilities through extensive online experiments. We also investigate how DL performance is affected by varying amounts of training data and collected more than 150 hours of BCI data that can be used to train new models. The results of this study provide valuable information for developing future DL-based BCI decoders which can improve performance and help bring BCIs closer to practical applications and wide-spread use.
25Recent advances and current trends in brain-computer interface research and their applications.PubMed
Martina Zabcikova, Zuzana Koudelkova, Roman Jasek, et al.
Int J Dev Neurosci. 2022 Apr;82(2):107-123. doi: 10.1002/jdn.10166. Epub 2022 Jan 24.
Brain-computer interface (BCI) provides direct communication between the brain and an external device. BCI systems have become a trendy field of research in recent years. These systems can be used in a variety of applications to help both disabled and healthy people. Concerning significant BCI progress, we may assume that these systems are not very far from real-world applications. This review has taken into account current trends in BCI research. In this survey, 100 most cited articles from the WOS database were selected over the last 4 years. This survey is divided into several sectors. These sectors are Medicine, Communication and Control, Entertainment, and Other BCI applications. The application area, recording method, signal acquisition types, and countries of origin have been identified in each article. This survey provides an overview of the BCI articles published from 2016 to 2020 and their current trends and advances in different application areas.
26Brain Computer Interfaces for Improving the Quality of Life of Older Adults and Elderly Patients.PubMed
Abdelkader Nasreddine Belkacem, Nuraini Jamil, Jason A Palmer, et al.
Front Neurosci. 2020 Jun 30;14:692. doi: 10.3389/fnins.2020.00692. eCollection 2020.
All people experience aging, and the related physical and health changes, including changes in memory and brain function. These changes may become debilitating leading to an increase in dependence as people get older. Many external aids and tools have been developed to allow older adults and elderly patients to continue to live normal and comfortable lives. This mini-review describes some of the recent studies on cognitive decline and motor control impairment with the goal of advancing non-invasive brain computer interface (BCI) technologies to improve health and wellness of older adults and elderly patients. First, we describe the state of the art in cognitive prosthetics for psychiatric diseases. Then, we describe the state of the art of possible assistive BCI applications for controlling an exoskeleton, a wheelchair and smart home for elderly people with motor control impairments. The basic age-related brain and body changes, the effects of age on cognitive and motor abilities, and several BCI paradigms with typical tasks and outcomes are thoroughly described. We also discuss likely future trends and technologies to assist healthy older adults and elderly patients using innovative BCI applications with minimal technical oversight.
27Towards BCI-actuated smart wheelchair system.PubMed
Jingsheng Tang, Yadong Liu, Dewen Hu, et al.
Biomed Eng Online. 2018 Aug 20;17(1):111. doi: 10.1186/s12938-018-0545-x.
BACKGROUND: Electroencephalogram-based brain-computer interfaces (BCIs) represent novel human machine interactive technology that allows people to communicate and interact with the external world without relying on their peripheral muscles and nervous system. Among BCI systems, brain-actuated wheelchairs are promising systems for the rehabilitation of severely motor disabled individuals who are unable to control a wheelchair by conventional interfaces. Previous related studies realized the easy use of brain-actuated wheelchairs that enable people to navigate the wheelchair through simple commands; however, these systems rely on offline calibration of the environment. Other systems do not rely on any prior knowledge; however, the control of the system is time consuming. In this paper, we have proposed an improved mobile platform structure equipped with an omnidirectional wheelchair, a lightweight robotic arm, a target recognition module and an auto-control module. Based on the you only look once (YOLO) algorithm, our system can, in real time, recognize and locate the targets in the environment, and the users confirm one target through a P300-based BCI. An expert system plans a proper solution for a specific target; for example, the planned solution for a door is opening the door and then passing through it, and the auto-control system then jointly controls the wheelchair and robotic arm to complete the operation. During the task execution, the target is also tracked by using an image tracking technique. Thus, we have formed an easy-to-use system that can provide accurate services to satisfy user requirements, and this system can accommodate different environments. RESULTS: To validate and evaluate our system, an experiment simulating the daily application was performed. The tasks included the user driving the system closer to a walking man and having a conversation with him; going to another room through a door; and picking up a bottle of water on the desk and drinking water. Three patients (cerebral infarction; spinal injury; and stroke) and four healthy subjects participated in the test and all completed the tasks. CONCLUSION: This article presents a brain-actuated smart wheelchair system. The system is intelligent in that it provides efficient and considerate services for users. To test the system, three patients and four healthy subjects were recruited to participate in a test. The results demonstrate that the system works smartly and efficiently; with this system, users only need to issue small commands to get considerate services. This system is of significance for accelerating the application of BCIs in the practical environment, especially for patients who will use a BCI for rehabilitation applications.
28Quadcopter Flight Control Using a Non-invasive Multi-Modal Brain Computer Interface.PubMed
Xu Duan, Songyun Xie, Xinzhou Xie, et al.
Front Neurorobot. 2019 May 31;13:23. doi: 10.3389/fnbot.2019.00023. eCollection 2019.
Brain-Computer Interfaces (BCIs) translate neuronal information into commands to control external software or hardware, which can improve the quality of life for both healthy and disabled individuals. Here, a multi-modal BCI which combines motor imagery (MI) and steady-state visual evoked potential (SSVEP) is proposed to achieve stable control of a quadcopter in three-dimensional physical space. The complete information common spatial pattern (CICSP) method is used to extract two MI features to control the quadcopter to fly left-forward and right-forward, and canonical correlation analysis (CCA) is employed to perform the SSVEP classification for rise and fall. Eye blinking is designed to switch these two modes while hovering. Real-time feedback is provided to subjects by a global camera. Two flight tasks were conducted in physical space in order to certify the reliability of the BCI system. Subjects were asked to control the quadcopter to fly forward along the zig-zag pattern to pass through a gate in the relatively simple task. For the other complex task, the quadcopter was controlled to pass through two gates successively according to an S-shaped route. The performance of the BCI system is quantified using suitable metrics and subjects are able to acquire 86.5% accuracy for the complicated flight task. It is demonstrated that the multi-modal BCI has the ability to increase the accuracy rate, reduce the task burden, and improve the performance of the BCI system in the real world.
29Continuous Reaching and Grasping With a BCI Controlled Robotic Arm in Healthy and Stroke-Affected Individuals.PubMed
Dylan Forenzo, Yisha Zhang, George F Wittenberg, et al.
IEEE Trans Neural Syst Rehabil Eng. 2025;33:3888-3899. doi: 10.1109/TNSRE.2025.3611821.
Recent advancements in signal processing techniques have enabled non-invasive Brain-Computer Interfaces (BCIs) to control assistive devices, like robotic arms, directly with users' EEG signals. However, the applications of these systems are currently limited by the low signal-to-noise ratio and spatial resolution of EEG from which brain intention is decoded. In this study, we propose a motor-imagery (MI) paradigm, inspired by the mechanisms of a computer mouse, that adds an additional "click" signal to an established 2D movement BCI paradigm. The additional output signal increases the degrees of freedom of the BCI system and may enable more complex tasks. We evaluated this paradigm using deep learning (DL) based signal processing on both healthy subjects and stroke-survivors in online BCI tasks derived from two potential applications: clicking on virtual targets and moving physical objects with a robotic arm in a continuous reach-and-grasp task. The results show that subjects were able to control both movement and clicking simultaneously to grab, move, and place up to an average of 7 cups in a 5-minute run using the robotic arm. The proposed paradigm provides an additional degree of freedom to EEG BCIs, and improves upon existing systems by enabling continuous control of reach-and-grasp tasks instead of selecting from a discrete list of predetermined actions. The tasks studied in these experiments show BCIs may be used to control computer cursors or robotic arms for complex real-world or clinical applications in the near future, potentially improving the lives of both healthy individuals and motor-impaired patients.
30Revolutionizing brain‒computer interfaces: overcoming biocompatibility challenges in implantable neural interfaces.PubMed
Weihang Gao, Zineng Yan, Hong Zhou, et al.
J Nanobiotechnology. 2025 Jul 10;23(1):498. doi: 10.1186/s12951-025-03573-x.
Brain‒computer interfaces (BCIs) exhibit significant potential for various applications, including neurofeedback training, neurological injury management, and language, sensory and motor rehabilitation. Neural interfacing electrodes are positioned between external electronic devices and the nervous system to capture complex neuronal activity data and promote the repair of damaged neural tissues. Implantable neural electrodes can record and modulate neural activities with both high spatial and high temporal resolution, offering a wide window for neuroscience research. Despite significant advancements over the years, conventional neural electrode interfaces remain insufficient for fully achieving these objectives, particularly in the context of long-term implantation. The primary limitation stems from the poor biocompatibility and mechanical mismatch between the interfacing electrodes and neural tissues, which induce a local immune response and scar tissue formation, thus decreasing the performance and useful lifespan. Therefore, neural interfaces should ideally exhibit appropriate stiffness and minimal foreign body reactions to mitigate neuroinflammation and enhance recording quality. This review provides an exhaustive analysis of the current understanding of the critical failure modes that may impact the performance of implantable neural electrodes. Additionally, this study provides a comprehensive overview of the current research on coating materials and design strategies for implanted neural interfaces and discusses the primary challenges currently facing long-term implantation of neural electrodes. Finally, we present our perspective and propose possible future research directions to improve implantable neural interfaces for BCIs.
31Chronic neural interfacing with cerebral cortex using single-walled carbon nanotube-polymer grids.PubMed
Luigi Pavone, Slavianka Moyanova, Federica Mastroiacovo, et al.
J Neural Eng. 2020 Jul 3;17(3):036032. doi: 10.1088/1741-2552/ab98db.
OBJECTIVE: The development of electrode arrays able to reliably record brain electrical activity is a critical issue in brain machine interface (BMI) technology. In the present study we undertook a comprehensive physico-chemical, physiological, histological and immunohistochemical characterization of new single-walled carbon nanotubes (SWCNT)-based electrode arrays grafted onto medium-density polyethylene (MD-PE) films. APPROACH: The long-term electrical stability, flexibility, and biocompatibility of the SWCNT arrays were investigated in vivo in laboratory rats by two-months recording and analysis of subdural electrocorticogram (ECoG). Ex-vivo characterization of a thin flexible and single probe SWCNT/polymer electrode is also provided. MAIN RESULTS: The SWCNT arrays were able to capture high quality and very stable ECoG signals across 8 weeks. The histological and immunohistochemical analyses demonstrated that SWCNT arrays show promising biocompatibility properties and may be used in chronic conditions. The SWCNT-based arrays are flexible and stretchable, providing low electrode-tissue impedance, and, therefore, high compliance with the irregular topography of the cortical surface. Finally, reliable evoked synaptic local field potentials in rat brain slices were recorded using a special SWCNT-polymer-based flexible electrode. SIGNIFICANCE: The results demonstrate that the SWCNT arrays grafted in MD-PE are suitable for manufacturing flexible devices for subdural ECoG recording and might represent promising candidates for long-term neural implants for epilepsy monitoring or neuroprosthetic BMI.
32Conducting polymer-based nanostructured materials for brain-machine interfaces.PubMed
Yasamin Ziai, Seyed Shahrooz Zargarian, Chiara Rinoldi, et al.
Wiley Interdiscip Rev Nanomed Nanobiotechnol. 2023 Sep-Oct;15(5):e1895. doi: 10.1002/wnan.1895. Epub 2023 May 4.
As scientists discovered that raw neurological signals could translate into bioelectric information, brain-machine interfaces (BMI) for experimental and clinical studies have experienced massive growth. Developing suitable materials for bioelectronic devices to be used for real-time recording and data digitalizing has three important necessitates which should be covered. Biocompatibility, electrical conductivity, and having mechanical properties similar to soft brain tissue to decrease mechanical mismatch should be adopted for all materials. In this review, inorganic nanoparticles and intrinsically conducting polymers are discussed to impart electrical conductivity to systems, where soft materials such as hydrogels can offer reliable mechanical properties and a biocompatible substrate. Interpenetrating hydrogel networks offer more mechanical stability and provide a path for incorporating polymers with desired properties into one strong network. Promising fabrication methods, like electrospinning and additive manufacturing, allow scientists to customize designs for each application and reach the maximum potential for the system. In the near future, it is desired to fabricate biohybrid conducting polymer-based interfaces loaded with cells, giving the opportunity for simultaneous stimulation and regeneration. Developing multi-modal BMIs, Using artificial intelligence and machine learning to design advanced materials are among the future goals for this field. This article is categorized under: Therapeutic Approaches and Drug Discovery > Nanomedicine for Neurological Disease.
33Multi-functional 3D printed hydrogel electrodes for brain-computer interfaces and wearable sensing.PubMed
Xinyu Wu, Haorui Ge, Wei Zhao, et al.
J Colloid Interface Sci. 2026 Feb 15;704(Pt 2):139418. doi: 10.1016/j.jcis.2025.139418. Epub 2025 Nov 8.
In this study, a 3D printing-based polyvinyl alcohol (PVA)/κ-carrageenan (κ-CA)/ carbon nanotubes (CNTs) hydrogel composite (referred to as PCC) was developed for the fabrication of flexible electrodes, targeting applications in brain-computer interfaces (BCIs) and wearable strain sensors. The hydrogel composite exhibited excellent mechanical properties, including a tensile strength of 633 kPa, an elastic modulus of 243 kPa, and a maximum tensile strain of 283 %. In BCI tests, the PCC hydrogel electrode achieved a scalp contact impedance of 76.08 kΩ across five channels, with signal quality comparable to wet electrodes (3.06 μV at 13 Hz stimulation) and significantly higher than dry electrodes (2.16 μV). The decoding accuracy for the PCC hydrogel electrode was 78.2 % with a 1.25 s window length, comparable to the wet electrode, and the information transfer rate (ITR) reached 71.3 bits/min. Furthermore, the hydrogel demonstrated excellent strain sensing performance, with a gauge factor (GF) of 2.7 in the 0-75 % strain range and fast self-recovery, making it a promising material for dynamic wearable sensing devices. This work highlights the successful integration of material optimization and structural design, offering a new approach for development of next-generation flexible bioelectronic devices.
34A Lubricated Nonimmunogenic Neural Probe for Acute Insertion Trauma Minimization and Long-Term Signal Recording.PubMed
Yeontaek Lee, Hyogeun Shin, Dongwon Lee, et al.
Adv Sci (Weinh). 2021 Aug;8(15):e2100231. doi: 10.1002/advs.202100231. Epub 2021 Jun 3.
Brain-machine interfaces (BMIs) that link the brain to a machine are promising for the treatment of neurological disorders through the bi-directional translation of neural information over extended periods. However, the longevity of such implanted devices remains limited by the deterioration of their signal sensitivity over time due to acute inflammation from insertion trauma and chronic inflammation caused by the foreign body reaction. To address this challenge, a lubricated surface is fabricated to minimize friction during insertion and avoid immunogenicity during neural signal recording. Reduced friction force leads to 86% less impulse on the brain tissue, and thus immediately increases the number of measured signal electrodes by 102% upon insertion. Furthermore, the signal measurable period increases from 8 to 16 weeks due to the prevention of gliosis. By significantly reducing insertion damage and the foreign body reaction, the lubricated immune-stealthy probe surface (LIPS) can maximize the longevity of implantable BMIs.
35Design and verification of a wearable wireless 64-channel high-resolution EEG acquisition system with wi-fi transmission.PubMed
Chin-Teng Lin, Yuhling Wang, Sheng-Fu Chen, et al.
Med Biol Eng Comput. 2023 Nov;61(11):3003-3019. doi: 10.1007/s11517-023-02879-y. Epub 2023 Aug 11.
Brain-computer interfaces (BCIs) allow communication between the brain and the external world. This type of technology has been extensively studied. However, BCI instruments with high signal quality are typically heavy and large. Thus, recording electroencephalography (EEG) signals is an inconvenient task. In recent years, system-on-chip (SoC) approaches have been integrated into BCI research, and sensors for wireless portable devices have been developed; however, there is still considerable work to be done. As neuroscience research has advanced, EEG signal analyses have come to require more accurate data. Due to the limited bandwidth of Bluetooth wireless transmission technology, EEG measurement systems with more than 16 channels must be used to reduce the sampling rate and prevent data loss. Therefore, the goal of this study was to develop a multichannel, high-resolution (24-bit), high-sampling-rate EEG BCI device that transmits signals via Wi-Fi. We believe that this system can be used in neuroscience research. The EEG acquisition system proposed in this work is based on a Cortex-M4 microcontroller with a Wi-Fi subsystem, providing a multichannel design and improved signal quality. This system is compatible with wet sensors, Ag/AgCl electrodes, and dry sensors. A LabVIEW-based user interface receives EEG data via Wi-Fi transmission and saves the raw data for offline analysis. In previous cognitive experiments, event tags have been communicated using Recommended Standard 232 (RS-232). The developed system was validated through event-related potential (ERP) and steady-state visually evoked potential (SSVEP) experiments. Our experimental results demonstrate that this system is suitable for recording EEG measurements and has potential in practical applications. The advantages of the developed system include its high sampling rate and high amplification. Additionally, in the future, Internet of Things (IoT) technology can be integrated into the system for remote real-time analysis or edge computing.
36Brain Wearables: Validation Toolkit for Ear-Level EEG Sensors.PubMed
Guilherme Correia, Michael J Crosse, Alejandro Lopez Valdes
Sensors (Basel). 2024 Feb 15;24(4):1226. doi: 10.3390/s24041226.
EEG-enabled earbuds represent a promising frontier in brain activity monitoring beyond traditional laboratory testing. Their discrete form factor and proximity to the brain make them the ideal candidate for the first generation of discrete non-invasive brain-computer interfaces (BCIs). However, this new technology will require comprehensive characterization before we see widespread consumer and health-related usage. To address this need, we developed a validation toolkit that aims to facilitate and expand the assessment of ear-EEG devices. The first component of this toolkit is a desktop application ("EaR-P Lab") that controls several EEG validation paradigms. This application uses the Lab Streaming Layer (LSL) protocol, making it compatible with most current EEG systems. The second element of the toolkit introduces an adaptation of the phantom evaluation concept to the domain of ear-EEGs. Specifically, it utilizes 3D scans of the test subjects' ears to simulate typical EEG activity around and inside the ear, allowing for controlled assessment of different ear-EEG form factors and sensor configurations. Each of the EEG paradigms were validated using wet-electrode ear-EEG recordings and benchmarked against scalp-EEG measurements. The ear-EEG phantom was successful in acquiring performance metrics for hardware characterization, revealing differences in performance based on electrode location. This information was leveraged to optimize the electrode reference configuration, resulting in increased auditory steady-state response (ASSR) power. Through this work, an ear-EEG evaluation toolkit is made available with the intention to facilitate the systematic assessment of novel ear-EEG devices from hardware to neural signal acquisition.
37Designing Multifunctional Microneedles in Biomedical Engineering: Materials, Methods, and Applications.PubMed
Ling Liu, Fangyan Wang, Xiang Chen, et al.
Int J Nanomedicine. 2025 Jul 4;20:8693-8728. doi: 10.2147/IJN.S531898. eCollection 2025.
This review focuses on the emerging technology of multifunctional microneedles (MNs) within the biomedical engineering (BME) field, highlighting their potential in drug delivery, diagnostics, and therapeutics. Previous studies have explored MNs in various applications; however, their diverse functionalities across different material types and advanced application domains have been rarely comprehensively explored. This review bridges this gap by providing insights into the application of MNs in materials science, drug delivery, diagnostic monitoring, and tissue engineering. The unique properties and skin effects of various inorganic (eg, silicon, metals) and organic materials (eg, polysaccharides, polymers, proteins) used in MNs are examined. The analysis emphasizes the advantages of different MN materials, ie, their biocompatibility, degradation rates, and application specificity. In addition, the preparation processes and application scenarios of each MN type, such as minimally invasive drug delivery in transdermal applications and their benefits in tissue engineering for promoting repair, regeneration, and precise delivery of cells and growth factors in tissues like skin, cartilage, muscle, bone, and nerves, are discussed. Furthermore, this review explores the innovative use of MNs in brain-computer interfaces-an area not yet thoroughly examined. This novel application offers significant opportunities in neuroscience and clinical practice. Overall, this review provides valuable insights into the current research landscape and unexplored areas of MNs, contributing to future advancements in BME.
38Group-level brain decoding with deep learning.PubMed
Richard Csaky, Mats W J van Es, Oiwi Parker Jones, et al.
Hum Brain Mapp. 2023 Dec 1;44(17):6105-6119. doi: 10.1002/hbm.26500. Epub 2023 Sep 27.
Decoding brain imaging data are gaining popularity, with applications in brain-computer interfaces and the study of neural representations. Decoding is typically subject-specific and does not generalise well over subjects, due to high amounts of between subject variability. Techniques that overcome this will not only provide richer neuroscientific insights but also make it possible for group-level models to outperform subject-specific models. Here, we propose a method that uses subject embedding, analogous to word embedding in natural language processing, to learn and exploit the structure in between-subject variability as part of a decoding model, our adaptation of the WaveNet architecture for classification. We apply this to magnetoencephalography data, where 15 subjects viewed 118 different images, with 30 examples per image; to classify images using the entire 1 s window following image presentation. We show that the combination of deep learning and subject embedding is crucial to closing the performance gap between subject- and group-level decoding models. Importantly, group models outperform subject models on low-accuracy subjects (although slightly impair high-accuracy subjects) and can be helpful for initialising subject models. While we have not generally found group-level models to perform better than subject-level models, the performance of group modelling is expected to be even higher with bigger datasets. In order to provide physiological interpretation at the group level, we make use of permutation feature importance. This provides insights into the spatiotemporal and spectral information encoded in the models. All code is available on GitHub (https://github.com/ricsinaruto/MEG-group-decode).
39Machine Learning for Neural Decoding.PubMed
Joshua I Glaser, Ari S Benjamin, Raeed H Chowdhury, et al.
eNeuro. 2020 Aug 31;7(4). doi: 10.1523/ENEURO.0506-19.2020. Print 2020 Jul/Aug.
Despite rapid advances in machine learning tools, the majority of neural decoding approaches still use traditional methods. Modern machine learning tools, which are versatile and easy to use, have the potential to significantly improve decoding performance. This tutorial describes how to effectively apply these algorithms for typical decoding problems. We provide descriptions, best practices, and code for applying common machine learning methods, including neural networks and gradient boosting. We also provide detailed comparisons of the performance of various methods at the task of decoding spiking activity in motor cortex, somatosensory cortex, and hippocampus. Modern methods, particularly neural networks and ensembles, significantly outperform traditional approaches, such as Wiener and Kalman filters. Improving the performance of neural decoding algorithms allows neuroscientists to better understand the information contained in a neural population and can help to advance engineering applications such as brain-machine interfaces. Our code package is available at github.com/kordinglab/neural_decoding.
40Blind source separation of event-related potentials using a recurrent neural network.PubMed
Jamie A O'Reilly, Hassapong Sunthornwiriya-Amon, Naradith Aparprasith, et al.
bioRxiv. 2024 Apr 28:2024.04.23.590794. doi: 10.1101/2024.04.23.590794.
Event-related potentials (ERPs) are a superposition of electric potential differences generated by neurophysiological activity associated with psychophysical events. Spatiotemporal dissociation of these signal sources can supplement conventional ERP analysis and improve source localization. However, results from established source separation methods applied to ERPs can be challenging to interpret. Hence, we have developed a recurrent neural network (RNN) method for blind source separation. The RNN transforms input step pulse signals representing events into corresponding ERP difference waveforms. Source waveforms are obtained from penultimate layer units and scalp maps are obtained from feed-forward output layer weights that project these source waveforms onto EEG electrode amplitudes. An interpretable, sparse source representation is achieved by incorporating L1 regularization of signals obtained from the penultimate layer of the network during training. This RNN method was applied to four ERP difference waveforms (MMN, N170, N400, P3) from the open-access ERP CORE database, and independent component analysis (ICA) was applied to the same data for comparison. The RNN decomposed these ERPs into eleven spatially and temporally separate sources that were less noisy, tended to be more ERP-specific, and were less similar to each other than ICA-derived sources. The RNN sources also had less ambiguity between source waveform amplitude, scalp potential polarity, and equivalent current dipole orientation than ICA sources. In conclusion, the proposed RNN blind source separation method can be effectively applied to grand-average ERP difference waves and holds promise for further development as a computational model of event-related neural signals.
41Spatio-temporal transformers for decoding neural movement control.PubMed
Benedetta Candelori, Giampiero Bardella, Indro Spinelli, et al.
J Neural Eng. 2025 Feb 4;22(1). doi: 10.1088/1741-2552/adaef0.
. Deep learning tools applied to high-resolution neurophysiological data have significantly progressed, offering enhanced decoding, real-time processing, and readability for practical applications. However, the design of artificial neural networks to analyze neural activityremains a challenge, requiring a delicate balance between efficiency in low-data regimes and the interpretability of the results.. To address this challenge, we introduce a novel specialized transformer architecture to analyze single-neuron spiking activity. The model is tested on multi-electrode recordings from the dorsal premotor cortex of non-human primates performing a motor inhibition task.. The proposed architecture provides an early prediction of the correct movement direction, achieving accurate results no later than 230 ms after the Go signal presentation across animals. Additionally, the model can forecast whether the movement will be generated or withheld before a stop signal, unattended, is actually presented. To further understand the internal dynamics of the model, we compute the predicted correlations between time steps and between neurons at successive layers of the architecture, with the evolution of these correlations mirrors findings from previous theoretical analyses.. Overall, our framework provides a comprehensive use case for the practical implementation of deep learning tools in motor control research, highlighting both the predictive capabilities and interpretability of the proposed architecture.
42A Two-Stage Deep Learning Approach for EEG Artifact Removal and Classification: Towards Reliable Wearable Applications.PubMed
Andrea Farabbi, Filippo Ballabio, Matteo Rossi, et al.
Annu Int Conf IEEE Eng Med Biol Soc. 2025 Jul;2025:1-5. doi: 10.1109/EMBC58623.2025.11254976.
EEG artifact removal remains a critical challenge in neural signal processing. In this paper, we present a novel two-stage approach combining a modified IC-UNet architecture for artifact removal with a modified VGGNet for artifact type identification. The system automatically triggers the classification stage when the difference between original and denoised signals exceeds a learned threshold, enabling the classification of ocular artifacts (eye blinks and saccadic movements) in the original signals. The denoising stage employs parallel encoding paths with channel-specific feature extraction, followed by a shared bottleneck and decoder network. The system was evaluated using EEG data from subjects performing controlled eye blink and saccadic movement tasks. The denoising network achieves high correlation values between predicted and ground truth signals, particularly in temporal and specific frontal regions (T5: 0.86 ± 0.01, T6: 0.85 ± 0.01, F3: 0.83 ± 0.01). The classification network shows excellent performance, achieving 99.35% accuracy on the test set with only four misclassifications out of 620 cases.Clinical relevance- This study demonstrates the feasibility of accurate artifact removal and classification in temporal and behind-the-ear EEG recordings, which is particularly relevant for the development of wearable EEG devices for continuous monitoring and hybrid BCI systems.
43Time-Variant Linear Discriminant Analysis Improves Hand Gesture and Finger Movement Decoding for Invasive Brain-Computer Interfaces.PubMed
Johannes Gruenwald, Andrei Znobishchev, Christoph Kapeller, et al.
Front Neurosci. 2019 Sep 26;13:901. doi: 10.3389/fnins.2019.00901. eCollection 2019.
Invasive brain-computer interfaces yield remarkable performance in a multitude of applications. For classification experiments, high-gamma bandpower features and linear discriminant analysis (LDA) are commonly used due to simplicity and robustness. However, LDA is inherently static and not suited to account for transient information that is typically present in high-gamma features. To resolve this issue, we here present an extension of LDA to the time-variant feature space. We call this method (TVLDA). It intrinsically provides a feature reduction stage, which makes external approaches thereto obsolete, such as feature selection techniques or common spatial patterns (CSPs). As well, we propose a time-domain whitening stage which equalizes the pronounced 1/-shape of the typical brain-wave spectrum. We evaluated our proposed architecture based on recordings from 15 epilepsy patients with temporarily implanted subdural grids, who participated in additional research experiments besides clinical treatment. The experiments featured two different motor tasks involving three high-level gestures and individual finger movement. We used log-transformed bandpower features from the high-gamma band (50-300 Hz, excluding power-line harmonics) for classification. On average, whitening improved the classification performance by about 11%. On whitened data, TVLDA outperformed LDA with feature selection by 11.8%, LDA with CSPs by 13.9%, and regularized LDA with vectorized features by 16.4%. At the same time, TVLDA only required one or two internal features to achieve this. TVLDA provides stable results even if very few trials are available. It is easy to implement, fully automatic and deterministic. Due to its low complexity, TVLDA is suited for real-time brain-computer interfaces. Training is done in less than a second. TVLDA performed particularly well in experiments with data from high-density electrode arrays. For example, the three high-level gestures were correctly identified at a rate of 99% over all subjects. Similarly, the decoding accuracy of individual fingers was 96% on average over all subjects. To our knowledge, these mean accuracies are the highest ever reported for three-class and five-class motor-control BCIs.
44Continuous 2D trajectory decoding from attempted movement: across-session performance in able-bodied and feasibility in a spinal cord injured participant.PubMed
Hannah S Pulferer, Brynja Ásgeirsdóttir, Valeria Mondini, et al.
J Neural Eng. 2022 May 9;19(3). doi: 10.1088/1741-2552/ac689f.
. In people with a cervical spinal cord injury (SCI) or degenerative diseases leading to limited motor function, restoration of upper limb movement has been a goal of the brain-computer interface field for decades. Recently, research from our group investigated non-invasive and real-time decoding of continuous movement in able-bodied participants from low-frequency brain signals during a target-tracking task. To advance our setup towards motor-impaired end users, we consequently chose a new paradigm based on attempted movement.. Here, we present the results of two studies. During the first study, data of ten able-bodied participants completing a target-tracking/shape-tracing task on-screen were investigated in terms of improvements in decoding performance due to user training. In a second study, a spinal cord injured participant underwent the same tasks. To investigate the merit of employing attempted movement in end users with SCI, data of the spinal cord injured participant were recorded twice; once within an observation-only condition, and once while simultaneously attempting movement.. We observed mean correlations well above chance level for continuous motor decoding based on attempted movement in able-bodied participants. Additionally, no global improvement over three sessions within five days, both in sensor and in source space, could be observed across all participants and movement parameters. In the participant with SCI, decoding performance well above chance was found.. No presence of a learning effect in continuous attempted movement decoding in able-bodied participants could be observed. In contrast, non-significantly varying decoding patterns may promote the use of source space decoding in terms of generalized decoders utilizing transfer learning. Furthermore, above-chance correlations for attempted movement decoding ranging between those of observation only and executed movement were seen in one spinal cord injured participant, suggesting attempted movement decoding as a possible link between feasibility studies in able-bodied and actual applications in motor impaired end users.
45Multimodal fusion of muscle and brain signals for a hybrid-BCI.PubMed
Robert Leeb, Hesam Sagha, Ricardo Chavarriaga, et al.
Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:4343-6. doi: 10.1109/IEMBS.2010.5626233.
Practical Brain-Computer Interfaces (BCIs) for disabled people should allow them to use all their remaining functionalities as control possibilities. Sometimes these people have residual activity of their muscles, most likely in the morning when they are not exhausted. In this work we fuse electromyographic (EMG) with electroencephalographic (EEG) activity in the framework of a so called "Hybrid-BCI" (hBCI) approach. Thereby, subjects could achieve a good control of their hBCI independently of their level of muscular fatigue. Furthermore, although EMG alone yields good performance, it is outperformed by the hybrid fusing of EEG and EMG. Two different fusion techniques are explored showing graceful performance degradation in the case of signal attenuation. Such a system allows a very reliable control and a smooth handover if the subjects get exhausted or fatigued during the day.
46Enhanced Recognition of Amputated Wrist and Hand Movements by Deep Learning Method Using Multimodal Fusion of Electromyography and Electroencephalography.PubMed
Sehyeon Kim, Dae Youp Shin, Taekyung Kim, et al.
Sensors (Basel). 2022 Jan 16;22(2):680. doi: 10.3390/s22020680.
Motion classification can be performed using biometric signals recorded by electroencephalography (EEG) or electromyography (EMG) with noninvasive surface electrodes for the control of prosthetic arms. However, current single-modal EEG and EMG based motion classification techniques are limited owing to the complexity and noise of EEG signals, and the electrode placement bias, and low-resolution of EMG signals. We herein propose a novel system of two-dimensional (2D) input image feature multimodal fusion based on an EEG/EMG-signal transfer learning (TL) paradigm for detection of hand movements in transforearm amputees. A feature extraction method in the frequency domain of the EEG and EMG signals was adopted to establish a 2D image. The input images were used for training on a model based on the convolutional neural network algorithm and TL, which requires 2D images as input data. For the purpose of data acquisition, five transforearm amputees and nine healthy controls were recruited. Compared with the conventional single-modal EEG signal trained models, the proposed multimodal fusion method significantly improved classification accuracy in both the control and patient groups. When the two signals were combined and used in the pretrained model for EEG TL, the classification accuracy increased by 4.18-4.35% in the control group, and by 2.51-3.00% in the patient group.
47Advanced multiscale machine learning for nerve conduction velocity analysis.PubMed
Hossein Sadeghi
Sci Rep. 2025 Jul 2;15(1):23399. doi: 10.1038/s41598-025-08367-7.
This paper presents an advanced machine learning (ML) framework for precise nerve conduction velocity (NCV) analysis, integrating multiscale signal processing with physiologically-constrained deep learning. Our approach addresses three fundamental limitations of conventional NCV techniques: (1) oversimplified nerve fiber modeling, (2) temperature sensitivity, and (3) static measurement interpretation. The proposed framework combines: (i) entropy-optimized wavelet analysis for adaptive multiscale signal decomposition, (ii) thermodynamically-regularized neural networks incorporating Arrhenius kinetics, and (iii) stochastic progression models for uncertainty-aware longitudinal tracking. Through data extracted from prior studies in this field, rigorously validated across 1842 patients from 28 medical centers, we demonstrate significant improvements: 23.4% enhancement in motor NCV accuracy ([Formula: see text]) and 28.7% for sensory fibers. The framework maintains physiological interpretability while achieving superior performance through: (a) wavelet-optimized resolution scales (2-8 ms for motor, 0.5-2 ms for sensory fibers), (b) temperature compensation accurate to [Formula: see text] across 20-[Formula: see text], and (c) probabilistic progression tracking with 88.9% treatment response prediction accuracy. This work establishes new standards for ML applications in clinical neurophysiology by rigorously combining biophysical first principles with data-driven learning, offering both theoretical advances and immediate clinical utility for neuropathy diagnosis and monitoring.
48Rehabilitation with brain-computer interface and upper limb motor function in ischemic stroke: A randomized controlled trial.PubMed
Anxin Wang, Xue Tian, Di Jiang, et al.
Med. 2024 Jun 14;5(6):559-569.e4. doi: 10.1016/j.medj.2024.02.014. Epub 2024 Apr 19.
BACKGROUND: Upper limb motor dysfunction is a major problem in the rehabilitation of patients with stroke. Brain-computer interface (BCI) is a kind of communication system that converts the "ideas" in the brain into instructions and has been used in stroke rehabilitation. This study aimed to investigate the efficacy and safety of BCI in rehabilitation training on upper limb motor function among patients with ischemic stroke. METHODS: This was an investigator-initiated, multicenter, randomized, open-label, blank-controlled clinical trial with blinded outcome assessment conducted at 17 centers in China. Patients were assigned in a 1:1 ratio to the BCI group or the control group based on traditional rehabilitation training. The primary efficacy outcome is the difference in improvement of the Fugl-Meyer Assessment upper extremity (FMA-UE) score between two groups at month 1 after randomization. The safety outcomes were any adverse events within 3 months. FINDINGS: A total of 296 patients with ischemic stroke were enrolled and randomly allocated to the BCI group (n = 150) and the control group (n = 146). The primary efficacy outcomes of FMA-UE score change from baseline to 1 month were 13.17 (95% confidence interval [CI], 11.56-14.79) in the BCI group and 9.83 (95% CI, 8.19-11.47) in the control group (mean difference between groups was 3.35; 95% CI, 1.05-5.65; p = 0.0045). Adverse events occurred in 33 patients (22.00%) in the BCI group and in 31 patients (21.23%) in the control group. CONCLUSIONS: BCI rehabilitation training can further improve upper limb motor function based on traditional rehabilitation training in patients with ischemic stroke. This study was registered at ClinicalTrials.gov: NCT04387474. FUNDING: This work was supported by the National Key R&D Program of China (2018YFC1312903), the National Key Research and Development Program of China (2022YFC3600600), the Training Fund for Open Projects at Clinical Institutes and Departments of Capital Medical University (CCMU2022ZKYXZ009), the Beijing Natural Science Foundation Haidian original innovation joint fund (L222123), the Fund for Young Talents of Beijing Medical Management Center (QML20230505), and the high-level public health talents (xuekegugan-02-47).
49Effects of brain-computer interface based training on post-stroke upper-limb rehabilitation: a meta-analysis.PubMed
Dan Li, Ruoyu Li, Yunping Song, et al.
J Neuroeng Rehabil. 2025 Mar 3;22(1):44. doi: 10.1186/s12984-025-01588-x.
BACKGROUND: Previous research has used the brain-computer interface (BCI) to promote upper-limb motor rehabilitation. However, the results of these studies were variable, leaving efficacy unclear. OBJECTIVES: This review aims to evaluate the effects of BCI-based training on post-stroke upper-limb rehabilitation and identify potential factors that may affect the outcome. DESIGN: A meta-analysis including all available randomized-controlled clinical trials (RCTs) that reported the efficacy of BCI-based training on upper-limb motor rehabilitation after stroke. DATA SOURCES AND METHODS: We searched PubMed, Cochrane Library, and Web of Science before September 15, 2024, for relevant studies. The primary efficacy outcome was the Fugl-Meyer Assessment-Upper extremity (FMA-UE). RevMan 5.4.1 with a random effect model was used for data synthesis and analysis. Mean difference (MD) and 95% confidence interval (95%CI) were calculated. RESULTS: Twenty-one RCTs (n = 886 patients) were reviewed in the meta-analysis. Compared with control, BCI-based training exerted significant effects on FMA-UE (MD = 3.69, 95%CI 2.41-4.96, P < 0.00001, moderate-quality evidence), Wolf Motor Function Test (WMFT) (MD = 5.00, 95%CI 2.14-7.86, P = 0.0006, low-quality evidence), and Action Research Arm Test (ARAT) (MD = 2.04, 95%CI 0.25-3.82, P = 0.03, high-quality evidence). Additionally, BCI-based training was effective on FMA-UE for both subacute (MD = 4.24, 95%CI 1.81-6.67, P = 0.0006) and chronic patients (MD = 2.63, 95%CI 1.50-3.76, P < 0.00001). BCI combined with functional electrical stimulation (FES) (MD = 4.37, 95%CI 3.09-5.65, P < 0.00001), robots (MD = 2.87, 95%CI 0.69-5.04, P = 0.010), and visual feedback (MD = 4.46, 95%CI 0.24-8.68, P = 0.04) exhibited significant effects on FMA-UE. BCI combined with FES significantly improved FMA-UE for both subacute (MD = 5.31, 95%CI 2.58-8.03, P = 0.0001) and chronic patients (MD = 3.71, 95%CI 2.44-4.98, P < 0.00001), and BCI combined with robots was effective for chronic patients (MD = 1.60, 95%CI 0.15-3.05, P = 0.03). Better results may be achieved with daily training sessions ranging from 20 to 90 min, conducted 2-5 sessions per week for 3-4 weeks. CONCLUSIONS: BCI-based training may be a reliable rehabilitation program to improve upper-limb motor impairment and function. TRIAL REGISTRATION: PROSPERO registration ID: CRD42022383390.
50Brain-actuated functional electrical stimulation elicits lasting arm motor recovery after stroke.PubMed
A Biasiucci, R Leeb, I Iturrate, et al.
Nat Commun. 2018 Jun 20;9(1):2421. doi: 10.1038/s41467-018-04673-z.
Brain-computer interfaces (BCI) are used in stroke rehabilitation to translate brain signals into intended movements of the paralyzed limb. However, the efficacy and mechanisms of BCI-based therapies remain unclear. Here we show that BCI coupled to functional electrical stimulation (FES) elicits significant, clinically relevant, and lasting motor recovery in chronic stroke survivors more effectively than sham FES. Such recovery is associated to quantitative signatures of functional neuroplasticity. BCI patients exhibit a significant functional recovery after the intervention, which remains 6-12 months after the end of therapy. Electroencephalography analysis pinpoints significant differences in favor of the BCI group, mainly consisting in an increase in functional connectivity between motor areas in the affected hemisphere. This increase is significantly correlated with functional improvement. Results illustrate how a BCI-FES therapy can drive significant functional recovery and purposeful plasticity thanks to contingent activation of body natural efferent and afferent pathways.
51Efficacy of brain-computer interface training with motor imagery-contingent feedback in improving upper limb function and neuroplasticity among persons with chronic stroke: a double-blinded, parallel-group, randomized controlled trial.PubMed
Myeong Sun Kim, Hyunju Park, Ilho Kwon, et al.
J Neuroeng Rehabil. 2025 Jan 6;22(1):1. doi: 10.1186/s12984-024-01535-2.
BACKGROUND: Brain-computer interface (BCI) technology can enhance neural plasticity and motor recovery in persons with stroke. However, the effects of BCI training with motor imagery (MI)-contingent feedback versus MI-independent feedback remain unclear. This study aimed to investigate whether the contingent connection between MI-induced brain activity and feedback influences functional and neural plasticity outcomes. We hypothesized that BCI training, with MI-contingent feedback, would result in greater improvements in upper limb function and neural plasticity compared to BCI training, with MI-independent feedback. METHODS: This randomized controlled trial included persons with chronic stroke who underwent BCI training involving functional electrical stimulation feedback on the affected wrist extensor. Primary outcomes included the Medical Research Council (MRC) scale score for muscle strength in the wrist extensor (MRC-WE) and active range of motion in wrist extension (AROM-WE). Resting-state electroencephalogram recordings were used to assess neural plasticity. RESULTS: Compared to the MI-independent feedback BCI group, the MI-contingent feedback BCI group showed significantly greater improvements in MRC-WE scores (mean difference = 0.52, 95% CI = 0.03-1.00, p = 0.036) and demonstrated increased AROM-WE at 4 weeks post-intervention (p = 0.019). Enhanced functional connectivity in the affected hemisphere was observed in the MI-contingent feedback BCI group, correlating with MRC-WE and Fugl-Meyer assessment-distal scores. Improvements were also observed in the unaffected hemisphere's functional connectivity. CONCLUSIONS: BCI training with MI-contingent feedback is more effective than MI-independent feedback in improving AROM-WE, MRC, and neural plasticity in individuals with chronic stroke. BCI technology could be a valuable addition to conventional rehabilitation for stroke survivors, enhancing recovery outcomes. TRIAL REGISTRATION: CRIS (KCT0009013).
52Brain-Computer Interface-Controlled Exoskeletons in Clinical Neurorehabilitation: Ready or Not?PubMed
Annalisa Colucci, Mareike Vermehren, Alessia Cavallo, et al.
Neurorehabil Neural Repair. 2022 Dec;36(12):747-756. doi: 10.1177/15459683221138751. Epub 2022 Nov 25.
The development of brain-computer interface-controlled exoskeletons promises new treatment strategies for neurorehabilitation after stroke or spinal cord injury. By converting brain/neural activity into control signals of wearable actuators, brain/neural exoskeletons (B/NEs) enable the execution of movements despite impaired motor function. Beyond the use as assistive devices, it was shown that-upon repeated use over several weeks-B/NEs can trigger motor recovery, even in chronic paralysis. Recent development of lightweight robotic actuators, comfortable and portable real-world brain recordings, as well as reliable brain/neural control strategies have paved the way for B/NEs to enter clinical care. Although B/NEs are now technically ready for broader clinical use, their promotion will critically depend on early adopters, for example, research-oriented physiotherapists or clinicians who are open for innovation. Data collected by early adopters will further elucidate the underlying mechanisms of B/NE-triggered motor recovery and play a key role in increasing efficacy of personalized treatment strategies. Moreover, early adopters will provide indispensable feedback to the manufacturers necessary to further improve robustness, applicability, and adoption of B/NEs into existing therapy plans.
53The potential of robotics: A systematic review of neuroplastic changes following advanced lower limb rehabilitation in neurological disorders.PubMed
Rocco Salvatore Calabrò, Andrea Calderone, Laura Simoncini, et al.
Neurosci Biobehav Rev. 2025 Nov 8;180:106459. doi: 10.1016/j.neubiorev.2025.106459.
BACKGROUND: Neurological diseases are among the most common pathologies that strongly influence a person's ability to walk and move, affecting the lower extremities. They disrupt motor brain networks that enable precise movement, leading to deficits in gait, balance, and coordination; while conventional therapies remain essential, advances in robotic technologies show growing promise for rehabilitation. AIM OF REVIEW: This systematic review aims to investigate the role of robotic rehabilitation in improving neuroplasticity and motor outcomes for individuals with neurological disorders, with a particular focus on studies incorporating neurophysiological or neuroimaging techniques to assess neuroplastic changes and their long-term impact on recovery. KEY SCIENTIFIC CONCEPTS OF REVIEW: A systematic review was carried out utilizing an online search of articles from 2014 to 2025 on the PubMed, Web of Science, Cochrane Library, Embase, EBSCOhost, and Scopus databases in accordance with PRISMA guidelines. Studies were chosen based on predetermined inclusion criteria, with an emphasis on robotic rehabilitation therapies targeted at improving neuroplasticity in lower limb rehabilitation for people with neurological conditions. This review has been registered on Prospero with the following number: CRD42025640347. The search identified 12,769 records; after screening and eligibility assessment, 25 studies met inclusion criteria. Studies demonstrate that robot-assisted gait training (RAGT) and exoskeleton-based therapies improve motor function, gait, balance, and neuroplasticity across stroke, spinal cord injury, cerebral palsy, and brain injury populations. Adjunctive approaches such as brain-computer interface (BCI) integration, virtual reality feedback, and neuromodulation further enhance outcomes, with increases in cortical activation and improvements in functional connectivity supported by convergent neurophysiological and neuroimaging data; changes in corticospinal excitability are also reported. Taken together, robotic interventions, often combined with neuromodulation or virtual reality (VR), appear to catalyze neuroplasticity in ways that align with clinically meaningful gains. These findings underscore their transformative potential for tailored, multimodal rehabilitation strategies in neurological recovery.
54Restoration of reaching and grasping movements through brain-controlled muscle stimulation in a person with tetraplegia: a proof-of-concept demonstration.PubMed
A Bolu Ajiboye, Francis R Willett, Daniel R Young, et al.
Lancet. 2017 May 6;389(10081):1821-1830. doi: 10.1016/S0140-6736(17)30601-3. Epub 2017 Mar 28.
BACKGROUND: People with chronic tetraplegia, due to high-cervical spinal cord injury, can regain limb movements through coordinated electrical stimulation of peripheral muscles and nerves, known as functional electrical stimulation (FES). Users typically command FES systems through other preserved, but unrelated and limited in number, volitional movements (eg, facial muscle activity, head movements, shoulder shrugs). We report the findings of an individual with traumatic high-cervical spinal cord injury who coordinated reaching and grasping movements using his own paralysed arm and hand, reanimated through implanted FES, and commanded using his own cortical signals through an intracortical brain-computer interface (iBCI). METHODS: We recruited a participant into the BrainGate2 clinical trial, an ongoing study that obtains safety information regarding an intracortical neural interface device, and investigates the feasibility of people with tetraplegia controlling assistive devices using their cortical signals. Surgical procedures were performed at University Hospitals Cleveland Medical Center (Cleveland, OH, USA). Study procedures and data analyses were performed at Case Western Reserve University (Cleveland, OH, USA) and the US Department of Veterans Affairs, Louis Stokes Cleveland Veterans Affairs Medical Center (Cleveland, OH, USA). The study participant was a 53-year-old man with a spinal cord injury (cervical level 4, American Spinal Injury Association Impairment Scale category A). He received two intracortical microelectrode arrays in the hand area of his motor cortex, and 4 months and 9 months later received a total of 36 implanted percutaneous electrodes in his right upper and lower arm to electrically stimulate his hand, elbow, and shoulder muscles. The participant used a motorised mobile arm support for gravitational assistance and to provide humeral abduction and adduction under cortical control. We assessed the participant's ability to cortically command his paralysed arm to perform simple single-joint arm and hand movements and functionally meaningful multi-joint movements. We compared iBCI control of his paralysed arm with that of a virtual three-dimensional arm. This study is registered with ClinicalTrials.gov, number NCT00912041. FINDINGS: The intracortical implant occurred on Dec 1, 2014, and we are continuing to study the participant. The last session included in this report was Nov 7, 2016. The point-to-point target acquisition sessions began on Oct 8, 2015 (311 days after implant). The participant successfully cortically commanded single-joint and coordinated multi-joint arm movements for point-to-point target acquisitions (80-100% accuracy), using first a virtual arm and second his own arm animated by FES. Using his paralysed arm, the participant volitionally performed self-paced reaches to drink a mug of coffee (successfully completing 11 of 12 attempts within a single session 463 days after implant) and feed himself (717 days after implant). INTERPRETATION: To our knowledge, this is the first report of a combined implanted FES+iBCI neuroprosthesis for restoring both reaching and grasping movements to people with chronic tetraplegia due to spinal cord injury, and represents a major advance, with a clear translational path, for clinically viable neuroprostheses for restoration of reaching and grasping after paralysis. FUNDING: National Institutes of Health, Department of Veterans Affairs.
55Training with noninvasive brain-machine interface, tactile feedback, and locomotion to enhance neurological recovery in individuals with complete paraplegia: a randomized pilot study.PubMed
Miguel A L Nicolelis, Eduardo J L Alho, Ana R C Donati, et al.
Sci Rep. 2022 Nov 29;12(1):20545. doi: 10.1038/s41598-022-24864-5.
In recent years, our group and others have reported multiple cases of consistent neurological recovery in people with spinal cord injury (SCI) following a protocol that integrates locomotion training with brain machine interfaces (BMI). The primary objective of this pilot study was to compare the neurological outcomes (motor, tactile, nociception, proprioception, and vibration) in both an intensive assisted locomotion training (LOC) and a neurorehabilitation protocol integrating assisted locomotion with a noninvasive brain-machine interface (L + BMI), virtual reality, and tactile feedback. We also investigated whether individuals with chronic-complete SCI could learn to perform leg motor imagery. We ran a parallel two-arm randomized pilot study; the experiments took place in São Paulo, Brazil. Eight adults sensorimotor-complete (AIS A) (all male) with chronic (> 6 months) traumatic spinal SCI participated in the protocol that was organized in two blocks of 14 weeks of training and an 8-week follow-up. The participants were allocated to either the LOC group (n = 4) or L + BMI group (n = 4) using block randomization (blinded outcome assessment). We show three important results: (i) locomotion training alone can induce some level of neurological recovery in sensorimotor-complete SCI, and (ii) the recovery rate is enhanced when such locomotion training is associated with BMI and tactile feedback (∆Mean Lower Extremity Motor score improvement for LOC = + 2.5, L + B = + 3.5; ∆Pinprick score: LOC = + 3.75, L + B = + 4.75 and ∆Tactile score LOC = + 4.75, L + B = + 9.5). (iii) Furthermore, we report that the BMI classifier accuracy was significantly above the chance level for all participants in L + B group. Our study shows potential for sensory and motor improvement in individuals with chronic complete SCI following a protocol with BMIs and locomotion therapy. We report no dropouts nor adverse events in both subgroups participating in the study, opening the possibility for a more definitive clinical trial with a larger cohort of people with SCI.Trial registration: http://www.ensaiosclinicos.gov.br/ identifier RBR-2pb8gq.
56Brain-controlled neuromuscular stimulation to drive neural plasticity and functional recovery.PubMed
C Ethier, J A Gallego, L E Miller
Curr Opin Neurobiol. 2015 Aug;33:95-102. doi: 10.1016/j.conb.2015.03.007. Epub 2015 Mar 28.
There is mounting evidence that appropriately timed neuromuscular stimulation can induce neural plasticity and generate functional recovery from motor disorders. This review addresses the idea that coordinating stimulation with a patient's voluntary effort might further enhance neurorehabilitation. Studies in cell cultures and behaving animals have delineated the rules underlying neural plasticity when single neurons are used as triggers. However, the rules governing more complex stimuli and larger networks are less well understood. We argue that functional recovery might be optimized if stimulation were modulated by a brain machine interface, to match the details of the patient's voluntary intent. The potential of this novel approach highlights the need for a better understanding of the complex rules underlying this form of plasticity.
57Motor imagery-based brain-computer interface rehabilitation programs enhance upper extremity performance and cortical activation in stroke patients.PubMed
Zhen-Zhen Ma, Jia-Jia Wu, Zhi Cao, et al.
J Neuroeng Rehabil. 2024 May 29;21(1):91. doi: 10.1186/s12984-024-01387-w.
BACKGROUND: The most challenging aspect of rehabilitation is the repurposing of residual functional plasticity in stroke patients. To achieve this, numerous plasticity-based clinical rehabilitation programs have been developed. This study aimed to investigate the effects of motor imagery (MI)-based brain-computer interface (BCI) rehabilitation programs on upper extremity hand function in patients with chronic hemiplegia. DESIGN: A 2010 Consolidated Standards for Test Reports (CONSORT)-compliant randomized controlled trial. METHODS: Forty-six eligible stroke patients with upper limb motor dysfunction participated in the study, six of whom dropped out. The patients were randomly divided into a BCI group and a control group. The BCI group received BCI therapy and conventional rehabilitation therapy, while the control group received conventional rehabilitation only. The Fugl-Meyer Assessment of the Upper Extremity (FMA-UE) score was used as the primary outcome to evaluate upper extremity motor function. Additionally, functional magnetic resonance imaging (fMRI) scans were performed on all patients before and after treatment, in both the resting and task states. We measured the amplitude of low-frequency fluctuation (ALFF), regional homogeneity (ReHo), z conversion of ALFF (zALFF), and z conversion of ReHo (ReHo) in the resting state. The task state was divided into four tasks: left-hand grasping, right-hand grasping, imagining left-hand grasping, and imagining right-hand grasping. Finally, meaningful differences were assessed using correlation analysis of the clinical assessments and functional measures. RESULTS: A total of 40 patients completed the study, 20 in the BCI group and 20 in the control group. Task-related blood-oxygen-level-dependent (BOLD) analysis showed that when performing the motor grasping task with the affected hand, the BCI group exhibited significant activation in the ipsilateral middle cingulate gyrus, precuneus, inferior parietal gyrus, postcentral gyrus, middle frontal gyrus, superior temporal gyrus, and contralateral middle cingulate gyrus. When imagining a grasping task with the affected hand, the BCI group exhibited greater activation in the ipsilateral superior frontal gyrus (medial) and middle frontal gyrus after treatment. However, the activation of the contralateral superior frontal gyrus decreased in the BCI group relative to the control group. Resting-state fMRI revealed increased zALFF in multiple cerebral regions, including the contralateral precentral gyrus and calcarine and the ipsilateral middle occipital gyrus and cuneus, and decreased zALFF in the ipsilateral superior temporal gyrus in the BCI group relative to the control group. Increased zReHo in the ipsilateral cuneus and contralateral calcarine and decreased zReHo in the contralateral middle temporal gyrus, temporal pole, and superior temporal gyrus were observed post-intervention. According to the subsequent correlation analysis, the increase in the FMA-UE score showed a positive correlation with the mean zALFF of the contralateral precentral gyrus (r = 0.425, P < 0.05), the mean zReHo of the right cuneus (r = 0.399, P < 0.05). CONCLUSION: In conclusion, BCI therapy is effective and safe for arm rehabilitation after severe poststroke hemiparesis. The correlation of the zALFF of the contralateral precentral gyrus and the zReHo of the ipsilateral cuneus with motor improvements suggested that these values can be used as prognostic measures for BCI-based stroke rehabilitation. We found that motor function was related to visual and spatial processing, suggesting potential avenues for refining treatment strategies for stroke patients. TRIAL REGISTRATION: The trial is registered in the Chinese Clinical Trial Registry (number ChiCTR2000034848, registered July 21, 2020).
58Effects of motor imagery based brain-computer interface on upper limb function and attention in stroke patients with hemiplegia: a randomized controlled trial.PubMed
Xiaolu Liu, Wendong Zhang, Weibo Li, et al.
BMC Neurol. 2023 Mar 31;23(1):136. doi: 10.1186/s12883-023-03150-5.
BACKGROUND: Seeking positive and comprehensive rehabilitation methods after stroke is an urgent problem to be solved, which is very important to improve the dysfunction of stroke. The aim of this study was to investigate the effects of motor imagery-based brain-computer interface training (MI-BCI) on upper limb function and attention in stroke patients with hemiplegia. METHODS: Sixty stroke patients with impairment of upper extremity function and decreased attention were randomly assigned to the control group (CR group) or the experimental group (BCI group) in a 1:1 ratio. Patients in the CR group received conventional rehabilitation. Patients in the BCI group received 20 min of MI-BCI training five times a week for 3 weeks (15 sessions) in addition to conventional rehabilitation. The primary outcome measures were the changes in Fugl-Meyer Motor Function Assessment of Upper Extremities (FMA-UE) and Attention Network Test (ANT) from baseline to 3 weeks. RESULTS: About 93% of the patients completed the allocated training. Compared with the CR group, among those in the BCI group, FMA-UE was increased by 8.0 points (95%CI, 5.0 to 10.0; P < 0.001). Alert network response time (32.4ms; 95%CI, 58.4 to 85.6; P < 0.001), orienting network response (5.6ms; 95%CI, 29.8 to 55.8; P = 0.010), and corrects number (8.0; 95%CI, 17.0 to 28.0; P < 0.001) also increased in the BCI group compared with the CR group. Additionally, the executive control network response time (- 105.9ms; 95%CI, - 68.3 to - 23.6; P = 0.002), the total average response time (- 244.8ms; 95%CI, - 155.8 to - 66.2; P = 0.002), and total time (- 122.0ms; 95%CI, - 80.0 to - 35.0; P = 0.001) were reduced in the BCI group compared with the CR group. CONCLUSION: MI-BCI combined with conventional rehabilitation training could better enhance upper limb motor function and attention in stroke patients. This training method may be feasible and suitable for individuals with stroke. TRIAL REGISTRATION: This study was registered in the Chinese Clinical Trial Registry with Portal Number ChiCTR2100050430(27/08/2021).
59The ReHand-BCI trial: a randomized controlled trial of a brain-computer interface for upper extremity stroke neurorehabilitation.PubMed
Jessica Cantillo-Negrete, Martín Emiliano Rodríguez-García, Paul Carrillo-Mora, et al.
Front Neurosci. 2025 Jun 18;19:1579988. doi: 10.3389/fnins.2025.1579988. eCollection 2025.
BACKGROUND: Brain-computer interfaces (BCI) are a promising complementary therapy for stroke rehabilitation due to the close-loop feedback that can be provided with these systems, but more evidence is needed regarding their clinical and neuroplasticity effects. METHODS: A randomized controlled trial was performed using the ReHand-BCI system that provides feedback with a robotic hand orthosis. The experimental group (EG) used the ReHand-BCI, while sham-BCI was given to the control group (CG). Both groups performed 30 therapy sessions, with primary outcomes being the Fugl-Meyer Assessment for the Upper Extremity (FMA-UE) and the Action Research Arm Test (ARAT). Secondary outcomes were hemispheric dominance, measured with electroencephalography and functional magnetic resonance imaging, white matter integrity via diffusion tensor imaging, and corticospinal tract integrity and excitability, measured with transcranial magnetic stimulation. RESULTS: At post-treatment, patients in both groups had significantly different FMA-UE scores (EG: baseline = 24.5[20, 36], post-treatment 28[23, 43], CG: baseline = 26[16, 37.5], post-treatment = 34[17.3, 46.5]), while only the EG had significantly different ARAT scores at post-treatment (EG: baseline = 8.5[5, 26], post-treatment = 20[7, 36], CG: baseline = 3[1.8, 30.5], post-treatment = 15[2.5, 40.8]). In addition, across the intervention, the EG showed trends of more pronounced ipsilesional cortical activity and higher ipsilesional corticospinal tract integrity, although these differences were not statistically different compared to the control group, likely due to the study's sample size. CONCLUSION: To the authors' knowledge, this is the first clinical trial that has assessed such a wide range of physiological effects across a long BCI intervention, implying that a more pronounced ipsilesional hemispheric dominance is associated with upper extremity motor recovery. Therefore, the study brings light into the neuroplasticity effects of a closed-loop BCI-based neurorehabilitation intervention in stroke. CLINICAL TRIAL REGISTRATION: https://clinicaltrials.gov/, identifier NCT04724824.
60Adaptive deep brain stimulation in Parkinson's disease.PubMed
M Beudel, P Brown
Parkinsonism Relat Disord. 2016 Jan;22 Suppl 1(Suppl 1):S123-6. doi: 10.1016/j.parkreldis.2015.09.028. Epub 2015 Sep 15.
Although Deep Brain Stimulation (DBS) is an established treatment for Parkinson's disease (PD), there are still limitations in terms of effectivity, side-effects and battery consumption. One of the reasons for this may be that not only pathological but also physiological neural activity can be suppressed whilst stimulating. For this reason, adaptive DBS (aDBS), where stimulation is applied according to the level of pathological activity, might be advantageous. Initial studies of aDBS demonstrate effectiveness in PD, but there are still many questions to be answered before aDBS can be applied clinically. Here we discuss the feedback signals and stimulation algorithms involved in adaptive stimulation in PD and sketch a potential road-map towards clinical application.
61Adaptive deep brain stimulation in advanced Parkinson disease.PubMed
Simon Little, Alex Pogosyan, Spencer Neal, et al.
Ann Neurol. 2013 Sep;74(3):449-57. doi: 10.1002/ana.23951. Epub 2013 Jul 12.
OBJECTIVE: Brain-computer interfaces (BCIs) could potentially be used to interact with pathological brain signals to intervene and ameliorate their effects in disease states. Here, we provide proof-of-principle of this approach by using a BCI to interpret pathological brain activity in patients with advanced Parkinson disease (PD) and to use this feedback to control when therapeutic deep brain stimulation (DBS) is delivered. Our goal was to demonstrate that by personalizing and optimizing stimulation in real time, we could improve on both the efficacy and efficiency of conventional continuous DBS. METHODS: We tested BCI-controlled adaptive DBS (aDBS) of the subthalamic nucleus in 8 PD patients. Feedback was provided by processing of the local field potentials recorded directly from the stimulation electrodes. The results were compared to no stimulation, conventional continuous stimulation (cDBS), and random intermittent stimulation. Both unblinded and blinded clinical assessments of motor effect were performed using the Unified Parkinson's Disease Rating Scale. RESULTS: Motor scores improved by 66% (unblinded) and 50% (blinded) during aDBS, which were 29% (p = 0.03) and 27% (p = 0.005) better than cDBS, respectively. These improvements were achieved with a 56% reduction in stimulation time compared to cDBS, and a corresponding reduction in energy requirements (p < 0.001). aDBS was also more effective than no stimulation and random intermittent stimulation. INTERPRETATION: BCI-controlled DBS is tractable and can be more efficient and efficacious than conventional continuous neuromodulation for PD.
62Adaptive Deep Brain Stimulation: From Experimental Evidence Toward Practical Implementation.PubMed
Wolf-Julian Neumann, Roee Gilron, Simon Little, et al.
Mov Disord. 2023 Jun;38(6):937-948. doi: 10.1002/mds.29415. Epub 2023 May 6.
Closed-loop adaptive deep brain stimulation (aDBS) can deliver individualized therapy at an unprecedented temporal precision for neurological disorders. This has the potential to lead to a breakthrough in neurotechnology, but the translation to clinical practice remains a significant challenge. Via bidirectional implantable brain-computer-interfaces that have become commercially available, aDBS can now sense and selectively modulate pathophysiological brain circuit activity. Pilot studies investigating different aDBS control strategies showed promising results, but the short experimental study designs have not yet supported individualized analyses of patient-specific factors in biomarker and therapeutic response dynamics. Notwithstanding the clear theoretical advantages of a patient-tailored approach, these new stimulation possibilities open a vast and mostly unexplored parameter space, leading to practical hurdles in the implementation and development of clinical trials. Therefore, a thorough understanding of the neurophysiological and neurotechnological aspects related to aDBS is crucial to develop evidence-based treatment regimens for clinical practice. Therapeutic success of aDBS will depend on the integrated development of strategies for feedback signal identification, artifact mitigation, signal processing, and control policy adjustment, for precise stimulation delivery tailored to individual patients. The present review introduces the reader to the neurophysiological foundation of aDBS for Parkinson's disease (PD) and other network disorders, explains currently available aDBS control policies, and highlights practical pitfalls and difficulties to be addressed in the upcoming years. Finally, it highlights the importance of interdisciplinary clinical neurotechnological research within and across DBS centers, toward an individualized patient-centered approach to invasive brain stimulation. © 2023 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.
63Adaptive deep brain stimulation (aDBS) controlled by local field potential oscillations.PubMed
Alberto Priori, Guglielmo Foffani, Lorenzo Rossi, et al.
Exp Neurol. 2013 Jul;245:77-86. doi: 10.1016/j.expneurol.2012.09.013. Epub 2012 Sep 27.
Despite their proven efficacy in treating neurological disorders, especially Parkinson's disease, deep brain stimulation (DBS) systems could be further optimized to maximize treatment benefits. In particular, because current open-loop DBS strategies based on fixed stimulation settings leave the typical parkinsonian motor fluctuations and rapid symptom variations partly uncontrolled, research has for several years focused on developing novel "closed-loop" or "adaptive" DBS (aDBS) systems. aDBS consists of a simple closed-loop model designed to measure and analyze a control variable reflecting the patient's clinical condition to elaborate new stimulation settings and send them to an "intelligent" implanted stimulator. The major problem in developing an aDBS system is choosing the ideal control variable for feedback. Here we review current evidence on the advantages of neurosignal-controlled aDBS that uses local field potentials (LFPs) as a control variable, and describe the technology already available to create new aDBS systems, and the potential benefits of aDBS for patients with Parkinson's disease.
64Chronic adaptive deep brain stimulation versus conventional stimulation in Parkinson's disease: a blinded randomized feasibility trial.PubMed
Carina R Oehrn, Stephanie Cernera, Lauren H Hammer, et al.
Nat Med. 2024 Nov;30(11):3345-3356. doi: 10.1038/s41591-024-03196-z. Epub 2024 Aug 19.
Deep brain stimulation (DBS) is a widely used therapy for Parkinson's disease (PD) but lacks dynamic responsiveness to changing clinical and neural states. Feedback control might improve therapeutic effectiveness, but the optimal control strategy and additional benefits of 'adaptive' neurostimulation are unclear. Here we present the results of a blinded randomized cross-over pilot trial aimed at determining the neural correlates of specific motor signs in individuals with PD and the feasibility of using these signals to drive adaptive DBS. Four male patients with PD were recruited from a population undergoing DBS implantation for motor fluctuations, with each patient receiving adaptive DBS and continuous DBS. We identified stimulation-entrained gamma oscillations in the subthalamic nucleus or motor cortex as optimal markers of high versus low dopaminergic states and their associated residual motor signs in all four patients. We then demonstrated improved motor symptoms and quality of life with adaptive compared to clinically optimized standard stimulation. The results of this pilot trial highlight the promise of personalized adaptive neurostimulation in PD based on data-driven selection of neural signals. Furthermore, these findings provide the foundation for further larger clinical trials to evaluate the efficacy of personalized adaptive neurostimulation in PD and other neurological disorders. ClinicalTrials.gov registration: NCT03582891 .
65Adaptive deep brain stimulation for Parkinson's disease using motor cortex sensing.PubMed
Nicole C Swann, Coralie de Hemptinne, Margaret C Thompson, et al.
J Neural Eng. 2018 Aug;15(4):046006. doi: 10.1088/1741-2552/aabc9b. Epub 2018 May 9.
OBJECTIVE: Contemporary deep brain stimulation (DBS) for Parkinson's disease is delivered continuously, and adjustments based on patient's changing symptoms must be made manually by a trained clinician. Patients may be subjected to energy intensive settings at times when they are not needed, possibly resulting in stimulation-induced adverse effects, such as dyskinesia. One solution is 'adaptive' DBS, in which stimulation is modified in real time based on neural signals that co-vary with the severity of motor signs or of stimulation-induced adverse effects. Here we show the feasibility of adaptive DBS using a fully implanted neural prosthesis. APPROACH: We demonstrate adaptive deep brain stimulation in two patients with Parkinson's disease using a fully implanted neural prosthesis that is enabled to utilize brain sensing to control stimulation amplitude (Activa PC + S). We used a cortical narrowband gamma (60-90 Hz) oscillation related to dyskinesia to decrease stimulation voltage when gamma oscillatory activity is high (indicating dyskinesia) and increase stimulation voltage when it is low. MAIN RESULTS: We demonstrate the feasibility of 'adaptive deep brain stimulation' in two patients with Parkinson's disease. In short term in-clinic testing, energy savings were substantial (38%-45%), and therapeutic efficacy was maintained. SIGNIFICANCE: This is the first demonstration of adaptive DBS in Parkinson's disease using a fully implanted device and neural sensing. Our approach is distinct from other strategies utilizing basal ganglia signals for feedback control.
66Chronic multisite brain recordings from a totally implantable bidirectional neural interface: experience in 5 patients with Parkinson's disease.PubMed
Nicole C Swann, Coralie de Hemptinne, Svjetlana Miocinovic, et al.
J Neurosurg. 2018 Feb;128(2):605-616. doi: 10.3171/2016.11.JNS161162. Epub 2017 Apr 14.
OBJECTIVE Dysfunction of distributed neural networks underlies many brain disorders. The development of neuromodulation therapies depends on a better understanding of these networks. Invasive human brain recordings have a favorable temporal and spatial resolution for the analysis of network phenomena but have generally been limited to acute intraoperative recording or short-term recording through temporarily externalized leads. Here, the authors describe their initial experience with an investigational, first-generation, totally implantable, bidirectional neural interface that allows both continuous therapeutic stimulation and recording of field potentials at multiple sites in a neural network. METHODS Under a physician-sponsored US Food and Drug Administration investigational device exemption, 5 patients with Parkinson's disease were implanted with the Activa PC+S system (Medtronic Inc.). The device was attached to a quadripolar lead placed in the subdural space over motor cortex, for electrocorticography potential recordings, and to a quadripolar lead in the subthalamic nucleus (STN), for both therapeutic stimulation and recording of local field potentials. Recordings from the brain of each patient were performed at multiple time points over a 1-year period. RESULTS There were no serious surgical complications or interruptions in deep brain stimulation therapy. Signals in both the cortex and the STN were relatively stable over time, despite a gradual increase in electrode impedance. Canonical movement-related changes in specific frequency bands in the motor cortex were identified in most but not all recordings. CONCLUSIONS The acquisition of chronic multisite field potentials in humans is feasible. The device performance characteristics described here may inform the design of the next generation of totally implantable neural interfaces. This research tool provides a platform for translating discoveries in brain network dynamics to improved neurostimulation paradigms. Clinical trial registration no.: NCT01934296 (clinicaltrials.gov).
67Noninvasive closed-loop acoustic brain-computer interface for seizure control.PubMed
Junjie Zou, Houminji Chen, Xiaoyan Chen, et al.
Theranostics. 2024 Sep 9;14(15):5965-5981. doi: 10.7150/thno.99820. eCollection 2024.
The brain-computer interface (BCI) is core tasks in comprehensively understanding the brain, and is one of the most significant challenges in neuroscience. The development of novel non-invasive neuromodulation technique will drive major innovations and breakthroughs in the field of BCI. We develop a new noninvasive closed-loop acoustic brain-computer interface (aBCI) for decoding the seizure onset based on the electroencephalography and triggering ultrasound stimulation of the vagus nerve to terminate seizures. Firstly, we create the aBCI system and decode the onset of seizure via a multi-level threshold model based on the analysis of wireless-collected electroencephalogram (EEG) signals recorded from above the hippocampus. Then, the different acoustic parameters induced acoustic radiation force were used to stimulate the vagus nerve in a rat model of epilepsy-induced by pentylenetetrazole. Finally, the results of epileptic EEG signal triggering ultrasound stimulation of the vagus nerve to control seizures. In addition, the mechanism of aBCI control seizures were investigated by real-time quantitative polymerase chain reaction (RT-qPCR). In a rat model of epilepsy, the aBCI system selectively actives mechanosensitive neurons in the nodose ganglion while suppressing neuronal excitability in the hippocampus and amygdala, and stops seizures rapidly upon ultrasound stimulation of the vagus nerve. Physical transection or chemical blockade of the vagus nerve pathway abolish the antiepileptic effects of aBCI. In addition, aBCI shows significant antiepileptic effects compared to conventional vagus nerve electrical stimulation in an acute experiment. Closed-loop aBCI provides a novel, safe and effective tool for on-demand stimulation to treat abnormal neuronal discharges, opening the door to next generation non-invasive BCI.
68iEEG: Dura-lining electrodes.PubMed
Dora Hermes, Kai J Miller
Handb Clin Neurol. 2020;168:263-277. doi: 10.1016/B978-0-444-63934-9.00019-6.
Intracranial electroencephalography (iEEG) is measured from electrodes placed in or on the brain. These measurements have an excellent signal-to-noise ratio and iEEG signals have often been used to decode brain activity or drive brain-computer interfaces (BCIs). iEEG recordings are typically done for seizure monitoring in epilepsy patients who have these electrodes placed for a clinical purpose: to localize both brain regions that are essential for function and others where seizures start. Brain regions not involved in epilepsy are thought to function normally and provide a unique opportunity to learn about human neurophysiology. Intracranial electrodes measure the aggregate activity of large neuronal populations and recorded signals contain many features. Different features are extracted by analyzing these signals in the time and frequency domain. The time domain may reveal an evoked potential at a particular time after the onset of an event. Decomposition into the frequency domain may show narrowband peaks in the spectrum at specific frequencies or broadband signal changes that span a wide range of frequencies. Broadband power increases are generally observed when a brain region is active while most other features are highly specific to brain regions, inputs, and tasks. Here we describe the spatiotemporal dynamics of several iEEG signals that have often been used to decode brain activity and drive BCIs.
69The Potential of Stereotactic-EEG for Brain-Computer Interfaces: Current Progress and Future Directions.PubMed
Christian Herff, Dean J Krusienski, Pieter Kubben
Front Neurosci. 2020 Feb 27;14:123. doi: 10.3389/fnins.2020.00123. eCollection 2020.
Stereotactic electroencephalogaphy (sEEG) utilizes localized, penetrating depth electrodes to measure electrophysiological brain activity. It is most commonly used in the identification of epileptogenic zones in cases of refractory epilepsy. The implanted electrodes generally provide a sparse sampling of a unique set of brain regions including deeper brain structures such as hippocampus, amygdala and insula that cannot be captured by superficial measurement modalities such as electrocorticography (ECoG). Despite the overlapping clinical application and recent progress in decoding of ECoG for Brain-Computer Interfaces (BCIs), sEEG has thus far received comparatively little attention for BCI decoding. Additionally, the success of the related deep-brain stimulation (DBS) implants bodes well for the potential for chronic sEEG applications. This article provides an overview of sEEG technology, BCI-related research, and prospective future directions of sEEG for long-term BCI applications.
71Brain-computer interfaces for patients with disorders of consciousness.PubMed
R M Gibson, A M Owen, D Cruse
Prog Brain Res. 2016;228:241-91. doi: 10.1016/bs.pbr.2016.04.003. Epub 2016 May 10.
The disorders of consciousness refer to clinical conditions that follow a severe head injury. Patients diagnosed as in a vegetative state lack awareness, while patients diagnosed as in a minimally conscious state retain fluctuating awareness. However, it is a challenge to accurately diagnose these disorders with clinical assessments of behavior. To improve diagnostic accuracy, neuroimaging-based approaches have been developed to detect the presence or absence of awareness in patients who lack overt responsiveness. For the small subset of patients who retain awareness, brain-computer interfaces could serve as tools for communication and environmental control. Here we review the existing literature concerning the sensory and cognitive abilities of patients with disorders of consciousness with respect to existing brain-computer interface designs. We highlight the challenges of device development for this special population and address some of the most promising approaches for future investigations.
72Brain-computer interfaces for consciousness assessment and communication in severely brain-injured patients.PubMed
Jitka Annen, Steven Laureys, Olivia Gosseries
Handb Clin Neurol. 2020;168:137-152. doi: 10.1016/B978-0-444-63934-9.00011-1.
Patients with disorders of consciousness (DOC) suffer from awareness deficits. Comorbidities such as motor disabilities or visual problems hamper clinical assessments, which can lead to misdiagnosis of the level of consciousness and render the patient unable to communicate. Objective measures of consciousness can reduce the risk of misdiagnosis and could enable patients to communicate by voluntarily modulating their brain activity. This chapter gives an overview of the literature regarding brain-computer interface (BCI) research in DOC patients. Different auditory, visual, and motor imagery paradigms are discussed, alongside their corresponding advantages and disadvantages. At this point, the use of BCIs for DOC patients in clinical applications is still preliminary. However, perspectives on the improvements in BCIs for DOC patients seem positive, and implementation during rehabilitation shows promise.
73Brain-Computer Interfaces in Acute and Subacute Disorders of Consciousness.PubMed
Ren Xu, Rossella Spataro, Brendan Z Allison, et al.
J Clin Neurophysiol. 2022 Jan 1;39(1):32-39. doi: 10.1097/WNP.0000000000000810.
Disorders of consciousness include coma, unresponsive wakefulness syndrome (also known as vegetative state), and minimally conscious state. Neurobehavioral scales such as coma recovery scale-revised are the gold standard for disorder of consciousness assessment. Brain-computer interfaces have been emerging as an alternative tool for these patients. The application of brain-computer interfaces in disorders of consciousness can be divided into four fields: assessment, communication, prediction, and rehabilitation. The operational theoretical model of consciousness that brain-computer interfaces explore was reviewed in this article, with a focus on studies with acute and subacute patients. We then proposed a clinically friendly guideline, which could contribute to the implementation of brain-computer interfaces in neurorehabilitation settings. Finally, we discussed limitations and future directions, including major challenges and possible solutions.
74Brain-Computer Interfaces for Awareness Detection, Auxiliary Diagnosis, Prognosis, and Rehabilitation in Patients with Disorders of Consciousness.PubMed
Jiahui Pan, Jun Xiao, Jing Wang, et al.
Semin Neurol. 2022 Jun;42(3):363-374. doi: 10.1055/a-1900-7261. Epub 2022 Jul 14.
In recent years, neuroimaging studies have remarkably demonstrated the presence of cognitive motor dissociation in patients with disorders of consciousness (DoC). These findings accelerated the development of brain-computer interfaces (BCIs) as clinical tools for behaviorally unresponsive patients. This article reviews the recent progress of BCIs in patients with DoC and discusses the open challenges. In view of the practical application of BCIs in patients with DoC, four aspects of the relevant literature are introduced: consciousness detection, auxiliary diagnosis, prognosis, and rehabilitation. For each aspect, the paradigm design, brain signal processing methods, and experimental results of representative BCI systems are analyzed. Furthermore, this article provides guidance for BCI design for patients with DoC and discusses practical challenges for future research.
75Brain-Computer Interfaces for Communication in Patients with Disorders of Consciousness: A Gap Analysis and Scientific Roadmap.PubMed
Nicholas D Schiff, Michael Diringer, Karin Diserens, et al.
Neurocrit Care. 2024 Aug;41(1):129-145. doi: 10.1007/s12028-023-01924-w. Epub 2024 Jan 29.
BACKGROUND: We developed a gap analysis that examines the role of brain-computer interfaces (BCI) in patients with disorders of consciousness (DoC), focusing on their assessment, establishment of communication, and engagement with their environment. METHODS: The Curing Coma Campaign convened a Coma Science work group that included 16 clinicians and neuroscientists with expertise in DoC. The work group met online biweekly and performed a gap analysis of the primary question. RESULTS: We outline a roadmap for assessing BCI readiness in patients with DoC and for advancing the use of BCI devices in patients with DoC. Additionally, we discuss preliminary studies that inform development of BCI solutions for communication and assessment of readiness for use of BCIs in DoC study participants. Special emphasis is placed on the challenges posed by the complex pathophysiologies caused by heterogeneous brain injuries and their impact on neuronal signaling. The differences between one-way and two-way communication are specifically considered. Possible implanted and noninvasive BCI solutions for acute and chronic DoC in adult and pediatric populations are also addressed. CONCLUSIONS: We identify clinical and technical gaps hindering the use of BCI in patients with DoC in each of these contexts and provide a roadmap for research aimed at improving communication for adults and children with DoC, spanning the clinical spectrum from intensive care unit to chronic care.
76Machine-learning-based coadaptive calibration for brain-computer interfaces.PubMed
Carmen Vidaurre, Claudia Sannelli, Klaus-Robert Müller, et al.
Neural Comput. 2011 Mar;23(3):791-816. doi: 10.1162/NECO_a_00089. Epub 2010 Dec 16.
Brain-computer interfaces (BCIs) allow users to control a computer application by brain activity as acquired (e.g., by EEG). In our classic machine learning approach to BCIs, the participants undertake a calibration measurement without feedback to acquire data to train the BCI system. After the training, the user can control a BCI and improve the operation through some type of feedback. However, not all BCI users are able to perform sufficiently well during feedback operation. In fact, a nonnegligible portion of participants (estimated 15%-30%) cannot control the system (a BCI illiteracy problem, generic to all motor-imagery-based BCIs). We hypothesize that one main difficulty for a BCI user is the transition from offline calibration to online feedback. In this work, we investigate adaptive machine learning methods to eliminate offline calibration and analyze the performance of 11 volunteers in a BCI based on the modulation of sensorimotor rhythms. We present an adaptation scheme that individually guides the user. It starts with a subject-independent classifier that evolves to a subject-optimized state-of-the-art classifier within one session while the user interacts continuously. These initial runs use supervised techniques for robust coadaptive learning of user and machine. Subsequent runs use unsupervised adaptation to track the features' drift during the session and provide an unbiased measure of BCI performance. Using this approach, without any offline calibration, six users, including one novice, obtained good performance after 3 to 6 minutes of adaptation. More important, this novel guided learning also allows participants with BCI illiteracy to gain significant control with the BCI in less than 60 minutes. In addition, one volunteer without sensorimotor idle rhythm peak at the beginning of the BCI experiment developed it during the course of the session and used voluntary modulation of its amplitude to control the feedback application.
77Representative-Based Cold Start for Adaptive SSVEP-BCI.PubMed
Nanlin Shi, Xiang Li, Bingchuan Liu, et al.
IEEE Trans Neural Syst Rehabil Eng. 2023;31:1521-1531. doi: 10.1109/TNSRE.2023.3245654. Epub 2023 Mar 7.
OBJECTIVE: The tradeoff between calibration effort and model performance still hinders the user experience for steady-state visual evoked brain-computer interfaces (SSVEP-BCI). To address this issue and improve model generalizability, this work investigated the adaptation from the cross-dataset model to avoid the training process, while maintaining high prediction ability. METHODS: When a new subject enrolls, a group of user-independent (UI) models is recommended as the representative from a multi-source data pool. The representative model is then augmented with online adaptation and transfer learning techniques based on user-dependent (UD) data. The proposed method is validated on both offline (N=55) and online (N=12) experiments. RESULTS: Compared with the UD adaptation, the recommended representative model relieved approximately 160 trials of calibration efforts for a new user. In the online experiment, the time window decreased from 2 s to 0.56±0.2 s, while maintaining high prediction accuracy of 0.89-0.96. Finally, the proposed method achieved the average information transfer rate (ITR) of 243.49 bits/min, which is the highest ITR ever reported in a complete calibration-free setting. The results of the offline result were consistent with the online experiment. CONCLUSION: Representatives can be recommended even in a cross-subject/device/session situation. With the help of represented UI data, the proposed method can achieve sustained high performance without a training process. SIGNIFICANCE: This work provides an adaptive approach to the transferable model for SSVEP-BCIs, enabling a more generalized, plug-and-play and high-performance BCI free of calibrations.
78Towards psychologically adaptive brain-computer interfaces.PubMed
A Myrden, T Chau
J Neural Eng. 2016 Dec;13(6):066022. doi: 10.1088/1741-2560/13/6/066022. Epub 2016 Nov 14.
OBJECTIVE: Brain-computer interface (BCI) performance is sensitive to short-term changes in psychological states such as fatigue, frustration, and attention. This paper explores the design of a BCI that can adapt to these short-term changes. APPROACH: Eleven able-bodied individuals participated in a study during which they used a mental task-based EEG-BCI to play a simple maze navigation game while self-reporting their perceived levels of fatigue, frustration, and attention. In an offline analysis, a regression algorithm was trained to predict changes in these states, yielding Pearson correlation coefficients in excess of 0.45 between the self-reported and predicted states. Two means of fusing the resultant mental state predictions with mental task classification were investigated. First, single-trial mental state predictions were used to predict correct classification by the BCI during each trial. Second, an adaptive BCI was designed that retrained a new classifier for each testing sample using only those training samples for which predicted mental state was similar to that predicted for the current testing sample. MAIN RESULTS: Mental state-based prediction of BCI reliability exceeded chance levels. The adaptive BCI exhibited significant, but practically modest, increases in classification accuracy for five of 11 participants and no significant difference for the remaining six despite a smaller average training set size. SIGNIFICANCE: Collectively, these findings indicate that adaptation to psychological state may allow the design of more accurate BCIs.
79Machine learning in biosignals processing for mental health: A narrative review.PubMed
Elena Sajno, Sabrina Bartolotta, Cosimo Tuena, et al.
Front Psychol. 2023 Jan 13;13:1066317. doi: 10.3389/fpsyg.2022.1066317. eCollection 2022.
Machine Learning (ML) offers unique and powerful tools for mental health practitioners to improve evidence-based psychological interventions and diagnoses. Indeed, by detecting and analyzing different biosignals, it is possible to differentiate between typical and atypical functioning and to achieve a high level of personalization across all phases of mental health care. This narrative review is aimed at presenting a comprehensive overview of how ML algorithms can be used to infer the psychological states from biosignals. After that, key examples of how they can be used in mental health clinical activity and research are illustrated. A description of the biosignals typically used to infer cognitive and emotional correlates (e.g., EEG and ECG), will be provided, alongside their application in Diagnostic Precision Medicine, Affective Computing, and brain-computer Interfaces. The contents will then focus on challenges and research questions related to ML applied to mental health and biosignals analysis, pointing out the advantages and possible drawbacks connected to the widespread application of AI in the medical/mental health fields. The integration of mental health research and ML data science will facilitate the transition to personalized and effective medicine, and, to do so, it is important that researchers from psychological/ medical disciplines/health care professionals and data scientists all share a common background and vision of the current research.
80Adaptive binary multi-objective harmony search algorithm for channel selection and cross-subject generalization in motor imagery-based BCI.PubMed
Bin Shi, Zan Yue, Shuai Yin, et al.
J Neural Eng. 2022 Jul 26;19(4). doi: 10.1088/1741-2552/ac7d73.
Multi-channel electroencephalogram data containing redundant information and noise may result in low classification accuracy and high computational complexity, which limits the practicality of motor imagery (MI)-based brain-computer interface (BCI) systems. Therefore, channel selection can improve BCI performance and contribute to user convenience. Additionally, cross-subject generalization is a key topic in the channel selection of MI-based BCI.In this study, an adaptive binary multi-objective harmony search (ABMOHS) algorithm is proposed to select the optimal set of channels. Furthermore, a new adaptive cross-subject generalization model (ACGM) is proposed. Three public MI datasets were used to validate the effectiveness of the proposed method.The Wilcoxon signed-rank test was performed on the test accuracies, and the results indicated that the ABMOHS method significantly outperformed all channels (< 0.001), the C3-Cz-C4 channels (< 0.001), and 20 channels (< 0.001) in the sensorimotor cortex. The ABMOHS algorithm based on Fisher's linear discriminant analysis (FLDA) and support vector machine (SVM) classifiers greatly reduces the number of selected channels, especially for larger channel sizes (Dataset 2), and obtains a comparative classification performance. Although there was no significant difference in test classification performance between ABMOHS and non-dominated sorting genetic algorithm II (NSGA-II) when FLDA and SVM were used, ABMOHS required less computational time than NSGA-II. Furthermore, the number of channels obtained by ABMOHS algorithm were significantly smaller than those obtained by common spatial pattern-Rank and correlation-based channel selection algorithm. Additionally, the generalization of ACGM to untrained subjects shows that the mean test classification accuracy of ACGM created by a small sample of trained subjects is significantly better than that of Special-16 and Special-32.The proposed method can reduce the calibration time in the training phase and improve the practicability of MI-BCI.
81A Usability Study of Low-cost Wireless Brain-Computer Interface for Cursor Control Using Online Linear Model.PubMed
Reza Abiri, Soheil Borhani, Justin Kilmarx, et al.
IEEE Trans Hum Mach Syst. 2020 Aug;50(4):287-297. doi: 10.1109/thms.2020.2983848. Epub 2020 May 14.
Computer cursor control using electroencephalogram (EEG) signals is a common and well-studied brain-computer interface (BCI). The emphasis of the literature has been primarily on evaluation of the objective measures of assistive BCIs such as accuracy of the neural decoder whereas the subjective measures such as user's satisfaction play an essential role for the overall success of a BCI. As far as we know, the BCI literature lacks a comprehensive evaluation of the usability of the mind-controlled computer cursor in terms of decoder efficiency (accuracy), user experience, and relevant confounding variables concerning the platform for the public use. To fill this gap, we conducted a two-dimensional EEG-based cursor control experiment among 28 healthy participants. The computer cursor velocity was controlled by the imagery of hand movement using a paradigm presented in the literature named imagined body kinematics (IBK) with a low-cost wireless EEG headset. We evaluated the usability of the platform for different objective and subjective measures while we investigated the extent to which the training phase may influence the ultimate BCI outcome. We conducted pre- and post- BCI experiment interview questionnaires to evaluate the usability. Analyzing the questionnaires and the testing phase outcome shows a positive correlation between the individuals' ability of visualization and their level of mental controllability of the cursor. Despite individual differences, analyzing training data shows the significance of electrooculogram (EOG) on the predictability of the linear model. The results of this work may provide useful insights towards designing a personalized user-centered assistive BCI.
82Increasing Robustness of Brain-Computer Interfaces Through Automatic Detection and Removal of Corrupted Input Signals.PubMed
Jordan L Vasko, Laura Aume, Sanjay Tamrakar, et al.
Front Neurosci. 2022 Apr 28;16:858377. doi: 10.3389/fnins.2022.858377. eCollection 2022.
For brain-computer interfaces (BCIs) to be viable for long-term daily usage, they must be able to quickly identify and adapt to signal disruptions. Furthermore, the detection and mitigation steps need to occur automatically and without the need for user intervention while also being computationally tractable for the low-power hardware that will be used in a deployed BCI system. Here, we focus on disruptions that are likely to occur during chronic use that cause some recording channels to fail but leave the remaining channels unaffected. In these cases, the algorithm that translates recorded neural activity into actions, the neural decoder, should seamlessly identify and adjust to the altered neural signals with minimal inconvenience to the user. First, we introduce an adapted statistical process control (SPC) method that automatically identifies disrupted channels so that both decoding algorithms can be adjusted, and technicians can be alerted. Next, after identifying corrupted channels, we demonstrate the automated and rapid removal of channels from a neural network decoder using a masking approach that does not change the decoding architecture, making it amenable for transfer learning. Finally, using transfer and unsupervised learning techniques, we update the model weights to adjust for the corrupted channels without requiring the user to collect additional calibration data. We demonstrate with both real and simulated neural data that our approach can maintain high-performance while simultaneously minimizing computation time and data storage requirements. This framework is invisible to the user but can dramatically increase BCI robustness and usability.
83Electroencephalogram-based adaptive closed-loop brain-computer interface in neurorehabilitation: a review.PubMed
Wenjie Jin, XinXin Zhu, Lifeng Qian, et al.
Front Comput Neurosci. 2024 Sep 20;18:1431815. doi: 10.3389/fncom.2024.1431815. eCollection 2024.
Brain-computer interfaces (BCIs) represent a groundbreaking approach to enabling direct communication for individuals with severe motor impairments, circumventing traditional neural and muscular pathways. Among the diverse array of BCI technologies, electroencephalogram (EEG)-based systems are particularly favored due to their non-invasive nature, user-friendly operation, and cost-effectiveness. Recent advancements have facilitated the development of adaptive bidirectional closed-loop BCIs, which dynamically adjust to users' brain activity, thereby enhancing responsiveness and efficacy in neurorehabilitation. These systems support real-time modulation and continuous feedback, fostering personalized therapeutic interventions that align with users' neural and behavioral responses. By incorporating machine learning algorithms, these BCIs optimize user interaction and promote recovery outcomes through mechanisms of activity-dependent neuroplasticity. This paper reviews the current landscape of EEG-based adaptive bidirectional closed-loop BCIs, examining their applications in the recovery of motor and sensory functions, as well as the challenges encountered in practical implementation. The findings underscore the potential of these technologies to significantly enhance patients' quality of life and social interaction, while also identifying critical areas for future research aimed at improving system adaptability and performance. As advancements in artificial intelligence continue, the evolution of sophisticated BCI systems holds promise for transforming neurorehabilitation and expanding applications across various domains.
84P300 Brain-Computer Interface-Based Drone Control in Virtual and Augmented Reality.PubMed
Soram Kim, Seungyun Lee, Hyunsuk Kang, et al.
Sensors (Basel). 2021 Aug 27;21(17):5765. doi: 10.3390/s21175765.
Since the emergence of head-mounted displays (HMDs), researchers have attempted to introduce virtual and augmented reality (VR, AR) in brain-computer interface (BCI) studies. However, there is a lack of studies that incorporate both AR and VR to compare the performance in the two environments. Therefore, it is necessary to develop a BCI application that can be used in both VR and AR to allow BCI performance to be compared in the two environments. In this study, we developed an opensource-based drone control application using P300-based BCI, which can be used in both VR and AR. Twenty healthy subjects participated in the experiment with this application. They were asked to control the drone in two environments and filled out questionnaires before and after the experiment. We found no significant ( > 0.05) difference in online performance (classification accuracy and amplitude/latency of P300 component) and user experience (satisfaction about time length, program, environment, interest, difficulty, immersion, and feeling of self-control) between VR and AR. This indicates that the P300 BCI paradigm is relatively reliable and may work well in various situations.
85How different immersive environments affect intracortical brain computer interfaces.PubMed
Ariana F Tortolani, Nicolas G Kunigk, Anton R Sobinov, et al.
J Neural Eng. 2025 Feb 10;22(1):016032. doi: 10.1088/1741-2552/adb078.
. As brain-computer interface (BCI) research advances, many new applications are being developed. Tasks can be performed in different virtual environments, and whether a BCI user can switch environments seamlessly will influence the ultimate utility of a clinical device.. Here we investigate the importance of the immersiveness of the virtual environment used to train BCI decoders on the resulting decoder and its generalizability between environments. Two participants who had intracortical electrodes implanted in their precentral gyrus used a BCI to control a virtual arm, both viewed immersively through virtual reality goggles and at a distance on a flat television monitor.. Each participant performed better with a decoder trained and tested in the environment they had used the most prior to the study, one for each environment type. The neural tuning to the desired movement was minimally influenced by the immersiveness of the environment. Finally, in further testing with one of the participants, we found that decoders trained in one environment generalized well to the other environment, but the order in which the environments were experienced within a session mattered.. Overall, experience with an environment was more influential on performance than the immersiveness of the environment, but BCI performance generalized well after accounting for experience.Clinical Trial: NCT01894802.
86Human-Avatar Symbiosis for the Treatment of Auditory Verbal Hallucinations in Schizophrenia through Virtual/Augmented Reality and Brain-Computer Interfaces.PubMed
Antonio Fernández-Caballero, Elena Navarro, Patricia Fernández-Sotos, et al.
Front Neuroinform. 2017 Nov 21;11:64. doi: 10.3389/fninf.2017.00064. eCollection 2017.
This perspective paper faces the future of alternative treatments that take advantage of a social and cognitive approach with regards to pharmacological therapy of auditory verbal hallucinations (AVH) in patients with schizophrenia. AVH are the perception of voices in the absence of auditory stimulation and represents a severe mental health symptom. Virtual/augmented reality (VR/AR) and brain computer interfaces (BCI) are technologies that are growing more and more in different medical and psychological applications. Our position is that their combined use in computer-based therapies offers still unforeseen possibilities for the treatment of physical and mental disabilities. This is why, the paper expects that researchers and clinicians undergo a pathway toward human-avatar symbiosis for AVH by taking full advantage of new technologies. This outlook supposes to address challenging issues in the understanding of non-pharmacological treatment of schizophrenia-related disorders and the exploitation of VR/AR and BCI to achieve a real human-avatar symbiosis.
87Digitally Assisted Mindfulness in Training Self-Regulation Skills for Sustainable Mental Health: A Systematic Review.PubMed
Eleni Mitsea, Athanasios Drigas, Charalabos Skianis
Behav Sci (Basel). 2023 Dec 10;13(12):1008. doi: 10.3390/bs13121008.
The onset of the COVID-19 pandemic has led to an increased demand for mental health interventions, with a special focus on digitally assisted ones. Self-regulation describes a set of meta-skills that enable one to take control over his/her mental health and it is recognized as a vital indicator of well-being. Mindfulness training is a promising training strategy for promoting self-regulation, behavioral change, and mental well-being. A growing body of research outlines that smart technologies are ready to revolutionize the way mental health training programs take place. Artificial intelligence (AI); extended reality (XR) including virtual reality (VR), augmented reality (AR), and mixed reality (MR); as well as the advancements in brain computer interfaces (BCIs) are ready to transform these mental health training programs. Mindfulness-based interventions assisted by smart technologies for mental, emotional, and behavioral regulation seem to be a crucial yet under-investigated issue. The current systematic review paper aims to explore whether and how smart technologies can assist mindfulness training for the development of self-regulation skills among people at risk of mental health issues as well as populations with various clinical characteristics. The PRISMA 2020 methodology was utilized to respond to the objectives and research questions using a total of sixty-six experimental studies that met the inclusion criteria. The results showed that digitally assisted mindfulness interventions supported by smart technologies, including AI-based applications, chatbots, virtual coaches, immersive technologies, and brain-sensing headbands, can effectively assist trainees in developing a wide range of cognitive, emotional, and behavioral self-regulation skills, leading to a greater satisfaction of their psychological needs, and thus mental wellness. These results may provide positive feedback for developing smarter and more inclusive training environments, with a special focus on people with special training needs or disabilities.
88Enhancing visual brain-computer interface through V1-targeted RTMS by modulating visual attention.PubMed
Xinyi Zhang, Shengpei Wang, Ying Gao, et al.
Imaging Neurosci (Camb). 2025 Nov 17;3. doi: 10.1162/IMAG.a.1013. eCollection 2025.
Brain-computer interfaces (BCIs) enable users to control devices directly through brain activity. Despite recent advancements in machine-learning algorithms, the signal-to-noise ratio (SNR) of the brain's responses still limits decoding performance, highlighting the necessity for targeted neuromodulation techniques to overcome this limitation. To evaluate whether 5 Hz repetitive transcranial magnetic stimulation (rTMS) targeting the primary visual cortex (V1) can enhance SSVEP-based BCI performance by improving neural signal SNR and modulating visual network dynamics. Twenty-four healthy subjects underwent both real and sham rTMS in a randomized order. The rTMS was precisely implemented through magnetic resonance imaging (MRI)-guided navigation to stimulate V1 in participants. Electroencephalograms (EEGs) were recorded during SSVEP tasks and resting-state before, immediately after, and 20 min after rTMS. SSVEP tasks were conducted across four frequency bands: low frequency (LF: 8-12 Hz), middle frequency (MF: 18-22 Hz), high frequency (HF: 28-32 Hz), and super high frequency (SHF: 38-42 Hz). The discriminability of BCI commands in the MF (+7.53%) and HF (+11.4%) bands significantly improved ( ), driven by enhanced prominence of both fundamental and harmonic components ( ). Quantitative analysis indicated that the improved SNR was due to the suppression of the background activity ( ). This effect was linked to rTMS-induced enhancements in visual attention, evidenced by increased occurrence and contribution of microstate B during the SSVEP task ( ). This study highlights the potential of 5 Hz rTMS as an effective neuromodulatory tool for optimizing BCI performance, particularly through facilitating visual attention.
895 Hz rTMS improves motor-imagery based BCI classification performance.PubMed
Tianyu Jia, Linhong Mo, Chong Li, et al.
Annu Int Conf IEEE Eng Med Biol Soc. 2021 Nov;2021:6116-6120. doi: 10.1109/EMBC46164.2021.9630102.
Brain-computer interface (BCI) based rehabilitation has been proven a promising method facilitating motor recovery. Recognizing motor intention is crucial for realizing BCI rehabilitation training. Event-related desynchronization (ERD) is a kind of electroencephalogram (EEG) inherent characteristics associated with motor intention. However, due to brain deficits poststroke, some patients are not able to generate ERD, which discourages them to be involved in BCI rehabilitation training. To boost ERD during motor imagery (MI), this paper investigates the effects of high-frequency repetitive transcranial magnetic stimulation (rTMS) on BCI classification performance. Eleven subjects participated in this study. The experiment consisted of two conditions: rTMS + MI versus sham rTMS + MI, which were arranged on different days. MI tests with 64-channel EEG recording were arranged immediately before and after rTMS and sham rTMS. Time-frequency analysis were utilized to measure ERD changes. Common spatial pattern was used to extract features and linear discriminant analysis was used to calculate offline classification accuracies. Paired-sample t-test and Wilcoxon signed rank tests with post-hoc analysis were used to compare performance before and after stimulation. Statistically stronger ERD (-13.93±12.99%) was found after real rTMS compared with ERD (-5.71±21.25%) before real rTMS (p<0.05). Classification accuracy after real rTMS (70.71±10.32%) tended to be higher than that before real rTMS (66.50±8.48%) (p<0.1). However, no statistical differences were found after sham stimulation. This research provides an effective method in improving BCI performance by utilizing neural modulation.Clinical Relevance- This study offers a promising treatment for patients who cannot be recruited in BCI rehabilitation training due to poor BCI classification performance.
90iTBS on RDLPFC improves performance of motor imagery: a brain-computer interface study combining EEG and fNIRS.PubMed
Jialin Chen, Quan Liu, Gengbin Chen, et al.
J Neuroeng Rehabil. 2025 Jul 31;22(1):172. doi: 10.1186/s12984-025-01688-8.
BACKGROUND: Some individuals using brain-computer interfaces (BCIs) exhibit ineffective control during motor imagery-based BCI (MI-BCI) training. MI-BCI performance correlates with the activation in the frontoparietal attention network, premotor-parietal network, and supplementary motor area (SMA). This study aimed to enhance motor imagery ability and MI-BCI performance by modulating the excitability of the right dorsolateral prefrontal cortex (RDLPFC) through intermittent theta-burst stimulation (iTBS), inducing neuroplastic changes. METHODS: Fifty-two healthy right-handed participants were randomly assigned to either the iTBS or sham group. They undertook two MI-BCI training sessions, with electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) used to assess acute neuroplasticity changes. The intervention was administered between sessions. Corticospinal excitability and motor imagery vividness were assessed using single-pulse transcranial magnetic stimulation (spTMS) and the Kinesthetic and Visual Imagery Questionnaire-20 (KVIQ-20) before and following the trial. RESULTS: The iTBS group significantly improved motor state percentage (MSP). Significant µ event-related desynchronization (µ-ERD) was observed at the F4 electrode in the iTBS group. Functional connectivity (FC) analyses revealed decreased connectivity among several electrodes during the post-intervention period. The hemodynamic response function (HRF) indicated significant activation in the right PMC and SMA, with reduced FC among motor areas. No significant differences in MEP, CSP, and KVIQ-20 scores were found between groups. CONCLUSION: iTBS targeting the RDLPFC may improve MI-BCI training performance and address the "BCI inefficiency" problem. RDLPFC stimulation induced changes in FC of brain regions associated with motor imagery and increased the activation of motor areas, suggesting that the RDLPFC could be a promising target for enhancing motor imagery and optimizing BCI systems.
91Cortical Activation Patterns Determine Effectiveness of rTMS-Induced Motor Imagery Decoding Enhancement in Stroke Patients.PubMed
Tianyu Jia, Linhong Mo, Ciaran McGeady, et al.
IEEE Trans Biomed Eng. 2025 Mar;72(3):1200-1208. doi: 10.1109/TBME.2024.3492977. Epub 2025 Feb 20.
Combination therapy with motor imagery (MI)-based brain-computer interface (BCI) and repetitive transcranial magnetic stimulation (rTMS) is a promising therapy for poststroke neurorehabilitation. However, with patients' individual differences, the clinical effects vary greatly. This study aims to explore the hypothesis that stroke patients show individualized cortical response to rTMS treatments, which determine the effectiveness of rTMS-induced MI decoding enhancement. We applied four kinds of rTMS treatments respectively to four groups of subacute stroke patients, twenty-six patients in total, and observed their EEG dynamics, MI decoding performance, and Fugl-Meyer assessment changes following 2-week neuromodulation. Four treatments consisted of ipsilesional 10 Hz rTMS, contralesional 1 Hz rTMS, ipsilesional 1 Hz rTMS, and sham stimulation. Results showed stroke patients with different neural reorganization patterns responded individually to rTMS therapy. Patients with cortical lesions mostly showed contralesional recruitment and patients without cortical lesions mostly presented ipsilesional focusing. Significant activation increases in the ipsilesional hemisphere (pre: -15.7% ∓ 8.2%, post: -17.3% ∓ 8.1%, p = 0.037) and MI decoding accuracy enhancement (pre: 76.3 ± 13.8%, post: 86.6 ± 8.2%, p = 0.037) were concurrently found in no-cortical-lesion patients with ipsilesional activation treatment. In the group of patients without cortical lesions, recovery rate in those receiving ipsilesional activation therapy (23.5 ± 10.4%) was higher than those receiving ipsilesional suppression therapy (9.9 ± 9.3%) (p = 0.041). This study reveals that tailoring neuromodulation therapy by recognizing cortical activation patterns is promising for improving effectiveness of the combination therapy with BCI and rTMS.
92Clinical Applications and Future Translation of Somatosensory Neuroprostheses.PubMed
Emily Graczyk, Brianna Hutchison, Giacomo Valle, et al.
J Neurosci. 2024 Oct 2;44(40):e1237242024. doi: 10.1523/JNEUROSCI.1237-24.2024.
Somatosensory neuroprostheses restore, replace, or enhance tactile and proprioceptive feedback for people with sensory impairments due to neurological disorders or injury. Somatosensory neuroprostheses typically couple sensor inputs from a wearable device, prosthesis, robotic device, or virtual reality system with electrical stimulation applied to the somatosensory nervous system via noninvasive or implanted interfaces. While prior research has mainly focused on technology development and proof-of-concept studies, recent acceleration of clinical studies in this area demonstrates the translational potential of somatosensory neuroprosthetic systems. In this review, we provide an overview of neurostimulation approaches currently undergoing human testing and summarize recent clinical findings on the perceptual, functional, and psychological impact of somatosensory neuroprostheses. We also cover current work toward the development of advanced stimulation paradigms to produce more natural and informative sensory feedback. Finally, we provide our perspective on the remaining challenges that need to be addressed prior to translation of somatosensory neuroprostheses.
93Restoring tactile and proprioceptive sensation through a brain interface.PubMed
Gregg A Tabot, Sung Shin Kim, Jeremy E Winberry, et al.
Neurobiol Dis. 2015 Nov;83:191-8. doi: 10.1016/j.nbd.2014.08.029. Epub 2014 Sep 6.
Somatosensation plays a critical role in the dexterous manipulation of objects, in emotional communication, and in the embodiment of our limbs. For upper-limb neuroprostheses to be adopted by prospective users, prosthetic limbs will thus need to provide sensory information about the position of the limb in space and about objects grasped in the hand. One approach to restoring touch and proprioception consists of electrically stimulating neurons in somatosensory cortex in the hopes of eliciting meaningful sensations to support the dexterous use of the hands, promote their embodiment, and perhaps even restore the affective dimension of touch. In this review, we discuss the importance of touch and proprioception in everyday life, then describe approaches to providing artificial somatosensory feedback through intracortical microstimulation (ICMS). We explore the importance of biomimicry--the elicitation of naturalistic patterns of neuronal activation--and that of adaptation--the brain's ability to adapt to novel sensory input, and argue that both biomimicry and adaptation will play a critical role in the artificial restoration of somatosensation. We also propose that the documented re-organization that occurs after injury does not pose a significant obstacle to brain interfaces. While still at an early stage of development, sensory restoration is a critical step in transitioning upper-limb neuroprostheses from the laboratory to the clinic.
94A brain-computer interface that evokes tactile sensations improves robotic arm control.PubMed
Sharlene N Flesher, John E Downey, Jeffrey M Weiss, et al.
Science. 2021 May 21;372(6544):831-836. doi: 10.1126/science.abd0380.
Prosthetic arms controlled by a brain-computer interface can enable people with tetraplegia to perform functional movements. However, vision provides limited feedback because information about grasping objects is best relayed through tactile feedback. We supplemented vision with tactile percepts evoked using a bidirectional brain-computer interface that records neural activity from the motor cortex and generates tactile sensations through intracortical microstimulation of the somatosensory cortex. This enabled a person with tetraplegia to substantially improve performance with a robotic limb; trial times on a clinical upper-limb assessment were reduced by half, from a median time of 20.9 to 10.2 seconds. Faster times were primarily due to less time spent attempting to grasp objects, revealing that mimicking known biological control principles results in task performance that is closer to able-bodied human abilities.
95A review of critical challenges in MI-BCI: From conventional to deep learning methods.PubMed
Zahra Khademi, Farideh Ebrahimi, Hussain Montazery Kordy
J Neurosci Methods. 2023 Jan 1;383:109736. doi: 10.1016/j.jneumeth.2022.109736. Epub 2022 Oct 29.
Brain-computer interfaces (BCIs) have achieved significant success in controlling external devices through the Electroencephalogram (EEG) signal processing. BCI-based Motor Imagery (MI) system bridges brain and external devices as communication tools to control, for example, wheelchair for people with disabilities, robotic control, and exoskeleton control. This success largely depends on the machine learning (ML) approaches like deep learning (DL) models. DL algorithms provide effective and powerful models to analyze compact and complex EEG data optimally for MI-BCI applications. DL models with CNN network have revolutionized computer vision through end-to-end learning from raw data. Meanwhile, RNN networks have been able to decode EEG signals by processing sequences of time series data. However, many challenges in the MI-BCI field have affected the performance of DL models. A major challenge is the individual differences in the EEG signal of different subjects. Therefore, the model must be retrained from the scratch for each new subject, which leads to computational costs. Analyzing the EEG signals is challenging due to its low signal to noise ratio and non-stationary nature. Additionally, limited size of existence datasets can lead to overfitting which can be prevented by using transfer learning (TF) approaches. The main contributions of this study are discovering major challenges in the MI-BCI field by reviewing the state of art machine learning models and then suggesting solutions to address these challenges by focusing on feature selection, feature extraction and classification methods.
96Intracortical recording stability in human brain-computer interface users.PubMed
John E Downey, Nathaniel Schwed, Steven M Chase, et al.
J Neural Eng. 2018 Aug;15(4):046016. doi: 10.1088/1741-2552/aab7a0. Epub 2018 Mar 19.
OBJECTIVE: Intracortical brain-computer interfaces (BCIs) are being developed to assist people with motor disabilities in communicating and interacting with the world around them. This technology relies on recordings from the primary motor cortex, which may vary from day to day. APPROACH: Here we quantify, in two long-term BCI subjects, the length of time that action potentials from the same neuron, or group of neurons, can be recorded from the motor cortex. MAIN RESULTS: These action potentials are identified by their extracellular waveforms and may change within a single day, although some of these identified units can be identified consistently for weeks and even months. Features of the extracellular waveforms allowed us to predict whether a specific unit was more or less likely to remain stable over a prolonged period. SIGNIFICANCE: A greater understanding of unit stability and instability can aid the development of motor BCIs, where the goal is to maintain a high level of performance despite changes in the recorded population. BCIs should be able to be operated without technician intervention for hours, and hopefully days, to provide the most benefit to the end-users of this technology.
97Subdural Soft Electrocorticography (ECoG) Array Implantation and Long-Term Cortical Recording in Minipigs.PubMed
Florian Fallegger, Alix Trouillet, Stéphanie P Lacour
J Vis Exp. 2023 Mar 31(193). doi: 10.3791/64997.
Neurological impairments and diseases can be diagnosed or treated using electrocorticography (ECoG) arrays. In drug-resistant epilepsy, these help delineate the epileptic region to resect. In long-term applications such as brain-computer interfaces, these epicortical electrodes are used to record the movement intention of the brain, to control the robotic limbs of paralyzed patients. However, current stiff electrode grids do not answer the need for high-resolution brain recordings and long-term biointegration. Recently, conformable electrode arrays have been proposed to achieve long-term implant stability with high performance. However, preclinical studies for these new implant technologies are needed to validate their long-term functionality and safety profile for their translation to human patients. In this context, porcine models are routinely employed in developing medical devices due to their large organ sizes and easy animal handling. However, only a few brain applications are described in the literature, mostly due to surgery limitations and integration of the implant system on a living animal. Here, we report the method for long-term implantation (6 months) and evaluation of soft ECoG arrays in the minipig model. The study first presents the implant system, consisting of a soft microfabricated electrode array integrated with a magnetic resonance imaging (MRI)-compatible polymeric transdermal port that houses instrumentation connectors for electrophysiology recordings. Then, the study describes the surgical procedure, from subdural implantation to animal recovery. We focus on the auditory cortex as an example target area where evoked potentials are induced by acoustic stimulation. We finally describe a data acquisition sequence that includes MRI of the whole brain, implant electrochemical characterization, intraoperative and freely moving electrophysiology, and immunohistochemistry staining of the extracted brains. This model can be used to investigate the safety and function of novel design of cortical prostheses; mandatory preclinical study to envision translation to human patients.
98Plug-and-play control of a brain-computer interface through neural map stabilization.PubMed
Daniel B Silversmith, Reza Abiri, Nicholas F Hardy, et al.
Nat Biotechnol. 2021 Mar;39(3):326-335. doi: 10.1038/s41587-020-0662-5. Epub 2020 Sep 7.
Brain-computer interfaces (BCIs) enable control of assistive devices in individuals with severe motor impairments. A limitation of BCIs that has hindered real-world adoption is poor long-term reliability and lengthy daily recalibration times. To develop methods that allow stable performance without recalibration, we used a 128-channel chronic electrocorticography (ECoG) implant in a paralyzed individual, which allowed stable monitoring of signals. We show that long-term closed-loop decoder adaptation, in which decoder weights are carried across sessions over multiple days, results in consolidation of a neural map and 'plug-and-play' control. In contrast, daily reinitialization led to degradation of performance with variable relearning. Consolidation also allowed the addition of control features over days, that is, long-term stacking of dimensions. Our results offer an approach for reliable, stable BCI control by leveraging the stability of ECoG interfaces and neural plasticity.
99A click-based electrocorticographic brain-computer interface enables long-term high-performance switch scan spelling.PubMed
Daniel N Candrea, Samyak Shah, Shiyu Luo, et al.
Commun Med (Lond). 2024 Oct 22;4(1):207. doi: 10.1038/s43856-024-00635-3.
BACKGROUND: Brain-computer interfaces (BCIs) can restore communication for movement- and/or speech-impaired individuals by enabling neural control of computer typing applications. Single command click detectors provide a basic yet highly functional capability. METHODS: We sought to test the performance and long-term stability of click decoding using a chronically implanted high density electrocorticographic (ECoG) BCI with coverage of the sensorimotor cortex in a human clinical trial participant (ClinicalTrials.gov, NCT03567213) with amyotrophic lateral sclerosis. We trained the participant's click detector using a small amount of training data (<44 min across 4 days) collected up to 21 days prior to BCI use, and then tested it over a period of 90 days without any retraining or updating. RESULTS: Using a click detector to navigate a switch scanning speller interface, the study participant can maintain a median spelling rate of 10.2 characters per min. Though a transient reduction in signal power modulation can interrupt usage of a fixed model, a new click detector can achieve comparable performance despite being trained with even less data (<15 min, within 1 day). CONCLUSIONS: These results demonstrate that a click detector can be trained with a small ECoG dataset while retaining robust performance for extended periods, providing functional text-based communication to BCI users.
100Advances in endovascular brain computer interface: Systematic review and future implications.PubMed
Julien Ognard, Gerard El Hajj, Onam Verma, et al.
J Neurosci Methods. 2025 Aug;420:110471. doi: 10.1016/j.jneumeth.2025.110471. Epub 2025 May 10.
BACKGROUND: Brain-computer interfaces (BCIs) translate neural activity into real-world commands. While traditional invasive BCIs necessitate craniotomy, endovascular BCIs offer a minimally invasive alternative using the venous system for electrode placement. NEW METHOD: This systematic review evaluates the technical feasibility, safety, and clinical outcomes of endovascular BCIs, discussing their future implications. A systematic review was conducted per PRISMA guidelines. The search spanned PubMed, Web of Science, and Scopus databases using keywords related to neural interfaces and endovascular approaches. Studies were included if they reported on endovascular BCIs in preclinical or clinical settings. Dual independent screening and extraction focused on electrode material, recording capabilities, safety parameters, and clinical efficacy. RESULTS: From 1385 initial publications, 26 met the inclusion criteria. Seventeen studies investigated the Stentrode device. Among the 24 preclinical studies, 16 used ovine or rodent models, and 9 addressed engineering or simulation aspects. Two clinical studies reported six ALS patients successfully using an endovascular BCI for digital communication. Preclinical data established the endovascular ovine model, demonstrating stable neural recordings and vascular changes with long-term implantation. Key challenges include thrombosis risk, long-term electrode stability, and anatomical variability. COMPARISON WITH EXISTING METHODS: Endovascular BCI reduced invasiveness, improved safety profiles, with comparable neural recording fidelity to invasive methods, and promising preliminary clinical outcomes in severely paralyzed patients. CONCLUSIONS: Early results are promising, but clinical data remain scarce. Further research is needed to optimize signal processing, enhance electrode biocompatibility, and refine endovascular procedures for broader clinical applications.
101Improving Generalization of CNN-Based Motor-Imagery EEG Decoders via Dynamic Convolutions.PubMed
Konstantinos Barmpas, Yannis Panagakis, Stylianos Bakas, et al.
IEEE Trans Neural Syst Rehabil Eng. 2023;31:1997-2005. doi: 10.1109/TNSRE.2023.3265304. Epub 2023 Apr 14.
Deep Convolutional Neural Networks (CNNs) have recently demonstrated impressive results in electroencephalogram (EEG) decoding for several Brain-Computer Interface (BCI) paradigms, including Motor-Imagery (MI). However, neurophysiological processes underpinning EEG signals vary across subjects causing covariate shifts in data distributions and hence hindering the generalization of deep models across subjects. In this paper, we aim to address the challenge of inter-subject variability in MI. To this end, we employ causal reasoning to characterize all possible distribution shifts in the MI task and propose a dynamic convolution framework to account for shifts caused by the inter-subject variability. Using publicly available MI datasets, we demonstrate improved generalization performance (up to 5%) across subjects in various MI tasks for four well-established deep architectures.
102Advocating for neurodata privacy and neurotechnology regulation.PubMed
Rafael Yuste
Nat Protoc. 2023 Oct;18(10):2869-2875. doi: 10.1038/s41596-023-00873-0. Epub 2023 Sep 11.
The ability to record and alter brain activity by using implantable and nonimplantable neural devices, while poised to have significant scientific and clinical benefits, also raises complex ethical concerns. In this Perspective, we raise awareness of the ability of artificial intelligence algorithms and data-aggregation tools to decode and analyze data containing highly sensitive information, jeopardizing personal neuroprivacy. Voids in existing regulatory frameworks, in fact, allow unrestricted decoding and commerce of neurodata. We advocate for the implementation of proposed ethical and human rights guidelines, alongside technical options such as data encryption, differential privacy and federated learning to ensure the protection of neurodata privacy. We further encourage regulatory bodies to consider taking a position of responsibility by categorizing all brain-derived data as sensitive health data and apply existing medical regulations to all data gathered via pre-registered neural devices. Lastly, we propose that a technocratic oath may instill a deontology for neurotechnology practitioners akin to what the Hippocratic oath represents in medicine. A conscientious societal position that thoroughly rejects the misuse of neurodata would provide the moral compass for the future development of the neurotechnology field.
103Ethical aspects of brain computer interfaces: a scoping review.PubMed
Sasha Burwell, Matthew Sample, Eric Racine
BMC Med Ethics. 2017 Nov 9;18(1):60. doi: 10.1186/s12910-017-0220-y.
BACKGROUND: Brain-Computer Interface (BCI) is a set of technologies that are of increasing interest to researchers. BCI has been proposed as assistive technology for individuals who are non-communicative or paralyzed, such as those with amyotrophic lateral sclerosis or spinal cord injury. The technology has also been suggested for enhancement and entertainment uses, and there are companies currently marketing BCI devices for those purposes (e.g., gaming) as well as health-related purposes (e.g., communication). The unprecedented direct connection created by BCI between human brains and computer hardware raises various ethical, social, and legal challenges that merit further examination and discussion. METHODS: To identify and characterize the key issues associated with BCI use, we performed a scoping review of biomedical ethics literature, analyzing the ethics concerns cited across multiple disciplines, including philosophy and medicine. RESULTS: Based on this investigation, we report that BCI research and its potential translation to therapeutic intervention generate significant ethical, legal, and social concerns, notably with regards to personhood, stigma, autonomy, privacy, research ethics, safety, responsibility, and justice. Our review of the literature determined, furthermore, that while these issues have been enumerated extensively, few concrete recommendations have been expressed. CONCLUSIONS: We conclude that future research should focus on remedying a lack of practical solutions to the ethical challenges of BCI, alongside the collection of empirical data on the perspectives of the public, BCI users, and BCI researchers.
104The ethical significance of user-control in AI-driven speech-BCIs: a narrative review.PubMed
O C van Stuijvenberg, D P S Samlal, M J Vansteensel, et al.
Front Hum Neurosci. 2024 Jun 27;18:1420334. doi: 10.3389/fnhum.2024.1420334. eCollection 2024.
AI-driven brain-computed interfaces aimed at restoring speech for individuals living with locked-in-syndrome are paired with ethical implications for user's autonomy, privacy and responsibility. Embedding options for sufficient levels of user-control in speech-BCI design has been proposed to mitigate these ethical challenges. However, how user-control in speech-BCIs is conceptualized and how it relates to these ethical challenges is underdetermined. In this narrative literature review, we aim to clarify and explicate the notion of user-control in speech-BCIs, to better understand in what way user-control could operationalize user's autonomy, privacy and responsibility and explore how such suggestions for increasing user-control can be translated to recommendations for the design or use of speech-BCIs. First, we identified types of user control, including executory control that can protect voluntariness of speech, and guidance control that can contribute to semantic accuracy. Second, we identified potential causes for a loss of user-control, including contributions of predictive language models, a lack of ability for neural control, or signal interference and external control. Such a loss of user control may have implications for semantic accuracy and mental privacy. Third we explored ways to design for user-control. While embedding initiation signals for users may increase executory control, they may conflict with other aims such as speed and continuity of speech. Design mechanisms for guidance control remain largely conceptual, similar trade-offs in design may be expected. We argue that preceding these trade-offs, the overarching aim of speech-BCIs needs to be defined, requiring input from current and potential users. Additionally, conceptual clarification of user-control and other (ethical) concepts in this debate has practical relevance for BCI researchers. For instance, different concepts of inner speech may have distinct ethical implications. Increased clarity of such concepts can improve anticipation of ethical implications of speech-BCIs and may help to steer design decisions.
105Decentralising the Self - Ethical Considerations in Utilizing Decentralised Web Technology for Direct Brain Interfaces.PubMed
David M Lyreskog, Hazem Zohny, Sebastian Porsdam Mann, et al.
Sci Eng Ethics. 2024 Jul 16;30(4):28. doi: 10.1007/s11948-024-00492-2.
The rapidly advancing field of brain-computer (BCI) and brain-to-brain interfaces (BBI) is stimulating interest across various sectors including medicine, entertainment, research, and military. The developers of large-scale brain-computer networks, sometimes dubbed 'Mindplexes' or 'Cloudminds', aim to enhance cognitive functions by distributing them across expansive networks. A key technical challenge is the efficient transmission and storage of information. One proposed solution is employing blockchain technology over Web 3.0 to create decentralised cognitive entities. This paper explores the potential of a decentralised web for coordinating large brain-computer constellations, and its associated benefits, focusing in particular on the conceptual and ethical challenges this innovation may pose pertaining to (1) Identity, (2) Sovereignty (encompassing Autonomy, Authenticity, and Ownership), (3) Responsibility and Accountability, and (4) Privacy, Safety, and Security. We suggest that while a decentralised web can address some concerns and mitigate certain risks, underlying ethical issues persist. Fundamental questions about entity definition within these networks, the distinctions between individuals and collectives, and responsibility distribution within and between networks, demand further exploration.
106Media Representation of the Ethical Issues Pertaining to Brain-Computer Interface (BCI) Technology.PubMed
Savannah Beck, Yuliya Liberman, Veljko Dubljević
Brain Sci. 2024 Dec 14;14(12):1255. doi: 10.3390/brainsci14121255.
BACKGROUND/OBJECTIVES: Brain-computer interfaces (BCIs) are a rapidly developing technology that captures and transmits brain signals to external sources, allowing the user control of devices such as prosthetics. BCI technology offers the potential to restore physical capabilities in the body and change how we interact and communicate with computers and each other. While BCI technology has existed for decades, recent developments have caused the technology to generate a host of ethical issues and discussions in both academic and public circles. Given that media representation has the potential to shape public perception and policy, it is necessary to evaluate the space that these issues take in public discourse. METHODS: We conducted a rapid review of media articles in English discussing ethical issues of BCI technology from 2013 to 2024 as indexed by LexisNexis. Our searches yielded 675 articles, with a final sample containing 182 articles. We assessed the themes of the articles and coded them based on the ethical issues discussed, ethical frameworks, recommendations, tone, and application of technology. RESULTS: Our results showed a marked rise in interest in media articles over time, signaling an increased focus on this topic. The majority of articles adopted a balanced or neutral tone when discussing BCIs and focused on ethical issues regarding privacy, autonomy, and regulation. CONCLUSIONS: Current discussion of ethical issues reflects growing news coverage of companies such as Neuralink, and reveals a mounting distrust of BCI technology. The growing recognition of ethical considerations in BCI highlights the importance of ethical discourse in shaping the future of the field.
107Understanding the Ethical Issues of Brain-Computer Interfaces (BCIs): A Blessing or the Beginning of a Dystopian Future?PubMed
Efstratios Livanis, Polychronis Voultsos, Konstantinos Vadikolias, et al.
Cureus. 2024 Apr 14;16(4):e58243. doi: 10.7759/cureus.58243. eCollection 2024 Apr.
In recent years, scientific discoveries in the field of neuroscience combined with developments in the field of artificial intelligence have led to the development of a range of neurotechnologies. Advances in neuroimaging systems, neurostimulators, and brain-computer interfaces (BCIs) are leading to new ways of enhancing, controlling, and "reading" the brain. In addition, although BCIs were developed and used primarily in the medical field, they are now increasingly applied in other fields (entertainment, marketing, education, defense industry). We conducted a literature review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to provide background information about ethical issues related to the use of BCIs. Among the ethical issues that emerged from the thematic data analysis of the reviewed studies included questions revolving around human dignity, personhood and autonomy, user safety, stigma and discrimination, privacy and security, responsibility, research ethics, and social justice (including access to this technology). This paper attempts to address the various aspects of these concerns. A variety of distinct ethical issues were identified, which, for the most part, were in line with the findings of prior research. However, we identified two nuances, which are related to the empirical research on ethical issues related to BCIs and the impact of BCIs on international relationships. The paper also highlights the need for the cooperation of all stakeholders to ensure the ethical development and use of this technology and concludes with several recommendations. The principles of bioethics provide an initial guiding framework, which, however, should be revised in the current artificial intelligence landscape so as to be responsive to challenges posed by the development and use of BCIs.
108Embodiment and Estrangement: Results from a First-in-Human "Intelligent BCI" Trial.PubMed
F Gilbert, M Cook, T O'Brien, et al.
Sci Eng Ethics. 2019 Feb;25(1):83-96. doi: 10.1007/s11948-017-0001-5. Epub 2017 Nov 11.
While new generations of implantable brain computer interface (BCI) devices are being developed, evidence in the literature about their impact on the patient experience is lagging. In this article, we address this knowledge gap by analysing data from the first-in-human clinical trial to study patients with implanted BCI advisory devices. We explored perceptions of self-change across six patients who volunteered to be implanted with artificially intelligent BCI devices. We used qualitative methodological tools grounded in phenomenology to conduct in-depth, semi-structured interviews. Results show that, on the one hand, BCIs can positively increase a sense of the self and control; on the other hand, they can induce radical distress, feelings of loss of control, and a rupture of patient identity. We conclude by offering suggestions for the proactive creation of preparedness protocols specific to intelligent-predictive and advisory-BCI technologies essential to prevent potential iatrogenic harms.
109The Mystery of Mental Integrity: Clarifying Its Relevance to Neurotechnologies.PubMed
Hazem Zohny, David M Lyreskog, Ilina Singh, et al.
Neuroethics. 2023;16(3):20. doi: 10.1007/s12152-023-09525-2. Epub 2023 Aug 21.
The concept of mental integrity is currently a significant topic in discussions concerning the regulation of neurotechnologies. Technologies such as deep brain stimulation and brain-computer interfaces are believed to pose a unique threat to mental integrity, and some authors have advocated for a legal right to protect it. Despite this, there remains uncertainty about what mental integrity entails and why it is important. Various interpretations of the concept have been proposed, but the literature on the subject is inconclusive. Here we consider a number of possible interpretations and argue that the most plausible one concerns neurotechnologies that bypass one's reasoning capacities, and do so specifically in ways that reliably lead to alienation from one's mental states. This narrows the scope of what constitutes a threat to mental integrity and offers a more precise role for the concept to play in the ethical evaluation of neurotechnologies.
110From Healthcare to Warfare and Reverse: How Should We Regulate Dual-Use Neurotechnology?PubMed
Marcello Ienca, Fabrice Jotterand, Bernice S Elger
Neuron. 2018 Jan 17;97(2):269-274. doi: 10.1016/j.neuron.2017.12.017.
Recent advances in military-funded neurotechnology and novel opportunities for misusing neurodevices show that the problem of dual use is inherent to neuroscience. This paper discusses how the neuroscience community should respond to these dilemmas and delineates a neuroscience-specific biosecurity framework. This neurosecurity framework involves calibrated regulation, (neuro)ethical guidelines, and awareness-raising activities within the scientific community.
111The Ethical and Responsible Development and Application of Advanced Brain Machine Interfaces.PubMed
Andrew David Maynard, Marissa Scragg
J Med Internet Res. 2019 Oct 31;21(10):e16321. doi: 10.2196/16321.
Advanced brain machine interfaces provide potentially transformative approaches to treating neurological conditions and enhancing the performance of users. Yet, as technological capabilities continue to progress in leaps and bounds, there is a possibility that these capabilities outstrip our collective understanding of how to ensure brain machine interfaces are developed and used ethically and responsibly. In this case, there is an overt danger of rapid technological developments leading to unanticipated harm through a lack of foresight including threats to privacy, autonomy, self-identity, and other areas of personal and social value which, while hard to quantify, represent substantial risks. There is also a very real likelihood of such risks undermining value creation around the technologies and the associated enterprises, as key stakeholders push back against perceived and actual threats to what they, in turn, hold to be of value. In order to successfully traverse the resulting risk landscape, researchers and developers will need to become increasingly adept at integrating a sophisticated understanding of ethical and socially responsible innovation into their enterprises. Here, we illustrate how a "risk innovation" approach may provide novel insights into mapping out this landscape and revealing potentially blindsiding risks. We show how this approach can be used to illuminate challenges and opportunities to the successful, ethical, and responsible development of advanced brain machine interfaces. In addition, we emphasize how success will ultimately depend on the willingness of innovators and others to take ethical and responsible innovation seriously and to draw on the interdisciplinary and transdisciplinary expertise that is necessary to translate good intentions into positive outcomes.
112Using brain-computer interfaces: a scoping review of studies employing social research methods.PubMed
Johannes Kögel, Jennifer R Schmid, Ralf J Jox, et al.
BMC Med Ethics. 2019 Mar 7;20(1):18. doi: 10.1186/s12910-019-0354-1.
BACKGROUND: The rapid expansion of research on Brain-Computer Interfaces (BCIs) is not only due to the promising solutions offered for persons with physical impairments. There is also a heightened need for understanding BCIs due to the challenges regarding ethics presented by new technology, especially in its impact on the relationship between man and machine. Here we endeavor to present a scoping review of current studies in the field to gain insight into the complexity of BCI use. By examining studies related to BCIs that employ social research methods, we seek to demonstrate the multitude of approaches and concerns from various angles in considering the social and human impact of BCI technology. METHODS: For this scoping review of research on BCIs' social and ethical implications, we systematically analyzed six databases, encompassing the fields of medicine, psychology, and the social sciences, in order to identify empirical studies on BCIs. The search yielded 73 publications that employ quantitative, qualitative, or mixed methods. RESULTS: Of the 73 publications, 71 studies address the user perspective. Some studies extend to consideration of other BCI stakeholders such as medical technology experts, caregivers, or health care professionals. The majority of the studies employ quantitative methods. Recurring themes across the studies examined were general user opinion towards BCI, central technical or social issues reported, requests/demands made by users of the technology, the potential/future of BCIs, and ethical aspects of BCIs. CONCLUSIONS: Our findings indicate that while technical aspects of BCIs such as usability or feasibility are being studied extensively, comparatively little in-depth research has been done on the self-image and self-experience of the BCI user. In general there is also a lack of focus or examination of the caregiver's perspective.
113[Ethics considerations on brain-computer interface technology].PubMed
Zhe Zhang, Xu Zhao, Yixin Ma, et al.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2023 Apr 25;40(2):358-364. doi: 10.7507/1001-5515.202208058.
The development and potential application of brain-computer interface (BCI) technology is closely related to the human brain, so that the ethical regulation of BCI has become an important issue attracting the consideration of society. Existing literatures have discussed the ethical norms of BCI technology from the perspectives of non-BCI developers and scientific ethics, while few discussions have been launched from the perspective of BCI developers. Therefore, there is a great need to study and discuss the ethical norms of BCI technology from the perspective of BCI developers. In this paper, we present the user-centered and non-harmful BCI technology ethics, and then discuss and look forward on them. This paper argues that human beings can cope with the ethical issues arising from BCI technology, and as BCI technology develops, its ethical norms will be improved continuously. It is expected that this paper can provide thoughts and references for the formulation of ethical norms related to BCI technology.