Sanchez Luis A, Pandiaraju Surya, Williams Autumn O, Daraie Amir H, Onyike Chiadi U, Sarma Sridevi V
Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA.
Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA.
Brain Commun. 2025 Jun 30;7(4):fcaf259. doi: 10.1093/braincomms/fcaf259. eCollection 2025.
Frontotemporal dementia is a complex neurodegenerative illness characterized by a progressive deterioration in temperament, judgement, behaviour, and communication. Misdiagnosis and late diagnosis occur frequently due to the complexity of the phenotypes, overlaps of features with other neurodegenerative syndromes and psychiatric disorders, and ill-defined preclinical phases of the illness. Diagnosis relies on structural or functional brain imaging to show characteristic atrophy, hypoperfusion or hypometabolism profiles. The sensitivity of neuroimaging is lower in the earliest phases of the illness, and there are few alternatives. Scalp electroencephalography (EEG) is a widely available, low-cost technology, but its utility in the differential diagnosis of dementia will require EEG indices of high sensitivity and discriminatory value. We have used scalp EEG to develop subject-specific Dynamic Network Models, from which we summarize the reciprocal relationships between the nodes (defined by the EEG channel). This index, the 'Sink-Index', characterizes how activity in each node (or channel) responds to activity in other nodes in the network. In this context, 'sources' are nodes that exert significant influence on the activity of different regions but are not themselves influenced, whereas 'sinks' represent influenced regions that do not affect activity in others. We hypothesized that brain regions associated with Frontotemporal dementia and Alzheimer's disease syndromes behave as sinks and have higher sink indices than healthy brain regions. This hypothesis was tested in a cohort of 88 subjects: 23 with frontotemporal dementia, 36 with Alzheimer's disease, and 29 healthy controls. The Sink-Index of nodes in the frontal-temporal and central-parietal-occipital brain regions differed between Frontotemporal dementia (1.3389 ± 0.0895 versus 0.8444 ± 0.0651), Alzheimer's disease (0.6015 ± 0.0188 versus 0.7766 ± 0.0158), and healthy controls (0.8978 ± 0.0453 versus 0.9116 ± 0.0457). These findings suggest the Sink-Index is an EEG marker with utility for the differential diagnosis of dementia syndromes.
额颞叶痴呆是一种复杂的神经退行性疾病，其特征为气质、判断力、行为和沟通能力逐渐衰退。由于该疾病表型复杂、特征与其他神经退行性综合征及精神疾病存在重叠，且临床前期阶段定义不明确，因此误诊和延迟诊断屡见不鲜。诊断依赖于结构或功能脑成像，以显示特征性萎缩、灌注不足或代谢减低情况。在疾病的最早阶段，神经成像的敏感性较低，且几乎没有其他替代方法。头皮脑电图（EEG）是一种广泛应用、成本低廉的技术，但其在痴呆鉴别诊断中的效用需要具有高敏感性和鉴别价值的EEG指标。我们利用头皮EEG开发了个体特异性动态网络模型，从中总结节点（由EEG通道定义）之间的相互关系。这个指标，即“汇指数”，表征每个节点（或通道）的活动如何响应网络中其他节点的活动。在此背景下，“源”是对不同区域活动施加显著影响但自身不受影响的节点，而“汇”代表受影响但不影响其他区域活动的区域。我们假设与额颞叶痴呆和阿尔茨海默病综合征相关的脑区表现为汇，且其汇指数高于健康脑区。在一个由88名受试者组成的队列中对这一假设进行了检验：23例额颞叶痴呆患者、36例阿尔茨海默病患者和29名健康对照者。额颞叶和中央顶枕叶脑区节点的汇指数在额颞叶痴呆（1.3389±0.0895对0.8444±0.0651）、阿尔茨海默病（0.6015±0.0188对0.7766±0.0158）和健康对照者（0.8978±0.0453对0.9116±0.0457）之间存在差异。这些发现表明汇指数是一种对痴呆综合征鉴别诊断有用的EEG标志物。
