Alarie Michaela E, Provenza Nicole R, Avendano-Ortega Michelle, McKay Sarah A, Waite Ayan S, Mathura Raissa K, Herron Jeffrey A, Sheth Sameer A, Borton David A, Goodman Wayne K
Brown University School of Engineering, Providence, RI, United States.
Department of Neurosurgery, Baylor College of Medicine, Houston, TX, United States.
Front Hum Neurosci. 2022 Oct 19;16:1016379. doi: 10.3389/fnhum.2022.1016379. eCollection 2022.
Bidirectional deep brain stimulation (DBS) platforms have enabled a surge in hours of recordings in naturalistic environments, allowing further insight into neurological and psychiatric disease states. However, high amplitude, high frequency stimulation generates artifacts that contaminate neural signals and hinder our ability to interpret the data. This is especially true in psychiatric disorders, for which high amplitude stimulation is commonly applied to deep brain structures where the native neural activity is miniscule in comparison. Here, we characterized artifact sources in recordings from a bidirectional DBS platform, the Medtronic Summit RC + S, with the goal of optimizing recording configurations to improve signal to noise ratio (SNR). Data were collected from three subjects in a clinical trial of DBS for obsessive-compulsive disorder. Stimulation was provided bilaterally to the ventral capsule/ventral striatum (VC/VS) using two independent implantable neurostimulators. We first manipulated DBS amplitude within safe limits (2-5.3 mA) to characterize the impact of stimulation artifacts on neural recordings. We found that high amplitude stimulation produces slew overflow, defined as exceeding the rate of change that the analog to digital converter can accurately measure. Overflow led to expanded spectral distortion of the stimulation artifact, with a six fold increase in the bandwidth of the 150.6 Hz stimulation artifact from 147-153 to 140-180 Hz. By increasing sense blank values during high amplitude stimulation, we reduced overflow by as much as 30% and improved artifact distortion, reducing the bandwidth from 140-180 Hz artifact to 147-153 Hz. We also identified artifacts that shifted in frequency through modulation of telemetry parameters. We found that telemetry ratio changes led to predictable shifts in the center-frequencies of the associated artifacts, allowing us to proactively shift the artifacts outside of our frequency range of interest. Overall, the artifact characterization methods and results described here enable increased data interpretability and unconstrained biomarker exploration using data collected from bidirectional DBS devices.
双向深部脑刺激（DBS）平台使得在自然环境中的记录时长大幅增加，有助于进一步深入了解神经和精神疾病状态。然而，高幅度、高频刺激会产生伪迹，这些伪迹会污染神经信号并阻碍我们对数据的解读能力。在精神疾病中尤其如此，因为高幅度刺激通常应用于深部脑结构，而相比之下，这些部位的原生神经活动非常微弱。在此，我们对双向DBS平台美敦力Summit RC + S记录中的伪迹来源进行了特征分析，目的是优化记录配置以提高信噪比（SNR）。数据来自一项针对强迫症的DBS临床试验中的三名受试者。使用两个独立的植入式神经刺激器对腹侧囊/腹侧纹状体（VC/VS）进行双侧刺激。我们首先在安全范围内（2 - 5.3 mA）操纵DBS幅度，以表征刺激伪迹对神经记录的影响。我们发现高幅度刺激会产生 slew 溢出，即超过模拟数字转换器能够准确测量的变化率。溢出导致刺激伪迹的频谱失真扩大，150.6 Hz刺激伪迹的带宽从147 - 153 Hz增加到140 - 180 Hz，增加了六倍。通过在高幅度刺激期间增加感测空白值，我们将溢出减少了多达30%，并改善了伪迹失真，将伪迹带宽从140 - 180 Hz减少到147 - 153 Hz。我们还识别出通过遥测参数调制而在频率上发生偏移的伪迹。我们发现遥测比率变化会导致相关伪迹的中心频率发生可预测的偏移，这使我们能够主动将伪迹移到我们感兴趣的频率范围之外。总体而言，此处描述的伪迹特征分析方法和结果能够提高数据的可解释性，并利用从双向DBS设备收集的数据进行无限制的生物标志物探索。
