抑郁症是全球最常见的精神健康疾病之一，其中约30%为对多种抗抑郁药物缺乏反应的难治性抑郁症患者。基于缰核的脑深部电刺激（Deep Brain Stimulation, DBS）是一种神经调控疗法，在治疗难治性抑郁症方面具有很大潜力。然而，缰核的细小体积、术中脑脊液流失或脑移位等不确定性，以及对缰核电生理特性的认知不足，都给DBS手术中电极的精准放置带来了挑战。本研究基于术中微电极记录（Microelectrode Recording, MER）和电生理知识，旨在研究DBS术中缰核定位方法，从而优化电极放置精度，提升治疗效果。我们采集了接受缰核DBS治疗的难治性抑郁症患者的术中MER信号，并通过动作电位波形分拣和神经元放电模式识别，实现了对丘脑与缰核的分类。进一步地，我们计算并比较了患者丘脑与缰核的神经电生理特征和MER信号特征，探讨了抑郁症的潜在病理机制。此外，我们还评估了缰核DBS术中MER的安全性、有效性、可行性及其研究价值，并基于术中MER和患者丘脑与缰核的电生理差异，建立了一种手动定位缰核DBS靶点的方法。我们成功在9个缰核DBS植入轨迹中获取了MER记录，且未发生术中出血或伤口感染。在这9个轨迹中，MER电极在缰核中的平均跨度为1.77 mm。在神经电生理方面，患者丘脑中的爆发式神经元比例（84.96%）显著高于缰核（29.26%），而缰核中的强直性和混合式神经元比例（分别为38.14%和14.74%）则显著高于丘脑（2.14%和5.20%）；患者丘脑爆发式神经元的放电间隔低于10 ms的百分比（24.4%）也显著高于缰核（15.0%）。在MER信号方面，患者丘脑的放电模式通常为爆发式，放电频率为20±9 Hz，而缰核则表现出强直性、爆发式或不规则放电模式，且存在高频强直性神经元，其放电频率在40-60 Hz；患者缰核的MER信号的信噪比和放电幅值均显著低于丘脑。本研究提出了一种简单有效的神经元放电模式分类方法，并建立了一种准确可靠的丘脑与缰核分类模型，丰富了我们对难治性抑郁症患者缰核电生理的认识，并对抑郁症的病理机制提出了合理假设。我们初步验证了MER在缰核DBS手术中的适用性，并为缰核DBS治疗难治性抑郁症的术中靶点定位提供了一种可行的方法，帮助医生根据术中MER信号准确判断缰核位置，以精确植入DBS电极。
Depression is one of the most common mental health disorders worldwide, with approximately 30% of patients suffering from treatment-resistant depression who do not respond to multiple antidepressant medications. Deep Brain Stimulation (DBS) based on habenula is a neuromodulation therapy that has great potential in the treatment of treatment-resistant depression. However, the small size of the habenula, the uncertainty of intraoperative cerebrospinal fluid loss or brain shift, and the lack of understanding of the electrophysiological properties of the habenula pose challenges for the precise placement of electrodes in DBS surgery. Based on Microelectrode Recording (MER) and electrophysiological knowledge, this study aims to establish an intraoperative habenula localization method, so as to optimize the electrode placement accuracy and improve the therapeutic effect.We collected the intraoperative MER signals of patients with treatment-resistant depression treated with habenula DBS, and realized the classification of thalamus and habenula through spike sorting and neuronal firing pattern recognition. Further, we calculated and compared the neuroelectrophysiological and MER signal characteristics of thalamus and habenula to explore the underlying pathological mechanism of depression. In addition, we evaluated the safety, efficacy, feasibility and research value of MER during habenula DBS surgery, and established a method to manually target habenula based on intraoperative MER and the electrophysiological differences between thalamus and habenula.We successfully obtained MERs in 9 habenula DBS implantation trajectories without intraoperative hemorrhage or wound infection. In these nine trajectories, the mean length of MER trajectories in habenula was 1.77 mm. In terms of neuroelectrophysiology, the proportion of burst neurons in thalamus (84.96%) was significantly higher than that in habenula (29.26%), while the proportion of tonic and mixed neurons in habenula (38.14% and 14.74%, respectively) was significantly higher than that in thalamus (2.14% and 5.20%, respectively). The percentage of inter-spike intervals less than 10 ms of burst neurons in the thalamus (24.4%) was significantly higher than that in the habenula (15.0%). In terms of MER signals, the thalamus usually has an bursting pattern with a firing frequency of 20±9 Hz, while the habenula shows a tonic, bursting or irregular firing pattern, and there are high-frequency tonic neurons with a firing frequency of 40-60 Hz. The signal-to-noise ratio and spike amplitude of MER signal in habenula were significantly lower than those in thalamus.In this study, we proposed a simple and effective classification method of neuronal firing patterns, established an accurate and reliable classification model of thalamus and habenula, revealed the neuroelectrophysiological characteristics and MER signal characteristics of human habenula, and proposed reasonable hypotheses for the pathological mechanism of depression. We preliminarily verified the applicability of MER in habenula DBS surgery, and provided a feasible method for intraoperative target localization, which can help neurosurgeons confirm the location of habenula according to the intraoperative MER signal, so as to accurately implant DBS electrodes.
