基于脑功能网络特征的抑郁症神经生物亚型分类及其疗效预测

Depression have become the greatest burden of disease around world, however the brain mechanism of depression is still under-determined, there is no objective diagnostic criteria, and with low treatment responsiveness. Functional neuroimaging demonstrate that abnormal function of dorso-lateral and dorso-medial prefrontal cortex plays a important role in the development and maintenance of depression. Initial research suggest that only a proportion of patients are sensitive to transcranial magnetic stimulation on either dorso-lateral or dorso-medial prefrontal cortex. These two regions are node within central executive network and default mode network, respectively, the disrupted interactions among these two networks and other brain systems may serve as distinct core brain mechanism of depression's heterogeneity. Therefore it is suggested that distinct patterns of brain network disruption are the neural basis of depression subtypes, and target key node for transcranial magnetic stimulation for each subtype may increase treatment efficacy. Here, we utilize transcranial magnetic stimulation on these two prefrontal target respectively, then to demonstrate the pattern of brain network changes associated with treatment efficacy and the complex modulation effects of inter- and intra- large-scale functional network by graph-theory analysis. Finally rely on the functional brain network signatures and machine learning techniques, we are going to build a diagnostic model for depression subtypes, and treatment efficacy prediction model.

抑郁症已成为世界头号疾病负担，然而其脑机制仍未完全厘清，诊断缺乏客观标准，治愈率低。功能磁共振成像研究表明，背内侧、背外侧额叶的功能异常分别在抑郁的发病、维持中起到重要作用；经颅磁刺激背内侧额叶或背外侧额叶治疗抑郁，只有一部分人病情缓解。这两个区域分别是默认网络和中央控制网络的节点，而默认网络、中央控制网络与大脑其他系统的异常相互作用类型可能是抑郁症异质性的核心脑机制。因此抑郁亚型可能以不同类型的脑功能网络异常模式为神经基础，而针对性地刺激不同亚型的关键节点可能有助于提高抗抑郁效果。本研究拟通过分别对两个位点行经颅磁刺激，通过图论分析方法来揭示脑功能网络的变化模式，明晰大脑对刺激位点及所属固有网络的调控方式；并最终基于发现的脑功能网络特征，采用机器学习建立抑郁症亚型的诊断及疗效预测模型。

抑郁症是严重影响公共健康的精神疾病，具有发病率高、致残率高、治愈率低的特点。我们前期的研究提示，脑网络核心节点的功能及结构连接异常可能是抑郁症的核心脑机制，此外，抑郁症可能存在多种脑亚型，提示抑郁症的发病机制多种多样，治疗方案也应区别针对。因此，抑郁症的纵向研究在其诊断及治疗均具有重要意义。抑郁患者脑固有网络之间的异常交互作用，可能是引发多种抑郁症状的脑机制，基于脑机制的抑郁亚型的分类诊断方面的研究十分匮乏。因此，本课题采用脑网络组学的研究思路以及磁共振影像学方法，结合抑郁症的治疗，初步探究抑郁患者的脑网络损害亚型，并探究可能的治疗靶点。.通过三年的研究，本项目基本完成了任务，取得了部分成果，目前发表已标注的SCI论文1篇，参与国内会议交流2次。我们观察到岛叶、额下回、缘上回与脑网络核心节点的异常联接是抑郁患者异常脑联接的主要表现，而短期治疗主要引入额-顶网络相关脑区介入，调控异常功能联接恢复。