中央环路模型(mesocircuit model) 假说为阐明意识障碍的产生机制和脑刺激治疗提供了研究思路，但支撑假说的关键依据如神经网络的构成和表型特征不详及脑刺激的理想部位不明，影响到意识障碍的精准诊断和治疗。我们前期研究发现极点对称模态分解方法（ESMD）可对血氧水平依赖（BOLD）信号进行频率特性分析，以及显微光学切片断层扫描（MOST）可清晰显示鼠脑神经网络；结合文献提出并验证假说：①神经网络的表型特征具有生物标志作用，依据fMRI和MOST技术可论证意识障碍神经网络的构成及其诊断价值；②结合ESMD对BOLD信号频率特性分析，可建立意识障碍的预后预测模型；③结合不同部位脑刺激前后的生物效应和神经网络表型的变化，可帮助找到脑刺激的理想部位。本项目拟对创伤性意识障碍患者和创伤性昏迷大鼠通过上述方法研究，获取神经网络的表型特征和脑刺激理想部位，为意识障碍的鉴别诊断及其治疗提供新策略。

The mesocircuit model provided train of thought and method for the explanation of the underlying mechanisms of disorders of consciousness (DOC) and brain stimulation. However, the challenges were faced for the differentiated diagnosis in the classification and the prediction of the accurate prognosis of DOC as well as the ideal sites for brain stimulation, due to lack of enough evidence of anatomical structure and characteristic patterns of neural network of DOC supporting the mesocircuit model as key elements. In previous study, we demonstrated that the frequency specificity of the BOLD signal could be achieved with Extreme-Point Symmetric Mode Decomposition (ESMD) Method as expected and the neural networks of cerebral cortex in rat model be documented by means of Micro-Optical Sectioning Tomography System (MOST) techniques. Based on our preliminary research and literature review, we propose and test the hypotheses as followings: ①the characteristic patterns of neural network could play an important role as a biomarker for differential diagnosis of DOC. The neural networks of DOC in patients of traumatic DOC and its corresponding coma-like rat model in brain injury will be documented by means of functional magnetic resonance imaging and MOST techniques.②the predicted model of accurate prognosis of DOC will be established by analyzing the frequency of the BOLD signal with ESMD Method as expected. ③ The ideal sites for brain stimulation will be identified with the aid of the observed changes of biological effects and the corresponding patterns of neural networks of DOC following brain stimulation at different sites. In this project, we plan to investigate the patterns of neural networks in the patients of traumatic DOC and its corresponding coma-like rat model via previously described methods and to achieve the characteristic patterns of neural networks and the best places to brain stimulation. The results will offer the novel strategies for the differential diagnosis of the classification and prediction of accurate prognosis of DOC as well as its therapeutic regimen of brain recovery.