面向意识障碍患者的脑机接口意识检测技术及其临床应用

意识障碍患者的精准意识检测是困扰国内外临床医生的世界性难题。针对意识障碍患者严重缺乏行为反应和临床误诊率高，本项目开展基于脑机接口的意识检测技术及其临床应用的研究，研究内容包括：（1）设计融合P300和情感脑电成分的刺激范式，提出基于长短期记忆网络的多特征联合情绪识别算法，建立基于多模态脑机接口技术的意识检测方法，鉴别意识障碍患者中具有明确意识的认知运动分离患者（CMD）；（2）研究正常人、CMD患者、和非CMD意识障碍患者之间的睡眠脑电模式差异，提出联合卷积神经网络和长短期记忆网络的睡眠脑电多尺度高维特征混合网络模型，建立基于睡眠脑电的意识检测方法；（3）设计面向意识障碍患者意识检测的系列脑机接口系统，分别实现意识检测和康复预测的示范性临床应用。本项目的研究成果可拓展意识障碍患者意识检测的现有知识，对基于脑机接口的意识障碍患者临床诊断和预后具有重要意义，为脑机接口临床应用奠定坚实基础。

Accurate awareness detection in patients with disorders of consciousness (DOC) is a worldwide problem that puzzles clinicians at home and abroad. To solve the lack of behavioral response and high rate of clinical misdiagnosis in patients with DOC, this project conduct a research of awareness detection technology based on brain computer interface (BCI) and its clinical application. The main research contents include three parts. (1) We design a stimuli paradigm that integrates P300 and emotion components, proposes a multi-feature joint emotion recognition algorithm based on Long Short-Term Memory (LSTM), and establishes a method of awareness detection based on multimodal BCI technology to identify the patients with cognitive motor dissociation (CMD) with clear evidence of consciousness. (2) To study the differences of sleep EEG patterns among normal people, CMD patients and non-CMD patients, we propose a Convolutional Neural Networks (CNN)-LSTM multi-scale and high-dimensional feature hybrid network model of sleep EEG, and establish a method of awareness detection based on sleep EEG. (3) We design a series of BCI systems for awareness detection in patients with DOC, and further realize the exemplary clinical applications for awareness detection and prognosis. The research results of this project can expand the existing knowledge of awareness detection of patients with DOC, which is of great significance for clinical diagnosis and prognosis in patients with DOC based on BCI, and lay a solid foundation for clinical application of BCI technology.