利用多模态心理生理计算研究抑郁症诊断和疗效评价方法

项目名称： 利用多模态心理生理计算研究抑郁症诊断和疗效评价方法

项目编号： No.61471291

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 无线电电子学、电信技术

项目作者： 闫相国

作者单位： 西安交通大学

项目金额： 80万元

中文摘要： 抑郁症是全球第四大疾病，患者自杀率高于普通人群数十倍，对患者、家庭和社会造成严重危害。目前临床诊疗以主观评价为主，缺乏有效的客观诊断方法，心理生理计算为抑郁症客观诊断提供了一条崭新的研究途径。本课题综合应用近红外光谱技术和脑电图，从社会行为特性、情绪面孔识别、长时间睡眠监测三个方面，对抑郁症诊断和疗效评价方法进行多角度、多时间尺度研究。重点研究：从人类社会交往中普遍存在的合作、竞争行为出发，通过行为特征建模，建立基于人-机交互研究抑郁症患者社会行为特征的新模式；利用情绪面孔识别任务，研究抑郁症患者出现负性偏向的信息加工机制；充分发挥长时间多模态监测优势，从不同角度研究睡眠过程中的神经血管耦合关系；利用现代信号处理方法和信号同步分析技术，从多层次提取抑郁症敏感特征信息；利用多模态生理数据、特征融合和分类技术，研究有效的抑郁症临床诊断及疗效评价方法。课题研究具有重要的社会意义和科学价值。

中文关键词： 近红外光谱；脑电图；抑郁症；数据融合；心理生理计算

英文摘要： Depression is the fourth cause of disease burden in the world. The suicide rate for patients with depression is higher than the rate for national population several tens times. It is a serious threat to patients, families, and society. However, currently clinical diagnosis for depression is based on subjective evaluation and is lack of effective and objective diagnostic methods. Psychophysiological computing paves a new way for objective evaluation of depression. The project will utilize comprehensively near-infrared spectroscopy(NIRS) and EEG measurements. Its object is to present methods of diagnosis and treatment evaluation for depression from three aspects of social-behavioral characteristics, emotional face recognition, and long-term sleep monitoring. The research contents: (1) According to the universal behaviors (cooperation and competition) of social interaction for humans, a cooperation model and a competition model are established. Then a new mode based on human-computer interaction is set up for the study of social behavior characteristics of patients with depression. (2)With the help of emotional face recognition tasks, the negative bias information processing mechanisms for patients with depression are studied. (3)Investigation of neurovascular coupling relationship during sleep from different angles, which will give full play to the unique advantage of long-term multi-modal monitoring techniques. (4)Extracting sensitive features associated with depression by means of modern signal processing methods and synchronization analysis techniques. (5)Study of effective methods of clinical diagnosis and treatment evaluation of depression by the aid of multi-modal physiological data and feature fusion and classification technology. The Investigation has important social significance and scientific value.

英文关键词： NIRS;EEG;Depression;Data Fusion;Psychophysiological Computing