基于时空图卷积神经网络的精神分裂症识别  

Spatial-temporal graph convolutional neural network for schizophrenia recognition

作　　者：徐信毅 李斌[2] 朱耿 周宇星 林萍[1] 李晓欧[1,2,3] XU Xinyi;LI Bin;ZHU Geng;ZHOU Yuxing;LIN Ping;LI Xiao'ou(School of Health Science and Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;Shanghai Yangpu District Mental Health Center,Shanghai 200093,China;School of Medical Instruments,Shanghai University of Medicine and Health Sciences,Shanghai 201318,China)

机构地区：[1]上海理工大学健康与科学工程学院,上海200093 [2]上海市杨浦区精神卫生中心,上海200093 [3]上海健康医学院医疗器械学院,上海201318

出　　处：《中国医学物理学杂志》2024年第2期227-232,共6页Chinese Journal of Medical Physics

基　　金：上海市科委地方院校能力建设项目(22010502400);上海健康医学院精神卫生临床研究中心项目(20MC2020005);上海市杨浦区技术委员会卫生健康委员会科研项目(YPM202114)。

摘　　要：提出一种基于时空图卷积神经网络的精神分裂症患者分类方法,与过往仅分析脑电中的时频特征而忽略各脑区之间空间特征的主流方法不同,模型主要通过用不同通道之间小波相干系数构成的邻接矩阵和脑电序列进行图卷积的方式获取其中的空频特征,再通过一维时间卷积获取其中的时频特征,经过多次卷积后将处理过的矩阵扁平化后输入分类模型。实验结果表明本文方法在公开数据集Zenodo上的分类准确率达96.32%,证明本文方法的有效性,也证明融合时频、空频特征对精神分裂症诊断的优势。A spatial-temporal convolutional neural network-based method is proposed for schizophrenia classification.Unlike the mainstream methods that only analyze the temporal frequency features in EEG and ignore the spatial features between brain regions,the model mainly obtains the spatial-frequency features by convolving the adjacency matrix composed of wavelet coherence coefficients between different channels and EEG sequences,and then extracts the temporal-frequency features through one-dimensional temporal convolution.The processed matrix is flattened after multiple convolutions and input to the classification model.Experimental results show that the method has a classification accuracy of 96.32% on the publicly available dataset Zenodo,demonstrating its effectiveness and exhibiting the advantages of fusing temporalfrequency and spatial-frequency features for schizophrenia diagnosis.

关 键 词：精神分裂症 时频特征 空频特征 图神经网络 

分 类 号：R318[医药卫生—生物医学工程] TP183[医药卫生—基础医学]