基于机器学习的急性期精神分裂症心理理论能力对社会功能的预测作用  

Predictive role of machine-learning-based theory of mind in acute phase schizophrenia on social functioning

作　　者：仲捷 朱虹 郑思思[1,2] 贾竑晓 Zhong Jie;Zhu Hong;Zheng Sisi;Jia Hongxiao(The National Clinical Research Center for Mental Disorders&Beijing Key Laboratory of Mental Disorders,Beijing Anding Hospital,Capital Medical University,Beijing 100088,China;Advanced Innovation Center for Human Brain Protection,Capital Medical University,Beijing 100069,China)

机构地区：[1]首都医科大学附属北京安定医院国家精神心理疾病临床医学研究中心精神疾病诊断与治疗北京市重点实验室,北京100088 [2]首都医科大学人脑保护高精尖创新中心,北京100069

出　　处：《首都医科大学学报》2023年第4期596-601,共6页Journal of Capital Medical University

基　　金：北京市医院管理中心临床医学发展专项(ZYLX202129);北京市医院管理中心“登峰”人才培养计划项目(DFL20191901)。

摘　　要：目的探讨基于机器学习算法的急性期精神分裂症患者心理理论能力对社会功能的预测作用。方法选取2013至2017年来自首都医科大学附属北京安定医院门诊的急性期精神分裂症患者90例,采用阳性和阴性症状量表(Positive and Negative Syndrome Scale,PANSS)评估急性期精神分裂症患者的精神症状;选取威斯康星卡片分类测验(Wisconsin Card Sorting Test,WCST)中的完成分类数、持续性操作测验(Continuous Performance Test,CPT)平均反应时间、成人韦氏智力量表(Wechsler Adult Intelligence Scale,WAIS)中的数字广度项目评估神经认知;选取心理理论能力中的一级错误信念、二级错误信念、失言识别任务和眼区阅读测试评估社会认知;选用个人与社会功能量表(Personal and Social Functioning Scale,PSP)评估急性期精神分裂症患者的社会功能。应用机器学习算法评估急性期精神分裂症心理理论能力对社会功能的预测作用。结果相关分析结果显示急性期精神分裂症患者的PSP评分与阴性症状(r=-0.271,P=0.010)、WCST完成分类数(r=-0.128,P=0.328)呈负相关,与失言识别评分呈正相关(r=0.410,P<0.001);二级错误信念及失言识别任务对急性期精神分裂症患者的社会功能具有良好的预测作用,曲线下面积(area under the curve,AUC)值为80.9%。结论急性期精神分裂症患者的社会功能损害与阴性症状及心理理论能力密切相关,心理理论能力对急性期精神分裂症患者的社会功能具有良好的预测作用。Objective To explore the predictive role of machine-learning-algorithm-based mind theory in social function of patients with schizophrenia in the acute phase.Methods Ninety patients with acute schizophrenia from the outpatient clinic of Beijing An Ding Hospital,Capital Medical University from 2013 to 2017 were selected,and the Positive and Negative Syndrome Scale(PANSS)was used to assess the psychiatric symptoms of the patients with acute schizophrenia.The number of completed categories on the Wisconsin Card Sorting Test(WCST),the mean reaction time on the Continuous Practice Test(CPT),and the number breadth items on the Wechsler Adult Intelligence Scale(WAIS)were selected to assess neurocognition.First-order false belief test,Second-order false belief test,Faux-pas test,and Reading the mind in the eyes test were selected to assess social cognition.The personal and social functioning scale(PSP)was selected to assess social function.The machine learning algorithms was applied to assess the predictive role of psychological theory of competence in acute phase schizophrenia on social functioning.Results The results of the correlation analysis showed that PSP scores of the acute schizophrenia patients were the negatively correlated with negative symptoms(r=-0.271,P=0.010),the number of WCST(r=-0.128,P=0.328),and positively correlated with faux-pas test(r=0.410,P<0.001).Second-false Belief Test and Faux-pas Test were good predictors of social function in the acute schizophrenia patients.The area under the curve(AUC)value was 80.9%.Conclusions The impairment of social function in the patients with acute schizophrenia is closely related to the negative symptoms and theory of mind,indicating that the theory of mind is a good predictor of social function in the patients with schizophrenia in the acute phase.

关 键 词：精神分裂症 心理理论能力 社会功能 机器学习 

分 类 号：R74[医药卫生—神经病学与精神病学]