基于潜类别分析模型的抑郁症分型的研究进展  

Progress in researches on latent class analysis based subtyping of depression

作　　者：王成蕾 吴志国[1] 方贻儒[1,2,3] WANG Cheng-lei;WU Zhi-guo;FANG Yi-ru(Division of Mood Disorder,Shanghai Mental Health Center,Shanghai Jiao Tong University School of Medicine,Shanghai 200030,China;Shanghai Institute for Biological Sciences,Chinese Academy of Sciences,CAS Center for Excellence in Brain Science and Intelligence Technology,Shanghai 200031,China;Shanghai Key Laboratory of Psychotic Disorders,Shanghai 200030,China)

机构地区：[1]上海交通大学医学院附属精神卫生中心心境障碍科,上海200030 [2]中国科学院上海生命科学研究院神经科学研究所,中国科学院脑科学与智能技术卓越创新中心,上海200031 [3]上海市重性精神病重点实验室,上海200030

出　　处：《上海交通大学学报（医学版）》2018年第6期676-679,共4页Journal of Shanghai Jiao tong University:Medical Science

基　　金：国家重点研发计划(2016YFC1307100);国家自然科学基金面上项目(91232719);上海申康医院发展中心项目(SHDC12015131,SHDC12015302);上海交通大学“医工交叉”研究基金(YG2015MS47)。

摘　　要：抑郁症具有高度临床异质性,根据症状学特征对抑郁症进行同质性分型可能有助于更具针对性地选择治疗决策,进行疗效评估和预后判断。潜类别分析可通过对潜变量的模型化分析对异质性人群进行同质性分类,有助于提高疾病临床分型的准确性,在抑郁症分型研究中逐渐获得广泛应用。文章对抑郁症临床描述性分型的现状及潜类别分析在疾病分型中的应用进行综述。Depression is a highly heterogeneous syndrome.Homogeneous subtypes according to symptomatology of illness may contribute to development of individualized treatment,assessment on outcomes and prognosis.Latent class analysis is a flexible statistical approach to determine classes with similar symptom profiles in a heterogeneous group,which has been widely used in data-driven subtyping of depression to increase accuracy of subtyping.This article reviewed existing symptom-based subtypes of depression and findings of researches on latent class analysis based illness subtyping.

关 键 词：抑郁症 分型 潜在类别分析 

分 类 号：R749.4[医药卫生—神经病学与精神病学]