精神分裂症亚型划分的数学模型一、数学模型的建立  

A Mathematic model for subtyping schizophrenia 1.Establishment of the model

作　　者：金卫东[1,2] 刘仁刚[1,2] 王高华[1,2] 刘铁榜 臧德馨 

机构地区：[1]湖北医学院附一院精神科 [2]湖南医科大学

出　　处：《临床精神医学杂志》1996年第5期267-269,共3页Journal of Clinical Psychiatry

摘　　要：应用Ａｎｄｒｅａｓｅｎ的精神分裂症阳性型、混合型、阴性型分型标准为金标准，并对金标准筛选的三类型病人进行简明精神病症状评定量表（ＢＰＲＳ）评定，由此得出阴性症状Ｘ１、阳性症状Ｘ２、和非特异症状Ｘ３，以糊模数学隶属度为方法，求得数学模型为μ（Ｇｊ）（Ｘｉ）＝Σ３ｉ＝１ＷｉｊμＧｊ（Ｘｉｊ），（ｉ＝１，２，３．ｊ＝１，２，３．），将原金标准病例Ｘ１、Ｘ２、Ｘ３代入后发现该模型对阳性型的符合率为８２．６１％，混合型符合率为７３．３３％，阴性型符合率为８８．８９％，总符合率为８２．１４％。bjective:To establish a mathematic model for classifying schizophrenia into the positive,mixed and negative subtypes according to Andreasens criteria.Method:Three groups of symptoms were selected from the BPRS:negative (X1),positive(X2),and nonspecific(X3).The mathematic model was established according to the fuzzy mathematic principle,from which the membership of different symptoms in different subtypes was derived.X1,X2and X3were put into the model to examine its accuracy. Results: The mathematic model is as following:μ (Gj) (Xij)= Σ3i=1 Wijμ Gj (Xij),(i=1,2,3.j=1,2,3.).Discussion: The accuracy of this model was satisfactory:82 61% for positive type,73 33% for mixed type,and 88 89% for negative type.Overall accuracy was 82 14%.It is helpful for differentiating the subtypes of schizophrenia.

关 键 词：精神分裂症 模糊数学 数学模型 分型 亚型 

分 类 号：R749.302[医药卫生—神经病学与精神病学]