功能磁共振脑连接度中心性对精神分裂症症状的预测评估  

Prediction and Evaluation of Schizophrenia Symptom by fMRI Brain Connectivity Centrality

作　　者：段明君[1] 李宁[1,2] 常鑫 李诚[1] 张楠[1,2] 任有为[2] 杨黄浩 贺辉 罗程 尧德中[2,3] Duan Mingjun;Li Ning;Chang Xin;Li Cheng;Zhang Nan;Ren Youwei;Yang Huanghao;He Hui;Luo Cheng;Yao Dezhong(The Clinical Hospital of Chengdu Brain Science Institute,University of Electronic Science and Technology of China,Chengdu 610030,China;School of life Science and technology,University of Electronic Science and Technology of China,Chengdu 611731,China;Sichuan Institute for Brain Science and Brain Inspired Intelligence,Chengdu 611731,China)

机构地区：[1]电子科技大学成都脑科学研究院临床医院,成都610030 [2]电子科技大学生命科学与技术学院,成都611731 [3]四川省脑科学与类脑智能研究院,成都611731

出　　处：《中国生物医学工程学报》2020年第1期19-25,共7页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金(81471638)。

摘　　要：精神分裂症是一种严重影响患者身心健康的精神类疾病,绝大部分患者需要终身服药以控制症状,因此药物疗效的评估预测对于患者治疗方法的选择十分重要。采用功能磁共振的脑连接度中心度(DC)指标,分析临床常用抗精神病药物利培酮和氯氮平对大脑功能的影响,并尝试建立预测评估两种药物疗效的方法。基于44例精神分裂症患者(20例利培酮治疗组、24例氯氮平治疗组)和30例健康志愿者的脑静息态功能磁共振成像数据,首先采用方差分析得到患者组与正常组间的差异脑区,然后根据这些脑区的DC值,用多变量支持向量回归算法构建回归模型,探讨用DC值评测患者症状的可行性。结果发现,两个患者组的丘脑、脑岛及初级感知运动相关皮层的功能连接特征存在显著的异常(P<0.05),且可以有效地预测氯氮平组患者的阴性症状(r=0.448,P<0.05),但未能较好地预测利培酮患者的症状。在统计对比分析的基础上,利用回归模型,标识出抗精神疾病药物作用的部分重要脑区,因而有助于将来患者治疗的药物选择。Schizophrenia is a kind of serious mental disorder.Most of patients with schizophrenia need lifetime treatment of antipsychotic drugs.It is very important to evaluate the effects of drugs for the choice of therapeutic approaches.In this study,the degree centrality(DC)of brain functional connectivity was used to investigate the effects of taking two clinical different kinds of antipsychotic drugs:risperidone and clozapine,aiming to construct a model to predict their effects.The resting-state functional magnetic resonance imaging data were collected from 20 schizophrenia patients therapized with risperidone,24 patients were treated with clozapine and30 healthy as controls.The voxel-based DC was evaluated using analysis of variance.The DC in the regions with difference between patients and healthy controls was used in support vector regression(SVR)to construct predictive model for the scores of syndromes in patients with different drugs respectively.Experimental results indicated that the DC was altered in thalamus,insula and primary perception and motor-related cortexes in schizophrenia contrast to healthy controls(P<0.05).In addition,The SVR model based on the DC in these regions could predict the negative syndrome of patients’treatment with clozapine(r=0.448,P<0.05),while the syndrome of patients was not predicted in the risperidone group.Therefore,using SVR for the regions with difference in analysis of variance,several main regions related with antipsychotic drugs in schizophrenia were determined,and these findings would contribute to the choice antipsychotic drugs in the future.

关 键 词：功能磁共振成像 度中心度 支持向量回归 精神分裂症 药物治疗 

分 类 号：R749.3[医药卫生—神经病学与精神病学] R445.2[医药卫生—临床医学]