摘要
目的:提出一种基于局部一致性(regional homogeneity,ReHo)和多变量模式分析(multivariate pattern analysis,MVPA)的颞叶癫痫(temporal lobe epilepsy,TLE)诊断方法。方法:收集17例TLE患者(作为TLE组)和32例正常被试(作为正常对照组)的静息态功能磁共振数据,并进行ReHo分析。在ReHo分析的基础上,利用MVPA对TLE患者和正常被试进行分类,并计算各脑区的权重。根据脑区权重选择感兴趣区(region of interest,ROI),分析TLE组ROI的ReHo值与智力评分的相关性;利用准确率、敏感度、精确度等评价指标和ROC曲线评估该方法的分类性能。结果:该方法分类的准确率为87.27%、敏感度为73.91%、精确度为94.44%,AUC值为0.94;右侧小脑后叶(CRBLCrus2_R)、右侧舌回(LING_R)、右侧楔叶(CUN_R)和左侧颞上回(STG_L)在识别TLE中贡献较大;言语商与STG_L的ReHo值之间存在明显的正相关,操作商和总智商与LING_R的ReHo值之间呈负相关。结论:提出的方法可以辅助TLE诊断,为探索TLE的发病机制提供了参考。
Objective To propose a diagnosis method for temporal lobe epilepsy(TLE) based on regional homogeneity(ReHo)and multivariate pattern analysis(MVPA). Methods Resting-state functional magnetic resonance data were collected from 17TLE patients(TLE group) and 32 normal controls(normal control group), and then the data collected underwent regional homogeneity(ReHo) analysis. Based on ReHo analysis the TLE patients and normal controls were classified with MVPA, and the weights of each brain region were calculated. Regions of interest(ROI) were selected according to the brain region weights, and the correlation between the ReHo values of ROI and IQ scores in the TLE group was analyzed;the classification performance of the method was evaluated using the indexes such as accuracy, sensitivity, precision and ROC curve. Results The method proposed had the classification accuracy being 87.27%, the sensitivity being 73.91%, the precision being94.44% and the AUC value being 0.94, confirming that the right posterior cerebellum lobe(CRBLCrus2-R), the right lingual gyrus(LING_R), the right cuneus(CUN-R) and the left superior temporal gyrus(STG_L) contributed a lot to the diagnosis of TLE. There was a significant positive correlation between verbal IQ and ReHo values of STG_L, and a negative correlation between performance IQ and full IQ and ReHo values of LING_R. Conclusion The method proposed assists in the diagnosis of TLE and provides references for exploring the pathogenesis of TLE. [Chinese Medical Equipment Journal,2022,43(6):1-6,17]
作者
郝思瑶
杨春兰
李志梅
任洁钏
HAO Si-yao;YANG Chun-lan;LI Zhi-mei;REN Jie-chuan(Department of Biomedical Engineering,Faculty of Environment and Life of Beijing University ofTechnology,Beijing 100124,China;Epilepsy Department of Neuromedical Center,Beijing Tiantan Hospital of Capital Medical University,Beijing 100050,China)
出处
《医疗卫生装备》
CAS
2022年第6期1-6,17,共7页
Chinese Medical Equipment Journal
基金
国家自然科学基金项目(31640035,81601126)
北京市自然科学基金项目(4162008)
北京市科技新星计划项目(Z161100004916157)
中国抗癫痫协会-UCB基金项目(2017010)。
关键词
静息态功能磁共振成像
颞叶癫痫
ReHo
MVPA
智力评分
resting-state functional magnetic resonance imaging
temporal lobe epilepsy
regional homogeneity
multivariate pattern analysis
IQ score