基于MRI图像后处理技术的双重病理致颞叶癫痫患者脑区结构特征分析  

Structural features of brain regions in temporal lobe epilepsy with dual pathology: An analysis based on MRI postprocessing technology

作　　者：孔钰 刘瑞寒 姚磊 王淑贞 刘春朝 孔庆霞[5] KONG Yu;LIU Ruihan;YAO Lei;WANG Shuzhen;LIU Chunzhao;KONG Qingxia(College of Materials Science and Engineering,Qingdao University,Qingdao 266071,China)

机构地区：[1]青岛大学材料与工程学院,山东青岛266071 [2]济宁医学院附属医院医学影像科 [3]济宁医学院附属医院儿科 [4]济宁医学院临床医学院 [5]济宁医学院附属医院神经内科

出　　处：《精准医学杂志》2024年第6期500-504,共5页Journal of Precision Medicine

基　　金：山东省自然科学基金项目(ZR2023QH076)。

摘　　要：目的基于磁共振成像(MRI)后处理技术,对双重病理(DP)致耐药性颞叶癫痫(TLE)患者脑区结构变化特征进行自动精细脑分割及脑区数据定量分析。方法选择2017年1月—2023年12月我院DP致TLE患者35例设为TLE-DP组,并选择同期32例健康成年人为对照组。收集两组受试者的MRI图像资料,进行自动精细脑分割及脑区数据定量分析。比较两组受试者各脑区的体积、体积分数、皮质厚度、皮质表面积及皮质曲率,使用受试者工作特征(ROC)曲线及曲线下面积(AUC)评估上述指标对DP致TLE的诊断效能。结果自动精细脑分割及脑区数据定量分析显示,两组受试者30个脑区体积、12个脑区体积分数、2个脑区皮质厚度、11个脑区皮质表面积、3个脑区皮质曲率,以及21个海马亚区体积和1个海马亚区体积分数差异显著(t=-2.151~3.882,P<0.05)。ROC曲线分析显示,部分脑区体积、脑区皮质表面积和海马亚区体积对DP致TLE诊断效能较高(AUC>0.7)。结论基于MRI后处理技术的自动精细脑分割及脑区数据定量分析对DP致TLE患者的诊断价值较高。Objective To perform automatic fine brain segmentation and brain region quantitative analysis based on magnetic resonance imaging(MRI)postprocessing technology,and to investigate the structural features of brain regions in drug-resis-tant temporal lobe epilepsy(TLE)with dual pathology(DP).Methods Thirty-five TLE patients with DP treated at our hospital from January 2017 to December 2023 were selected as TLE-DP group,and 32 healthy adults from the same period were selected as control group.The MRI images of the two groups were collected for automatic fine brain segmentation and brain region quantitative analysis.The volume,volume fraction,cortical thickness,cortical surface area,and cortical curvature of each brain region were compared between the two groups.The receiver operating characteristic(ROC)curve and area under the curve(AUC)were used to evaluate the above parameters for the efficacy of diagnosis of TLE with DP.Results Automatic fine brain segmentation and brain region quantitative analysis indicated that the volume of 30 brain regions,volume fraction of 12 brain regions,cortical thickness of 2 brain regions,cortical surface area of 11 brain regions,cortical curvature of 3 brain regions,volume of 21 hippocampal subregions,and volume fraction of 1 hippocampal subregion were significantly different between the two groups(t=-2.151-3.882,P<0.05).The ROC curve analysis showed that the volume and the cortical surface area of some brain regions and the vo-lume of some hippocampal subregions were effective in the diagnosis of TLE with DP(AUC>0.7).Conclusion Automatic fine brain segmentation and brain region quantitative analysis based on MRI postprocessing technology are of high diagnostic value in TLE patients with DP.

关 键 词：癫痫 颞叶 图像处理 计算机辅助 磁共振成像 皮质发育畸形 海马 硬化 

分 类 号：R742.1[医药卫生—神经病学与精神病学]