基于MR图像的AD和MCI与海马体子域的关联研究

Association of AD and MCI with Hippocampus Subfields based on MR Images

目的研究磁共振(Magnetic resonance,MR)脑图像中海马体子域的提取方法及海马子域体积与阿尔茨海默病(Alzheimer’s disease,AD)及其早期阶段轻度认知障碍(mild cognitive impairment,MCI)进行关联分析的方法,为AD的早期诊断提供依据。方法本实验共选取310组T1加权的MRI扫描图像,其中55例AD患者,180例为MCI患者,75例为老年对照CN。首先,对图像进行颅骨剔除、体积标签化、强度归一化、白质分割、表层图谱配准等处理。然后,使用贝叶斯建模方法,建立一个明确的计算模型,生成海马区周围的MRI图像,并使用这个模型来获得海马体子域的全自动分割结果。最后,运用统计学方法获得待试人群MR图像中海马体子域体积与AD组、MCI组及正常对照组的关联关系。结果海马体子域的体积在AD患者、MCI患者与健康人相比,海马体整体体积有明显的萎缩,具体表现为海马体的各子域体积都存在不同程度的萎缩。

Objective To study the extraction method of hippocampal subfields in magnetic resonance （MR） brain imaging and the relationship between the volume of hippocampus and Alzheimer＇s disease （AD） and its early stage mild cognitive impairment （MCI） The method of association analysis provides the basis for the early diagnosis of AD. Methods In this study, a total of 310 T1-weighted MRI scans were selected, of which 55 cases of AD patients, 180 cases of MCI patients, 75 cases of elderly controls CN. First of all, the images were skull removed, volume labeling, intensity normalization, white matter segmentation, surface map registration and others. Then, using the Bayesian modeling method, a clear computational model is built to generate MRI images around the hippocampus and use this model to obtain fully automated segmentation of the hippocampus subfields. Finally, the correlation between hippocampal subdomains volume and AD group, MCI group and normal control group was obtained by statistical methods.Results The volume of hippocampal subfields in AD patients, MCI patients compared with healthy people, the overall volume of the hippocampus significantly atrophy, the specific performance of the hippocampus volume of each sub-domain there are varying degrees of atrophy. Conclusion The volume of hippocampus and its subfields is feasible as a marker for the early diagnosis of AD.