优化VBM算法和DARTEL算法分析阿尔茨海默病MRI

目的 探讨优化VBM算法和DARTEL算法在分析阿尔茨海默病（AD）患者MRI的差别。方法 利用两种算法对14例AD患者（AD组）和23名健康对照（NC组）的MRI进行分析，并对其全脑灰质进行基于体素的统计学比较。结果 两种算法均报告颞叶右侧、海马、海马沟、海马旁回、杏仁核、枕叶深部存在灰质萎缩，但DARTEL算法报告的簇明显多于优化VBM算法；另外，部分区域在设置P<0.005时优化VBM算法仍未见报告。结论 分析AD患者的MRI时，采用DARTEL算法比优化VBM算法更合理。

Optimized VBM and DARTEL algorithm in analyzing MRI of Alzheimer's disease

Objective To compare optimized voxel-based morphometry (VBM) algorithm and diffeomorphic anatomical registration through exponentiated lie (DARTEL) algorithm in analyzing MRI of Alzheimer's disease (AD). Methods MRI of 14 patients with AD and 23 normal controls were analyzed with VBM and DARTEL algorithm, respectively. The differences of gray matter volume in the whole brain were analyzed statistically based on voxel. Results Both algorithms showed grey matter atrophy in right temporal lobe, hippocampus, hyppocampal suleus, parahippocampal, amygdala, sub-gyral, but the clusters displayed with DARTE algorithm were obviously more than that with optimized VBM algorithm. Further more, some anatomic regions were not displayed with optimized VBM algorithm even when P<0.005. Conclusion DARTEL algorithm is more robust than optimized VBM algorithm on MRI analysis of AD.