Microstructural White Matter Alterations in Mild Cognitive Impairment and Alzheimer’s Disease

Fu Xiuwei,Shrestha Susan,Sun Man,Wu Qiaoling,Luo Yuan,Zhang Xianchang,Yin Jianzhong,Ni Hongyan
DOI: https://doi.org/10.1007/s00062-019-00805-0
2019-01-01
Clinical Neuroradiology
Abstract:Purpose To investigate microstructural alterations in white matter in mild cognitive impairment (MCI) and Alzheimer's disease (AD) using neurite orientation dispersion and density imaging (NODDI) and to assess the potential diagnostic performance of NODDI-derived parameters. Methods In this study 14 MCI patients, 14 AD patients, and 14 healthy controls (HC) were recruited. The diffusion tensor imaging(DTI)-derived fractional anisotropy (FA) and NODDI-derived neurite density index (NDI), orientation dispersion index (ODI), and volume fraction of isotropic water molecules (Viso) were calculated from the diffusion data. The tract-based spatial statistics (TBSS) method was used for statistical analysis with one-way ANOVA. The correlations between the parameter values and mini-mental state examination (MMSE) and Montreal cognitive assessment (MoCA) scores were examined. A receiver operating characteristic (ROC) curve was conducted to assess the diagnostic performance of different parameters. Results Compared with the HC group, the NDI and ODI values decreased significantly and the Viso values were significantly increased in the MCI and AD groups (p< 0.01, threshold-free cluster enhancement (TFCE)-corrected); however, there were no significant differences in FA values in the MCI group. The NDI, ODI, and Viso values of multiple fibers were significantly correlated with MMSE and MoCA scores. For the diagnosis of AD, the area under the ROC curve (AUC) for the NDI value of the splenium of corpus callosum was larger than the FA value (AUC= 0.885, 0.714,p= 0.042). The AUC of the Viso value of the right cerebral peduncle was larger than FA value (AUC= 0.934, 0.531,p= 0.004). Conclusion The NDI is more sensitive to white matter microstructural changes than FA and NODDI could be superior to DTI in the diagnosis of AD.
What problem does this paper attempt to address?