Whole brain white matter histogram analysis of diffusion tensor imaging data detects microstructural damage in mild cognitive impairment and Alzheimer’s disease patients

ABSTRACT Background: Amnestic mild cognitive impairment (MCI) is a transitional stage between normal aging and Alzheimer’s disease (AD). However, the clinical conversion from MCI to AD is unpredictable. Hence, identification of non-invasive biomarkers able to detect early changes induced by dementia is a pressing need. Purpose: To explore the added value of histogram analysis applied to measures derived from diffusion tensor imaging (DTI) for detecting brain tissue differences between AD, MCI and healthy subjects (HS). Study type: Retrospective. Population/subjects: Local cohort (57 AD, 28 MCI, 23 HS), Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort (41 AD, 58 MCI, 41 HS). Field Strength: 3T. Dual echo TSE; FLAIR; MDEFT; IR-SPGR; DTI. Assessment: Normal appearing white matter (NAWM) masks were obtained using the T1-weighted volumes for tissue segmentation and T2-weighted images for removal of hyperintensities/lesions. From DTI images, fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AXD) and radial diffusivity (RD) were obtained. NAWM histograms of FA, MD, AXD and RD were derived and characterized estimating: peak height, peak location, mean value (MV), and quartiles (C25, C50, C75), which were compared between groups. Receiver operating characteristic (ROC) and area under ROC curves (AUC) were calculated. To confirm our results, the same analysis was repeated on ADNI dataset. Statistical tests: One-way ANOVA, post-hoc Student’s t-test, multi-class ROC analysis. Results: For the local cohort, C25 of AXD had the maximum capability of group discrimination with AUC of 0.80 for “HS vs patients” comparison and 0.74 for “AD vs others” comparison. For ADNI cohort, MV of AXD revealed the maximum group discrimination capability with AUC of 0.75 for “HS vs patients” comparison and 0.75 for “AD vs others” comparison. Data conclusion: AXD of NAWM might be an early marker of microstructural brain tissue changes occurring during AD course and might be useful for assessing disease progression.