P2-175: diffusion tensor imaging detection of cognitive changes in amnestic mci

Diffusion Tensor Imaging (DTI) is an MRI technique for defining anatomical structure and detecting changes in brain white matter pathways. AD is known to cause significant white matter (WM) degeneration late in its course. However, less is known about DTI's capability of detecting WM changes in the earliest clinically detectable stage of the disease, early amnestic mild cognitive impairment (eMCI). DTI is a promising method of investigating microscopic pathologic changes in single WM tracts. Patient data was retrieved from the Alzheimer's Disease Neuroimaging Initiative (ADNI), including three cohorts of patients: cognitively normal (CN), n=20; eMCI, n=21; and AD patients, n=9. We designed 5 regions of interest (ROIs) to compare methods of measuring DTI data against known diffusion characteristics in cognitively normal brains. Changes in the axial diffusivity (AxD), radial diffusivity (RD), mean diffusivity (MD), and fractional anistropy (FA) were measured for each ROI for every patient. To compare various approaches to sampling, we created a sub-set of 3 ROIs for each WM tract: eroded, baseline, and expanded To batch-process MRIs of patients and extract useable DTI data, the ROIs had to fit a wide range of brain architecture for data extraction, accounting for skewing during image compression and overlay. ROIs were evaluated for their ability to show significant differences between patient cohorts in FA, AxD, RD, and MD, and for internal variability. Among all the WM tracts, the CING showed the best statistical differentiation between CN, eMCI, and AD patients.). This was validated against known and hypothesized pathologic changes. Analysis of variance across the patient cohorts showed significant trends for progressing disease pathology in the CING. e MCI was significantly different from CN demonstrated (P