Abstract #0989

Voxel-Based Analysis of Alzheimer's Disease Using Apparent Fibre Density

David Raffelt¹,², J-Donald Tournier¹,³, Stuart Crozier⁴, Kathryn A. Ellis⁵, Ralph Martins⁶, Victor L. Villemagne,³⁷, Colin L. Masters⁷, David Ames⁸, Christopher C. Rowe³, Olivier Salvado², Alan Connelly¹,³

¹Brain Research Institute, Florey Neuroscience Institutes, Melbourne, VIC, Australia; ²The Australian E-Health Research Centre, CSIRO, Brisbane, QLD, Australia; ³Department of Medicine, University of Melbourne, Melbourne, VIC, Australia; ⁴Biomedical Engineering, School of ITEE, University of Queensland, Brisbane, QLD, Australia; ⁵Department of Psychiatry, University of Melbourne, Melbourne, VIC; ⁶Edith Cowan University, Centre of Excellence for Alzheimers Disease Research & Care, Joondalup, WA, Australia; ⁷The Mental Health Research Institute, University of Melbourne, Melbourne, VIC, Australia; ⁸Department of Radiology, University of Melbourne, Melbourne, VIC, Australia

In this work, we investigate Alzheimers Disease (AD) using a recently developed method for voxel-based analysis (VBA) of diffusion-weighted MRI called Apparent Fibre Density (AFD). Unlike existing measures such as Fractional Anisotropy, AFD permits population differences to be localised in both the spatial and orientation domains. This enables pathology-induced changes to be attributed to a single fibre population in a region containing multiple fibres. Our findings demonstrate a significant decrease in AFD in AD patients compared to healthy subjects within voxels and orientations corresponding to the cingulum, corpus callosum, uncinate fasciculus, and superior longitudinal fasciculus and anterior commissure.