Structural analysis of high-resolution magnetic resonance images has revolutionized the ability to diagnose and follow progression of neurodegenerative disease. Changes in the spatial characteristics of the cerebral cortex may herald disease well before clinical signs become apparent. Additionally, changes in cortical shape could be used as a surrogate biomarker for disease progression or for the success of therapeutic intervention. The purpose my research is to apply a fractal analysis technique to quantitatively characterize changes in the shape of the cerebral cortex caused by atrophy. The images in this study were selected from the Alzheimer?s Disease Neuroimaging Initiative database. Two-dimensional coronal and axial profiles of the cerebral cortex were created from pial and grey/white boundaries generated using the segmentation software FreeSurfer. The images were then analyzed using a freely available fractal analysis program called Fractal Dimension Calculator. The fractal analysis yielded a quantitative metric that separated patients with normal cognition from those with moderate to advanced degenerative disease. This novel application of fractal analysis may be useful for quantifying changes in shape not only in Alzheimer?s disease, but also in other neurological diseases associated with cerebral atrophy. Analysis of cortical fractal dimension may improve the quality of neuroimaging biomarkers when combined with other complementary structural analysis methods.
