Robust Assessment of Cortical Thinning in Brain MRI by Generalized Complexity Measures
Young K, Schuff N
University of California San Francisco, Veterans Affairs Medical Center
Recent brain MRI studies of aging and dementia have suggested methods for measuring cortical thinning. Though promising, these methods depend on a number of parameters. A robust method for detecting cortical thinning via complexity measures, requiring a single choice, the distance scale for the analysis, is presented in this study. The methods are demonstrated using a data set of 50 segmented T1 MRI images, 25 of which had the cortical surface in the right superior temporal gyrus artificially thinned. The method proved robust to normal population variations, and sensitive to even slight focal changes in cortical thinning.