3D morphological analysis of brain MRI using wavelets

This article describes the use of the 2D wavelet transform to estimate the multiscale gradient of two types of cortical sulci, from which two potentially discriminating features, namely the gradient mean magnitude and entropy of its orientation distributions, are calculated. The obtained results corroborate the potential of such multiscale geometrical features for identification of narrow and wide cortical sulci, and important problem in medical imaging. The spatial scale leading to better discrimination is also identified.