Edge-Preserving Segmentation and Fusion of Medical Images by using Enhanced Mean Shift

Konferensbidrag (offentliggjort, men ej förlagsutgivet), 2008

This paper addresses the issue of medical image segmentation by using an enhanced spatial-range mean shift. Mean shift is a method for estimating local modes (maxima) of pdf (probability density function) using a kernel-based approach. This paper describes an enhanced spatial-range mean shift segmentation method for biomedical (MRI) image segmentation. Preliminary work and the results on fusion of segmented brain images from different sensors (e.g. MRI, CT) are presented and discussed.