Abstract

In this paper we proposed, a novel framework to assist and automate the diagnosis of diseases from computer-based image analysis method using Content-based image retrieval (CBIR). CBIR is the process of retrieving related images from large database collections by using low level image features such as color, texture and shape etc. we have used fuzzy texton and discrete shearlet transform to extract texture and shape features. The aim is to support decision making by retrieving and displaying relevant past cases visually similar to the one under examination with relevance feedback using Support Vector Machines.