New learning approach for computer-aided diagnostic

Recently, computer-aided diagnosis is beginning to be applied widely in the detection and differential diagnosis of many different types of abnormalities in medical images obtained in various examinations by use of different imaging modalities. Content-based image retrieval (CBIR) is a promising method for computer-aided diagnostics leading early diagnosis. In this paper, we perform FOCT (Facial and Oral Cancer Tracker). Our new platform can automatically classify images based on their content; give them text annotation and assist surgeons in decisions regarding new cases by supplying visually similar past cases. This tool may guide diagnostic, treatment, management and monitoring oral cancer through comparison of long-term outcomes in similar cases. Our application is based on a web interface, able to classify suspicious regions. This paper presents a novel feature selection techniques based on a hierarchical model, that can find what features best represent a given set of images. In order to improve the retrieval performance, a machine learning approach based on support vector machines (SVM) and relevance feedback strategies are investigated in this paper.