Description

We propose a new technique for subject identification and semantic classification of MR brain images in a multimodal environment. The proposed technique not only finds and retrieves similar slices but it also associates the query slices with a particular patient. Moreover, it identifies the 3-D volume of the patient based on the query slice. We propose to make use of multiscale wavelet transform to extract features from MR slices. These features are fed to SVM (Support Vector Machines) classifier to predict the status of query data. We feel that given a query slice, identifying the specific subject in a multimodal environment and associating it with a specific semantic area would be extremely useful. For example, this technique will help human experts in quantifying, localizing, and tracking of disease progression among various subjects and within a specific semantic area of the brain. Our objective is to retrieve similar images from the dataset of classified images. To improve the prediction and classifier accuracy and also to limit the time consumption for the improved retrieval of similar images.