A Novel Approach For MRI Image Classification And Image Retrieval

Abstract

OBJECTIVE: Classification of MRI image data based on its attribute characteristics is very important for storing the image data and its efficient retrieval. This project aims to find a solution to the problem of classification of the images that are stored on a hard drive and image retrieval from the image database with optimum automation in a user friendly way. METHOD: Matlab was used to create a software tool for image classification and, Java and SQL were used to create a software tool for image retrieval. RESULTS: Image classification tool could read multiple files, extract attribute values from the files and store them in a directory hierarchy automatically. Image retrieval tool could query the database and route the user to the actual image data that was stored on the hard drive. CONCLUSIONS: The image classification tool was efficient at classifying images in relatively short time as compared to other tools and with minimum labor. Image retrieval tool could successfully execute queries via a user friendly Graphical User Interface (GUI), without the user having to know the programming language of the database.