Pattern Similarity of Medical Images for Texture Based Data Base Retrieval and Data Mining

A novel scheme for efficient content based medical image retrieval, formalized according to the PANDA (Patterns for Next generation Database systems) framework. The proposed scheme involves low-level feature extraction from image regions followed by clustering of the feature space to form higher-level patterns. The component of each pattern include a cluster representation and a measure of quality of the image content representation achieved by the pattern. The similarity between two patterns are estimated as a function of the similarity between both structure and the measured components of the patterns. Indexing of digital imagesand querying techniques have extensively been studied and few systems are dedicated to medical images today however the need for content based analysis and retrieval tools increases with the growth of digital medical image databases and data mining. Analyzing the medical image properties and evaluated the Gabor filter based features extraction for medical images indexing and classification. The goal is to perform clinically relevant queries on large image databases that do not require user supervision. After the demonstration on the concrete case of human imaging that these techniques can be used for indexing, retrieval by similarity queries, and 108 P.K. Kumaresan et al to some extent, extracting clinically relevant information out of the images.