A New Approach for Content Based Medical Image Retrieval Based on Wavelet Lifting and Fuzzy C-Means Clustering

This paper provides a novel scheme for efficient Content-Based Image Retrieval. With the vast amount of multimedia data generated and transmitted in digital formats, content-based data management has emerged as an important area. Content-Based Image Retrieval (CBIR) systems, is used to retrieve images based on their similarity with one or more query images. This is one of the primary tools used by physicians for the comparison of previous and current medical images associated with pathologic conditions. The novel approach used is formalized according to the PAtterns for Next generation DAtabase systems (PANDA) framework for pattern representation and management. This scheme involves block-based low-level feature extraction from images followed by the clustering of the feature space to form meaningful patterns. The Wavelet Lifting Technique is used to extract the features from the medical images. An Fuzzy C Means algorithm is used to cluster the feature space. It is an iterative approach to automatically determine the number of clusters. The clusters resulting from the Fuzzy C Means algorithm are considered as patterns extracted from the image database, and are represented and handled according to the PANDA formalization. The similarity between two clusters is estimated as a function of the similarity of both their structures and the measure components. A large set of reference radiographic images are used as the experimental datasets.