A Chain Code Based Fuzzy Features for Brain MRI Retrieval

In the content based image retrieval is an important to retrieve all closely matching images. Human brain MR Image retrieval for similar MR Images is an important application in Radiology field of medical diagnostics. Morphological segmentation is a promising method for highlighting and extraction of the region based features of human brain T₂ - weighted MR Images. Fuzzy representation of these features and its use in retrieval of brain MRI is demonstrated in our previous work. Unified Fuzzy Matching based fuzzy feature matching is used to assess the image-to-image similarity of MRI in it. The pattern recognition and the identification of similarity of Brain MRI is a complex problem. In addition to the general region based features, additional chain code based features are proposed in this paper. Use of integration of region based features and the chain code based features is demonstrated in this work. This application of Brain MRI retrieval is found to give better retrieval results in terms of precision and average rank.