Retrieval of Brain Image Using Soft Computing Technique

CBIR in medical image databases are used to assist physician in diagnosis the diseases and also used to aid diagnosis by identifying similar past cases. Retrieval of brain in fast, accurate and an effective from the large data set needs pre-processing steps as novel approach. This approach contains two methods. One method is removal of artifacts in brain image and another one is extraction of fine brain portion using Fuzzy Expectation Maximization Algorithm (FEMA). For the feature extraction use shape feature like 2D-Zernike moment and finally for retrieval of brain image this paper proposes soft computing technique of Extreme Learning Machine (ELM) with four distance metric measures viz., Euclidean Distance, Quasi Euclidean, City Block distance, Hamming Distance. For the experimental results 1500 MRI images are taken from scan centre and some brain images are collected from web and these have been implemented with these methods. The experiments which show that the retrieval of brain image using ELM with hamming distance advice promising results are reported.