Investigation of Modified Fuzzy C-Means Clustering with Content based Retrieval Image Technique for Medical Diagnosis

In medical images, Content-based image retrieval (CBIR) is a primary technique for computer-aided diagnosis. Many research works were developed in content based medical image retrieval, but the techniques have the drawback of low efficiency and high computation cost. To avoid such negative aspects a new enhanced Content Based Medical Image Retrieval (CBMIR) based on MFCM clustering technique is proposed in this paper. To retrieve the images, initially the features are extracted from the database medical images by using Haar. Thus the extracted features from the training database images are clustered by the FCM clustering technique. After in the retrieval process, the training image database feature values are differenced with the training images clustered result values. Based on the distance values, the most relevant images are retrieved. Hence, our proposed technique will efficiently retrieve the most relevant medial images via MFCM technique with retrieval rate.