A NOVEL HYBRID PCA-CNN BASED CBIR MODEL FOR MEDICAL IMAGE ANALYSIS

The rapid development of image processing in medical field is increased due to information technology and its impact over data analysis. Retrieval of medical images increased the importance in medical applications due to its challenging and unique characteristics of images. Content based image retrieval is one of the familiar options in medical image retrieval due to its popularity in effective image retrieval technique. CBIR uses a set of features to locate, retrieve and display the images are alike to the given query image. with advanced feature processing, it accesses the data in medical archives and medical equipment to infer the necessary output. Evidence based diagnosis, text-based retrieval are some of the advantages of CBIR, which could be applicable in various fields like data administration, healthcare. This research work proposed a hybrid CBIR model to improve the retrieval performance, retrieval accuracy of the medical image retrieval process through principal component analysis and convolutional neural network.