Investigation of content based information retrieval Algorithms for medical images and applications using Multidimensional features

Abstract

The research of the thesis is concluded that an efficient two methods are applied in CBIR technique to the problem of cirrhosis diagnosis in general have achieved successful detection of newlineaffected disease part using the EMD method of distance measure between the nuclei and the classification of cells using the ARKFCM method. An OR-ACM approach is used for segmenting the nuclei and non-nuclei cells from the liver histopathology image. After segmenting, the nuclei cells are characterized into two levels such as macro nuclei and micro nuclei cells, which are evaluated based on semantic feature extraction (combination of both low and high level features) and classified by applying multi-class SVM classifier. The GLCM with haralick features are extract the features from input image. After that using SVM classifier classify the image as benign, malignant and noncancerous. newline