Deep Belief CNN Feature Representation Based Content Based Image Retrieval for Medical Images.

Avascular Necrosis (AN) is a cause of muscular-skeletal disability. As it is common amongst the younger people, early intervention and prompt diagnosis is requisite. This disease normally affects the femoral bones, in order that the bones' shape gets altered due to the fracture. Other common sites encompass knees, humerus, shoulders, jaw, and ankles. The retrieval of the AN affected bone images is challenging due to its varied fracture locations. This work proposes an effectual methodology for retrieval of AN images utilizing Deep Belief CNN Feature Representation. Initially, the input dataset undergoes preprocessing. The image noise is eradicated utilizing Median Filter (MF) and is resized in the preprocessing stage. Features are represented using Deep Belief Convolutional Neural Network (DB-CNN). Now, the image feature representations are transmuted to binary codes. Then, the similarity measurement is computed utilizing Modified-Hamming Distance. Finally, the images are retrieved centered on the similarity values. The test outcomes evinced that the proposed work is better than the other existent techniques.