An Improved Content Based Medical Image Retrieval System Using Integrated Steerable Texture Components and User Interactive Feedback Method
B. Jyothi, Y. Madhavee Latha, P. G. Krishna Mohan · Advances in intelligent systems and computing · 2016
The advancement in medical technology has resulted in a huge number of medical images saved in a data-base. Content Based Medical Image Retrieval (CBMIR) mechanisms help the radiologist in retrieving the required medical images from an immense database. This paper envisages an effective content based procedure in which the region of the image is taken into account by determining the borders of the image region using gray level gradient method instead of considering the image as a whole. Later, the content within the boundary region of the image is described through the steerable filter in different orientations followed by extracting the second-order statistical components as feature vectors. Medical images correlated to the query image are retrieved by computing the Euclidean distance as a similarity measure between database images and the query image. To enhance the accuracy of the medical retrieval system, Instant Based Relevance Feedback has been used. In this procedure, the user interacts with the system and selects the most relevant image for searching again. The above search procedure is repeated for finding out more precise images by sorting out the first search and the second search similarity distances. Eventually, the corresponding top ranked images are displayed. These results reveal that the proposed algorithm outperforms by of increasing Recall Rate and reducing Rate of Error.