The analysis of whole-body bone scan images by using quantitative improvement in brightness

Abstract

Today's, the image processing applications are obvious in the various sciences. The medical image processing has achieved to acceptable results in the field of diagnosing disease. One of these applications which is discussed in this paper is diagnosis injury in whole body bone scan. Whole body bone scan has some applications in the diagnosis of benign lesions such as cardiovascular disease and metabolic bone and sports injuries and also to investigate the pervasiveness of bone. One of the main criteria to diagnose complications in the body is area with high brightness ‎to the brightness of adjacent areas in the image or asymmetry of the brightness in the two-sided areas of the body. In this study, a software has been implemented about the detection of disease-suspected areas that strategies to increase the pixel intensity have been considered to make easier the image segmentation and innovative methods of researcher have been implemented in image preparation to convert to binary mode and find suspected areas. In addition, the software automatically creates reports, this feature will be a big contribution to the doctor. Software assessment has been done on 11 whole body bone scan images and provided the results to the physician, the software succeeded to find the 41 correct areas from 44 injured areas, 93.18% of cases have been successfully identified. In terms of the number of errors in the areas of diagnosis, 30 cases were observed in 11 images which shows an average of 2.72 errors on each image.