A Robust Approach For Detection of the type of Fracture from X-Ray Images

Abstract: A bone is broken or a fracture in bone happens when an external force applied upon the bone is more than what the bone can tolerate or bear. As a result, the shape and strength of the bone is disturbed which causes excruciating pain on the bone and ends up in the loss of functioning of the bone. In some instances there will be bleeding around the injured site. The modern developments in medical imaging contributes a lot in examining the fractures in different kinds and classes of bones found in the body of the patient without much difficulty. Segmentation of an image involves a process called edge detection. The purpose of segmentation is to modify the representation of an image into fragments that is bits and pieces that are more important, useful and easier to analyse. To identify the edges, this function scans for regions in the image where the change of intensity occurs rapidly. Edge detection returns a binary image containing ones where edges are found and zeros in all other places. Different edge detection methods exist like Sobel operator, Prewitt operator, Laplacian of a Gaussian and Canny. These techniques can be applied on X-Ray medical images. Quality metrics like mean and standard deviation are applied to analyze, compare and evaluate the results. The process of detecting the type of fracture has to be done with a great precision. To improve these quality metrics,a novel edge detection algorithm is proposed and the accuracy is measured for this cause.

Keywords: Fracture detection, Bone segmentation, Edge detection, Fracture classification.