Detection of Fractures in Orthopedic X-Ray Images

Abstract

Computer detection of fractures in orthopedic X-ray images is very difficult and stimulating process. Bones can crack in multiple ways having a variety of degrees of severity. The Severe fractures effect drastically to the shape of the longest bones, mild fractures do not affect gammon’s shape and only elusive indications are observed in orthopedic X-ray images. A Large number of image processing procedures have been tested with orthopedic X-ray images having several types of fractures. The proposed work is to designate the fractures in orthopedic X-Ray images using edge detection, intensity based segmentation and frequency transformation based methods. An analysis has been performed using proposed algorithm on various orthopedic X-ray images depending upon the image characteristics and fracture type. The proposed work discovered an image processing algorithm that first performs orthopedic X-ray image enhancement, then performs the segmentation of region of interest of bone along with edge detection and feature extraction of broken area of bones. Then by applying morphological filtering, determines the expected region of fracture in test orthopedic X-Ray image. The analysis reveals improvement in timeliness and accuracy in clinical diagnosis. Results outshined both quantitatively as well as qualitatively when compared with other existing techniques. Also the clinical validation of Results has been performed.