Identification And Classification of Bone Deformity Using Machine Learning

Identification And Classification of Bone Deformity Using Machine Learning

Shubhangi Mane, Prof. Pramod Patil, Rutuja Shinde, Hansika Nemade, Vaishnavi Chaudhari

Abstract -- Automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have work on a technique for the detection and evaluation of orthopaedic fractures from X-ray images. We have also developed a software that can be ordinarily used by paramedics or bone specialist doctors. The proposed tool first segments the bone part of an input digital X-ray image from its surrounding region and then generates the outline of bone. After that, it performs correction of discontinuity or breaks of outline of bone that might have been generated because of segmentation process, and then finally it detects the existence of fracture. In X-ray image, the bone is visible with the surrounding tissues or muscles (i.e. flesh). therefore, segmentation process is required to generate the outline of bone. Extraction of bone part from muscles and tissue shadow is a tedious task in digital X-ray image segmentation. Each X-ray image has three distinct regions - a black background region, a flesh region, and bone region.

Keywords—Bone fracture, Image processing, Edge detection, Classification,

Convolutional neural network, Machine learning, Python, spyder, Anaconda