Abstract:

Bone fractures are one of the most common problems in humans because of accidents or other causes. By using the x-ray, MRI is a manual detection of bone, and ignoring the fracture may cause severe consequences to the patient that may risk their life. Automated fracture detection is an essential part of a computer-aided telemedicine system that reduces the patient's risk. It is useful to medical clinicians who lack subspecialized expertise in orthopedics, and misdiagnosed fractures account for upward of four of every five reported diagnostic errors in certain EDs. We found that the usage of CNN with edge detectors, which causes misclassification, was the main flaw in the base articles. The position and orientation of objects are not encoded by CNN. Hence, we use the deep learning model Faster RCNN to precisely detect fractures. The model detects modest bone fractures that are difficult to notice with the naked eye in x-ray pictures. In comparison to conventional approaches, the faster RCNN model is more accurate. The model in this paper uses efficient fracture location prediction and fracture accuracy visualization. The fracture location is high spotted with the bounding box. The model has a 90% accuracy rate

Description:

CNN, Faster RCNN, Edge detectors