Fracture Detection in X-ray Images Using Object Detection Algorithm

Abstract

Bone fractures are a common medical issue and the patient suffer a lot and it is so painful . The majority of the time as a result of pressure put on it from an accident, practicing sports, or other causes .Many prominent hospitals in Bangladesh invest a lot of money each year on X-ray machines, radiologists, and medical personnel to enhance patient care. After taking x-ray of the fracture the doctor and radiologist detect the fracture manually. For that reason many many doctors and radiologist miss small fracture or hair line fracture and give patient wrong treatment and for this wrong treatment patient suffer a long time . Many modern scanners can create digital radiography picture excellent detailing and resolution. But we cannot thinking to improve the current way of diagnosis. we may use the huge amount of data with high quality and resolution to apply some computer vision technique to speed up the detection process . My object is to analyze medical x-ray images using object detection algorithm, and I use YOLO V:7 for object detection .Yolo v7 is faster object detection algorithm because YOLO V7 use end to end Neural Network it makes predictions of bounding boxes and class probabilities all at once. YOLO uses a single fully connected layer to carry out all of its predictions. it is the fastest and most accurate real-time object detector to date.