Deep Learning-Based Wrist Fracture Localization and Segmentation in Bone X-rays

An incomplete or full break in the bone is known as a fracture. The primary cause of fracture is high impact or force applied to a bone that is structurally capable of withstanding. Traumatic and stress are commonly found bone fractures in the human body. Stress fractures are common among sports (such as gymnasts, dancers, and long-distance runners) and military people and are caused by repetitive load-carrying strain on a healthy bone. Traumatic fractures are caused by vehicle accidents, serious falls, or purposeful causes such as physical abuse. A fracture can also happen for several other reasons, such as osteoporosis (a disease that weakens bones), cancer, or the brittle bone condition known as ontogenesis imperfect. According to the World Health Organization (WHO) report, 1.66 million people suffer from hip fractures every year throughout the world, and the rate is expected to rise by three to four times by the year 2050 because of the worldwide increase in the number of older people.

X-rays, Magnetic Resonance Imaging (MRI), and Computed Tomography (CT) are various medical imaging modalities used to capture images of the affected body area. A radiology specialist interprets these recorded images to make a medical diagnosis and then recommends therapy. The oldest, quickest, and most popular imaging technique is an X-ray, which analyses potential fractures by taking pictures of the body's interior organs. It has emerged as the go-to analytical tool for examining patients for fractures because it is widely accessible in locations where other, more expensive imaging modalities would not be. Radiologists or physicians use visual inspection to evaluate X-ray samples to determine the existence and type of fractures in different bones. To acquire a more detailed, cross-sectional image of the bone that may be missed during an X-ray examination, the need for advanced imaging technologies such as MRI and CT scans emerges