Enhanced Bone Fracture Prediction Using Transfer Learning and YOLOv8 Deep Learning Models
Abstract
Bone is a rigid organ that provides strong structure to protect organs and enables movement of blood cells. Bones are made up of spongy and compact tissue and it provides continuous remodeling. Several accidents are happening every day and many cases show patients affected by bone breakage. Bones are fractured when a person gets into an accident. Many surgeries are done by mistake due to miss classification of brain fracture. Artificial intelligence techniques such deep learning models introduced for human bone fracture prediction. Proposed research uses YOLOv8, CNN and ResNet50 architectures for bone fracture prediction. YOLOv8 is an efficient object detection model that produces highest accuracy 96% when compared to ResNet50 transfer learning model and CNN deep learning model. Obtained accuracy results from proposed research enhanced when compared to existing research models such as machine learning algorithms. This research is more useful for bone fracture patients to take care of further treatment.