Bone fracture detection is important for medical diagnosis because prompt and accurate identification is required for effective care and patient care. Conventional methods often have low user participation complex interfaces and poor accuracy. This study describes a technique for detecting bone fractures using Convolutional Neural Networks (CNNs) concentrating on the ResNet-50 architecture and YOLOv8 framework. This suggested method offers automated result processing and real-time prediction. The suggested system presents a viable way to expedite fracture diagnosis processes which could lessen the workload for medical personnel and enhance patient outcomes. We validate the effectiveness and practical utility of our proposed YOLOv8-based bone fracture detection system and recent CNN technology by achieving fracture recognition with high reliability using CNN models specifically ResNet50 through thorough evaluation and comparison with conventional approaches. Physicians now have a practical and accurate way to identify bone fractures thanks to the suggested method.