A convolutional network based on YOLOv8 as an AI-assisted CT diagnosis technology for bone fracture

The utilization of an artificial intelligence (AI) algorithm based on deep learning is a valuable tool in the field of diagnostic imaging. In the past, medical images were diagnosed by medical professionals, a process that was timeconsuming and yielded limited diagnostic information. This paper proposes an automatic detection algorithm based on the YOLOv8 deep learning method to study CT images of trauma patients' arms, and its clinical effect is evaluated. The effectiveness of the proposed method was evaluated using Kaggle's fracture data, with the results confirmed by a radiologist with an accuracy of 91%. The algorithm demonstrated sufficient sensitivity to rapidly detect human upper arm fractures in a validated set of evaluated CT images.