Comparative analysis of YOLO variants and EfficientNet for detecting bone fractures in X-ray images

This research examines the use of deep learning models for the automatic identification of bone fractures in X-ray images, focusing on the necessity for timely and precise diagnoses in the field of medicine. With the aid of comprehensive datasets that cover seventeen different types of fractures, the author addresses the issues of class imbalance and the identification of subtle fractures through extensive initial processing (data augmentation, image resizing). The author performs a comparative analysis of seven advanced deep learning models: YOLOv8, YOLOv9, YOLOv10, YOLOv11, EfficientNetB0, DenseNet169, and ResNet50. In this study, the author evaluates the performance of the models according to traditional measurement criteria: precision, recall, F1-score, and mean average

precision (mAP). The author concludes that YOLOv11 surpasses the other models in terms of precision, mAP, and the balance of precision and recall. Such performance is due to design improvements, such as having a deeper backbone and hybrid feature fusion, which support its versatility in different bone fracture detection situations and its consistent use of memory for images of varying quality and resolutions.

Message: Hi FE, I need help with “Comparative analysis of YOLO variants and EfficientNet for detecting bone fractures in X-ray images” in “Electronics & Communication Engineering”