Advanced Hybrid Model for Hand Fracture Detection: Combining XGBoost and LightGBM

Fractures of the hand are frequent injuries that need prompt and precise diagnosis to guarantee the right treatment. Thanks to developments in deep learning, radiologists now have a practical tool for automatically detecting fractures in X-ray pictures. The authors of this paper suggest a hybrid method for X-ray bone fracture detection in hands by using a XGBoost and LightGBM model. XGBoost and LightGBM automates the building of an optimum neural network architecture for this job by combining the versatility of LightGBM with the efficient object identification capabilities of the XGBoost architecture. To start, the X-ray pictures are preprocessed using methods to make them clearer, such as noise reduction and picture enhancement. Afterwards, the hybrid XGBoost and LightGBM model is used to identify and pinpoint hand bone fractures. By combining global and local information, the model is able to accurately identify complex fractures, even those that are tiny and hard to see. In addition, the LightGBM component optimizes the detection accuracy while lowering processing costs by tailoring the design to the unique dataset. The hybrid XGBoost and LightGBM model performs better than conventional CNN-based approaches in terms of accuracy, inference speed, and resilience when tested on a variety of fracture types. This method has potential for improving orthopedic diagnostic clinical decision support systems.