BoneScanAI: Advanced Machine Learning for Precision Bone Fracture Diagnosis

The creation and use of a hybrid machine learning model integrating convolutional neural networks (CNN) and Random Forest classifiers to raise the accuracy of bone fracture prediction from X-ray pictures is investigated in this research article. Multiple layers of CNNs for deep feature extraction follow by a Random Forest to classify into seven different fracture kinds forms the model architecture. The 2,738 X-ray pictures in the dataset are carefully labelled by qualified radiologists to guarantee consistent training and testing. With strict data preparation methods to improve picture quality and relevance for efficient learning, the model was trained utilising a split of $\mathbf{7 0 \%}$ for training, $\mathbf{1 5 \%}$ for validation, and 15% for testing. Precision, recall, and F1-score measures were used to assess the model’s performance; total accuracy of $\mathbf{8 6. 9 9 \%}$ resulted. This great degree of accuracy shows how well the model supports doctors in precisely identifying many bone fractures. Considering the study’s implications, especially in enhancing the diagnostication processes in radiology, the conclusions suggest that employing CNN and Random Forest as a good method in medical image processing. For future work, more extensive datasets will be accumulated to enhance the model’s effectiveness and applicability in various clinical contexts, besides model refinement by further training.