Hybrid Deep Learning Framework for Femur Bone Osteoporosis Classification: Leveraging Tuna Jellyfish Optimization and Segmentation Techniques

Osteoporosis weakens bones, making them more fragile and prone to fractures. This condition can result from changes in bone strength or a decline in bone mineral density and mass. Accurate classification of osteoporosis is crucial for diagnosing skeletal disorders, particularly in aging individuals. This study introduces a novel approach to classify femur bones from X-ray images using a hybrid segmentation method and an enhanced Tuna Jelly Fish Optimizer (TJFO) with Deep Convolutional Neural Networks (Deep CNN). The hybrid method integrates traditional image processing and deep learning for precise femur isolation, while TJFO fine-tunes the Deep CNN’s hyperparameters for improved classification. The model’s performance was evaluated using metrics like Accuracy, True Positive Rate (TPR), True Negative Rate (TNR), Positive Predictive Value (PPV), and Negative Predictive Value (NPV), showing significant improvements in femur bone classification with balanced results across all metrics.