AUTOMATED OSTEOPOROSIS IDENTIFICATION IN BONE X-RAY IMAGES USING DEEP FEATURE LEARNING

Abstract

Osteoporosis is a progressive skeletal disorder characterized by reduced bone mineral density (BMD) and deterioration of bone microarchitecture, leading to an increased risk of fractures. Early diagnosis is essential for minimizing complications and improving patient outcomes. Conventional diagnostic methods such as Dual-Energy X-ray Absorptiometry (DEXA) are accurate but expensive and not always accessible in resource-constrained settings. This research proposes a deep learning framework for automated osteoporosis identification from bone X-ray images using a Multi-View CT Network (MVCTNet). Although trained primarily on X-ray images, the proposed architecture integrates multi-view feature learning inspired by CT imaging principles to capture complementary structural information from different anatomical perspectives. The framework utilizes transfer learning, multi-scale feature extraction, attention mechanisms, and feature fusion to improve classification performance. Extensive experiments demonstrate that the proposed model achieves superior classification accuracy compared with conventional CNN architectures including ResNet50, DenseNet121, EfficientNetB0, and MobileNetV2. The proposed approach obtained an accuracy of 98.64%, precision of 98.58%, recall of 98.72%, F1-score of 98.65%, and AUC of 99.41%, demonstrating its effectiveness for early osteoporosis screening.