They developed an explainable artificial intelligence (AI) system for histology classification of vertebral compression fractures from digital pathology images that integrates histopathology, clinical data, and transcriptomic profiles. The system achieved an accuracy of 86–91% (F1 score 0.83–0.88) in distinguishing between osteoporotic, traumatic and neoplastic fractures. Interpretability was confirmed by Grad-CAM temperature maps and SHAP analysis, which highlighted biologically significant features such as trabecular thinning, nuclear atypia, and marrow fibrosis. Multi-omic risk scores showed that high-risk fractures have upregulated TNF–NF-κB signaling, reduced cytotoxic T-cell infiltration, and significantly worse 3-year survival (log-rank p < 0.001). Drug sensitivity modeling predicted response to bisphosphonates and RANKL inhibitors in low-risk patients, whereas high-risk cases showed resistant phenotypes. The system operates on modest computing resources and supports telepathology in resource-constrained environments.