Development of Multimodal AI Predictive Models for Degenerative Disc Disease Progression and Optimization of Regenerative Therapeutic Approaches

Degenerative Disc Disease (DDD) is a major contributor to chronic low back pain and disability worldwide. The unpredictable nature of its progression, along with variable patient responses to regenerative therapies such as platelet-rich plasma (PRP) and mesenchymal stem cells (MSCs), poses significant clinical challenges. This study proposes a novel approach using multimodal Artificial Intelligence (AI) models to predict disease progression and optimize therapeutic strategies. By integrating heterogeneous data sources including MRI imaging, clinical variables, and molecular biomarkers we developed advanced predictive models employing XGBoost and multimodal deep learning architectures. Feature fusion and attention mechanisms were used to enhance the interpretability and robustness of the predictions. Our results demonstrate that the proposed model achieved high predictive performance, with an AUC of ۰.۹۱ and overall accuracy of ۹۱% in forecasting DDD progression. Moreover, the AI framework was able to suggest optimal treatment timing and modality based on individual patient profiles, improving the potential for personalized intervention planning. These findings suggest that multimodal AI models can serve as powerful decision-support tools, enabling precision diagnostics and personalized regenerative therapy in spine care. The study underscores the clinical utility of combining advanced machine learning with comprehensive biomedical data to transform the management of degenerative spine disorders.

کلیدواژه ها:

Degenerative Disc Disease (DDD) ، Multimodal Artificial Intelligence ، Predictive Modeling ، Regenerative Therapy Optimization ، Personalized Medicine