Comprehensive Summary
Low back pain (LBP) is one of the leading causes of disability worldwide, significantly impacting quality of life and placing a heavy socioeconomic burden on healthcare systems. This systematic review examines how artificial intelligence (AI)-based Decision Support Systems (DSS) are being utilized to assist in the prevention and management of lumbar degenerative spine disorders, a major contributor to chronic LBP. The study involved a systematic review of PubMed and Scopus databases, identifying 25 relevant studies that applied machine learning and deep learning models to clinical, demographic, imaging, and psychosocial data. The AI-based models were primarily used for three key tasks: (1) defining clinical scores for predicting disease progression and treatment suitability, (2) assessing patient conditions through risk factor identification and diagnostic prediction, and (3) determining eligibility for surgical or conservative treatment. The reported accuracy was high, with AI models achieving AUC scores between 0.93 and 0.99 for different prediction tasks. Some models also combined radiological features, biomarker analysis, and patient-reported outcomes to provide a more comprehensive decision-making framework.
Outcomes and Implications
This review's findings highlight AI's growing role in spine care, particularly in improving diagnostic precision and optimizing treatment strategies. By integrating AI into clinical decision-making, healthcare providers can reduce variability in diagnoses, offer personalized treatment recommendations, and potentially minimize unnecessary surgeries. AI-based DSS models have demonstrated strong potential in predicting surgical success, identifying high-risk patients, and improving guideline compliance, which could lead to better long-term outcomes for individuals suffering from degenerative lumbar spine disorders. Although AI is proving to be a powerful tool in spine surgery and rehabilitation, challenges remain, particularly in data integration, standardization, and real-world implementation. Future research should focus on multicenter validation, real-time AI integration in clinical settings, and regulatory compliance to ensure these AI-driven solutions can be widely adopted in everyday medical practice. If these hurdles are addressed, AI-based DSS could become a cornerstone of modern spinal care, making treatment more efficient, evidence-based, and tailored to individual patient needs.