Potentialities and challenges of using artificial intelligence in chronic pain management: a scoping review

This study aims to map the scientific literature on using artificial intelligence (AI) in managing chronic pain, the leading cause of disability worldwide. A scoping review was conducted following PRISMA-ScR guidelines. We searched for articles in MEDLINE, LILACS, SPORTDiscus, SCOPUS, CENTRAL, Science Direct, IEEE Xplore, Association for Computing Machinery Digital Library, and arXiv databases. Descriptors "Chronic Pain" and its synonyms were combined with "Artificial Intelligence" or "Machine Learning" or their synonyms. Initially, 2,767 articles were identified, and 109 studies were selected and analyzed. The volume of scientific production has been increasing over the years. Most adopted AI algorithms focused on classifying individuals by type of chronic pain and predicting risk groups. Support Vector Machine, Random Forest, Neural Networks, Logistic Regression, and K-means were the most adopted algorithms. Few studies focused on AI use in primary health care. We underscore AI's potential to prevent and improve chronic pain management.