Artificial intelligence in perioperative pain management: A review

Abstract

Artificial intelligence (AI) leverages its swift, precise, and fatigue-resistant problem-solving abilities to significantly influence anesthetic practices, ranging from monitoring the depth of anesthesia to controlling its delivery and predicting events. Within the domain of anesthesia, pain management plays a pivotal role. This review examines the promises and challenges of integrating AI into perioperative pain management, offering an in-depth analysis of their converging interfaces. Given the breadth of research in perioperative pain management, the review centers on the quality of training datasets, the integrity of experimental outcomes, and the diversity of algorithmic approaches. We conducted a thorough examination of studies from electronic databases, grouping them into three core themes: pain assessment, therapeutic interventions, and the forecasting of pain management-related adverse effects. Subsequently, we addressed the limitations of AI application, such as the need for enhanced predictive accuracy, privacy concerns, and the development of a robust database. Building upon these considerations, we propose avenues for future research that harness the potential of AI to effectively contribute to perioperative pain management, aiming to refine the clinical utility of this technology.