USE OF ARTIFICIAL INTELLIGENCE IN INTRAOPERATIVE HYPOTENSION PREDICTION

Abstract

Intraoperative hypotension (IOH) is a frequent hemodynamic complication, affecting approximately 29% of surgical patients, and is associated with severe adverse outcomes, such as acute kidney injury, perioperative myocardial injury, and increased mortality. With advancements in artificial intelligence (AI) techniques, machine learning and deep learning algorithms have been developed to predict these events prior to their clinical onset. The objective of this study was to evaluate the efficacy and safety of artificial intelligence algorithms in predicting intraoperative hypotension. A narrative literature review was conducted using the PubMed, Scopus, and Web of Science databases, and a total of 10 articles were selected after applying the inclusion and exclusion criteria. The studies primarily utilized the Hypotension Prediction Index (HPI), based on arterial waveform analysis, as well as deep learning models with CNN-RNN architectures. The results demonstrated a sensitivity of 0.75, a specificity of 0.73, and a pooled AUROC of 0.83 for machine learning algorithms in general, with deep learning models achieving an AUROC of 0.94. However, controversies remain regarding their impact on hard clinical outcomes, such as acute kidney injury and myocardial injury, which have not yet been consistently demonstrated. Furthermore, a substantial portion of the literature is limited by short-term evaluations and small sample sizes. In this regard, further research in this field is imperative, as the early prediction of IOH represents a fundamental advancement in surgical patient safety.