AI-Enhanced Monitoring and Early Detection of Intraoperative Events During Cardiac Surgery

Abstract

Background: Artificial intelligence (AI) is revolutionizing healthcare, particularly in cardiac surgery, where intraoperative monitoring plays a vital role in ensuring patient safety. Traditional monitoring systems depend on clinicians’ manual interpretation of physiological signals, which may be limited by human fatigue, delayed responses, and data complexity.

Aim and Objectives: This review aims to explore the role of AI-driven technologies in enhancing intraoperative monitoring during cardiac surgery. The objectives include evaluating AI applications for early event detection, analyzing algorithmic performance in predicting critical intraoperative conditions, and identifying challenges and future research directions.

Methodology: A comprehensive literature review was conducted using recent studies (2019-2025) focusing on machine learning (ML) and deep learning (DL) models applied to intraoperative cardiac monitoring. Databases such as PubMed, IEEE Xplore, and ScienceDirect were searched for studies related to hypotension prediction, arrhythmia detection, and myocardial ischemia monitoring.

Results: AI systems, particularly those utilizing multimodal physiological data such as electrocardiography (ECG), arterial pressure, and oxygen saturation demonstrated high accuracy in detecting and predicting adverse intraoperative events. Algorithms like the Hypotension Prediction Index (HPI) and deep neural networks for ECG analysis showed improved sensitivity and timeliness compared to conventional methods.

Conclusion: AI-enhanced intraoperative monitoring holds transformative potential for precision cardiac surgery. Despite challenges in model interpretability, generalizability, and clinical integration, future innovations in explainable AI, federated learning, and real-time decision support promise safer surgical outcomes.

KEYWORDS: Artificial intelligence, cardiac surgery, intraoperative monitoring, machine learning, hypotension prediction, arrhythmia detection, patient safety, deep learning.