Artificial intelligence on interventional cardiology

This chapter discusses the use of AI in interventional cardiology. The chapter outlines how machine learning is focused on building automated clinical decision systems that help physicians make more accurate predictions, rather than those obtained through simplified estimated scoring systems. Studies illustrating the promise of AI techniques, such as Deep Learning (DL) and computer vision, for image quantification in cardiology are discussed and recent developments in the use of Machine Learning (ML) for physiologic measurements are also summarized. Studies showing the use of ML in intravascular ultrasound (IVUS) imaging technology are summarized, with examples describing the accurate diagnosis of angiographic lesions, plaque characterization, microvascular dysfunction, and post-procedural stent area and the performance of IVUS-based ML algorithms compared with that of human experts. The chapter discusses how AI-assisted precision PCI is an extremely promising area within interventional cardiology in order to optimize lesion preparation, obtain complete revascularization and, achieve a durable long-term result, and examples of how AI can aid in this area are given. The use of AI models for risk stratification and outcome prediction in cardiology is described, and a selection of studies in this area is summarized. The use of DL in robotic-assisted procedures is also touched upon briefly. The use of Natural Language Processing (NLP) to predict post PCI outcomes is described and the possible utilization of augmented reality and AI to assist in medical education and catheterization laboratory training, such as procedural training, procedural assistance, and patient or physician education is also outlined. The potential for AI to detect and identify various challenging cardiological conditions is also discussed. The chapter ends by discussing the challenges for the use of AI-assisted decision-making, and providing a brief overview of the future of AI-integrated interventional cardiology.