Cardio-Ai: Compliance and Artificial Intelligence Regulation for Deep Learning in Electrocardiogram and Cardiac Signal Interpretation Publisher



Rajabi MZ ; Fazilati F ; Alihosseini N ; Esmaeili Farsani M ; Biriaei F ; Alirezaei Farahani M ; Khorasgani ANN ; Sadeghipour F ; Marateb HR ; Mananas M
Source: Deep Learning for Cardiac Signal Analysis in Robotic Applications Published:2026

Abstract

In this chapter, we provide a comprehensive overview of the regulatory, technical, and evaluative aspects of applying artificial intelligence (AI) and deep learning to cardiac signal analysis, with a specific focus on electrocardiogram (ECG) data. We explore general and ECG-specific regulations, including standards for arrhythmia detection and cardio-signal processing, as well as key considerations in dataset validation, governance, preprocessing, and augmentation. Evaluation techniques are addressed in terms of performance measures, validation techniques, and robustness, along with specific indicators for arrhythmia and cardiac monitoring, such as heart rate variability and the QRS complex. The security and confidentiality of ECG telemetry systems are also addressed, including secure storage, transmission, and real-time inspection. Pre- and postmarket regulatory frameworks are reviewed, including lifecycle management, industrial compliance, and continuous model updating. Finally, we highlight future directions, including explainable AI (XAI), transfer learning (TL), and the use of large language models (LLMs) and multimodal learning. These approaches can be applied to enhance cardiac diagnostics by leveraging XAI, TL, and LLMs s, as well as multimodal learning. The CARDIO-AI-CHECK checklist is publicly available online at https://dx.doi.org/10.6084/m9.figshare.30022759. © 2026 Elsevier Inc. All rights reserved.