Edge-XAI: An Offline-First Real-Time Explainable AI Framework for 12-Lead ECG Diagnosis in Resource-Constrained Healthcare Environments
Internet of Things · 2026-11 · 102155
DOI: 10.1016/j.iot.2026.102155 ↗Kurz gefasst
Fachartikel in der Zeitschrift Internet of Things, 2026-11. Verfasst von Sirapop Nuannimnoi und 6 weiteren Autor:innen. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Cardiovascular diseases, notably acute myocardial infarction (AMI), continue to be a principal cause of global morbidity and mortality. Prompt and accurate AMI diagnosis is vital for reducing its impact; however, resource-constrained settings—such as rural areas and ambulance services—present significant challenges for implementing effective diagnostic technologies. To address these challenges, we introduce an offline-first, real-time Edge Explainable AI (Edge-XAI) system tailored for 12-lead electrocardiogram (ECG) analysis in rural healthcare environments. Our system is specifically optimized for settings with limited computational resources and intermittent internet connectivity, enabling immediate diagnostic decisions. Additionally, we propose a novel explainable AI framework that facilitates real-time Shapley value estimation, thereby enhancing the interpretability of the diagnostic model. This integrated approach not only advances the deployment of lightweight and accurate ECG classification in resource-limited scenarios but also promotes transparency in AI-driven decision-making.
Zitieren
APA
Nuannimnoi, S., Lee, Y., Saharani, S., Baghban, H., Baskoro, A., Hu, Y., & Huang, C. (2026). Edge-XAI: An Offline-First Real-Time Explainable AI Framework for 12-Lead ECG Diagnosis in Resource-Constrained Healthcare Environments. Internet of Things, 102155. https://doi.org/10.1016/j.iot.2026.102155
Vancouver
Nuannimnoi S, Lee Y, Saharani S, Baghban H, Baskoro A, Hu Y, et al. Edge-XAI: An Offline-First Real-Time Explainable AI Framework for 12-Lead ECG Diagnosis in Resource-Constrained Healthcare Environments. Internet of Things. 2026:102155. doi:10.1016/j.iot.2026.102155
BibTeX
@article{Nuannimnoi2026,
author = {Nuannimnoi, Sirapop and Lee, Yun-Wei and Saharani, Sofya and Baghban, Hojjat and Baskoro, Angger and Hu, Yu-Yen and Huang, Ching-Yao},
title = {Edge-XAI: An Offline-First Real-Time Explainable AI Framework for 12-Lead ECG Diagnosis in Resource-Constrained Healthcare Environments},
journal = {Internet of Things},
year = {2026},
pages = {102155},
doi = {10.1016/j.iot.2026.102155}
}
EUVIMED