%0 Journal Article
%9 ACL : Articles dans des revues avec comité de lecture répertoriées par l'AERES
%A Lence, A.
%A Fall, Ahmad
%A Cohen, S. D.
%A Granese, F.
%A Zucker, Jean-Daniel
%A Salem, J. E.
%A Prifti, Edi
%T ECGTIzER : an open-source, fully automated pipeline for digitization and signal recovery from paper electrocardiograms
%D 2026
%L fdi:010095306
%G ENG
%J Biomedical Signal Processing and Control
%@ 1746-8094
%K Electrocardiograms ; Digitization ; Signal recovery ; Fully automated ; Torsades-de-pointes ; Artificial intelligence
%M ISI:001584508700001
%N D
%P 108710 [12 ]
%R 10.1016/j.bspc.2025.108710
%U https://www.documentation.ird.fr/hor/fdi:010095306
%> https://horizon.documentation.ird.fr/exl-doc/pleins_textes/2025-11/010095306.pdf
%V 112
%W Horizon (IRD)
%X Background and Objective: Electrocardiograms (ECGs) are essential for diagnosing cardiac pathologies, yet traditional paper-based ECG storage poses significant challenges for automated analyses. Despite the growing interest in leveraging AI for ECG analysis, there remains a lack of accessible, fully automated tools for digitizing paper-based ECGs. Existing solutions are often incomplete, behind paywalls, or not suited for large-scale use. To address this gap, we present ECGTIzER: an open-source, fully automated tool that enables high-fidelity digitization of paper ECGs, ensuring long-term preservation of clinical data and unlocking their potential for modern AI-driven analysis. Methods: ECGTIzER employs automated lead detection, three different pixel-based signal extraction algorithms, and a deep learning-based signal reconstruction module. We evaluated ECGTIzER on two datasets: a real-life cohort from the COVID-19 pandemic (JOCOVID) and a publicly available dataset (PTB-XL). Performance was compared with two existing methods: the fully automated ECGMINER and the semi-automated PAPERECG, which requires human intervention. The tools' digitization performance was assessed in terms of signal recovery, the fidelity of clinically relevant feature measurement and downstream AI classification tasks on a third dataset (GENEREPOL). Results: Results show that ECGTIzER outperforms state-of-the-art methods, with its ECGTIzERFrag algorithm delivering superior signal recovery performance. While PAPERECG demonstrated better outcomes than ECGMINER, it also requires human input. Conclusions: ECGTIzER enhances the usability of historical ECG data and supports advanced AI-based diagnostic methods, making it a valuable addition to the field of AI in ECG analysis.
%$ 122 ; 050