@article{fdi:010095306,
  title = {{ECGTI}z{ER} : an open-source, fully automated pipeline for digitization and signal recovery from paper electrocardiograms},
  author = {{L}ence, {A}. and {F}all, {A}hmad and {C}ohen, {S}. {D}. and {G}ranese, {F}. and {Z}ucker, {J}ean-{D}aniel and {S}alem, {J}. {E}. and {P}rifti, {E}di},
  editor = {},
  language = {{ENG}},
  abstract = {{B}ackground and {O}bjective: {E}lectrocardiograms ({ECG}s) are essential for diagnosing cardiac pathologies, yet traditional paper-based {ECG} storage poses significant challenges for automated analyses. {D}espite the growing interest in leveraging {AI} for {ECG} analysis, there remains a lack of accessible, fully automated tools for digitizing paper-based {ECG}s. {E}xisting solutions are often incomplete, behind paywalls, or not suited for large-scale use. {T}o address this gap, we present {ECGTI}z{ER}: an open-source, fully automated tool that enables high-fidelity digitization of paper {ECG}s, ensuring long-term preservation of clinical data and unlocking their potential for modern {AI}-driven analysis. {M}ethods: {ECGTI}z{ER} employs automated lead detection, three different pixel-based signal extraction algorithms, and a deep learning-based signal reconstruction module. {W}e evaluated {ECGTI}z{ER} on two datasets: a real-life cohort from the {COVID}-19 pandemic ({JOCOVID}) and a publicly available dataset ({PTB}-{XL}). {P}erformance was compared with two existing methods: the fully automated {ECGMINER} and the semi-automated {PAPERECG}, which requires human intervention. {T}he 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}). {R}esults: {R}esults show that {ECGTI}z{ER} outperforms state-of-the-art methods, with its {ECGTI}z{ERF}rag algorithm delivering superior signal recovery performance. {W}hile {PAPERECG} demonstrated better outcomes than {ECGMINER}, it also requires human input. {C}onclusions: {ECGTI}z{ER} 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.},
  keywords = {{E}lectrocardiograms ; {D}igitization ; {S}ignal recovery ; {F}ully automated ; {T}orsades-de-pointes ; {A}rtificial intelligence},
  booktitle = {},
  journal = {{B}iomedical {S}ignal {P}rocessing and {C}ontrol},
  volume = {112},
  numero = {{D}},
  pages = {108710 [12 p.]},
  ISSN = {1746-8094},
  year = {2026},
  DOI = {10.1016/j.bspc.2025.108710},
  URL = {https://www.documentation.ird.fr/hor/fdi:010095306},
}