Evaluating the Harmonization of Native Digital and Digitized ECGs for ECG-AI Research

Abstract

Large epidemiologic studies have historical electrocardiogram (ECG) collections limited to paper tracings or scanned images rather than native digital files, limiting ECG-AI applications. We compared 15 ECGs available as native digital XML files and as PDF tracings digitized using ECGScan software. Both versions were processed using an ensemble of five convolutional neural networks predicting 10-year heart failure risk. Predictions were strongly correlated (Pearson r=0.804; Spearman {rho}=0.893). Predicted probabilities were (mean {+/-} SD) 0.195 {+/-} 0.048 for native digital ECGs and 0.171 {+/-} 0.042 for digitized ECGs. These preliminary findings support further evaluation of digitized ECGs for ECG-AI applications.

Evaluating SynthSR for MRI harmonization and quantitative analysis of hippocampal volume in autoimmune neurologic diseases
Nabil Seery, Ella Rowsthorn, Adil Zia +10
Scientific Reports20 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Full textMedical imaging & radiologyCrossref
Evaluating the digital dispensation of HIV self-testing (HIVST) among men who have sex with men (MSM) in Pakistan: Seeing through the Consolidated Framework for Implementation Research (CFIR) lens
Usman Ali, Muhammad Safdar Kamal Pasha, Ramesh Kumar +3
PLOS Global Public Health11 Sep 2026Peer-reviewedCC BY 4.0Full textMicrobiology NeurosciencePLOS
Retrospective harmonization of neuropsychiatric questionnaires in a large multi-site research consortium using expert mappings and semantic textual similarity
Julia-Katharina Pfarr, Eva van Heese, Joshua Curtiss +7
PsyArXiv22 Oct 2026PreprintCC BY 4.0Full textNeuroscience Psychiatry & mental healthOSF
Polysomnography harmonization schemes: A multi-institutional sleep research resource
Yujin Han, Gayeon Ryu, Yeongdon Kim +1
PLOS One5 Nov 2026Peer-reviewedCC BY 4.0Neuroscience PhysiologyPLOS
Evaluating Large Language Models as Tools to Navigate Researchers in Rapidly Evolving Research Landscapes: A Case Study in Cancer Drug Response Prediction
E. Mourelatou, I. Katakis
bioRxiv2 Sep 2026PreprintAll rights reservedFull textEpidemiology & public healthbioRxiv

Cite

BibTeX RIS

BibTeX

@misc{soliman2026evaluating,
  title = {{Evaluating the Harmonization of Native Digital and Digitized ECGs for ECG-AI Research}},
  author = {E. Z. Soliman and H. Bukhari and L. Patterson and S. Moldibi and I. Karabayir and O. Akbilgic},
  howpublished = {medRxiv},
  year = {2026},
  month = sep,
  doi = {10.64898/2026.09.21.26363586},
  url = {https://www.medrxiv.org/content/10.64898/2026.09.21.26363586v1},
  note = {CC BY-NC},
}

RIS

TY  - UNPB
TI  - Evaluating the Harmonization of Native Digital and Digitized ECGs for ECG-AI Research
AU  - E. Z. Soliman
AU  - H. Bukhari
AU  - L. Patterson
AU  - S. Moldibi
AU  - I. Karabayir
AU  - O. Akbilgic
PY  - 2026
DA  - 2026/09/21
PB  - medRxiv
DO  - 10.64898/2026.09.21.26363586
UR  - https://www.medrxiv.org/content/10.64898/2026.09.21.26363586v1
AB  - Large epidemiologic studies have historical electrocardiogram (ECG) collections limited to paper tracings or scanned images rather than native digital files, limiting ECG-AI applications. We compared 15 ECGs available as native digital XML files and as PDF tracings digitized using ECGScan software. Both versions were processed using an ensemble of five convolutional neural networks predicting 10-year heart failure risk. Predictions were strongly correlated (Pearson r=0.804; Spearman {rho}=0.893). Predicted probabilities were (mean {+/-} SD) 0.195 {+/-} 0.048 for native digital ECGs and 0.171 {+/-} 0.042 for digitized ECGs. These preliminary findings support further evaluation of digitized ECGs for ECG-AI applications.
ER  -

Abstract

Related

Cite