| Title: | Pooling cohorts for deep learning analysis: a potential source of bias for electrocardiogram analysis |
| Journal: | Biomedical Signal Processing and Control |
| Published: | 1 Aug 2026 |
| DOI: | https://doi.org/10.1016/j.bspc.2026.110424 |
Publication 21253
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
Background and Objectives Deep learning models can isolate device characteristics in medical images, enabling the models to identify the origin of a medical image, which creates a bias. It is unknown whether such bias can arise with raw medical signals such as electrocardiograms (ECGs), so we aimed to quantify to what extent deep learning models can identify the study cohort and site from which an ECG originates. Methods We used 70,075 12-lead ECGs from four Danish cohorts and the UK Biobank. We trained a convolutional neural network (CNN) model using 5-fold cross-validation to identify the origin of each ECG. We also tested the effect of easy vs. difficult to predict outcomes (sex vs. diabetes) and equal vs. skewed outcome distributions. We reported accuracy and F1 score, which is less sensitive to unequal sample sizes. Results The CNN model was able to distinguish ECGs from five different cohorts with an accuracy of 93.4% and an F1 score of 77.1%. We found no bias with easily detected outcomes or equal distributions. We were unable to ascribe the CNN performance to any known factors, including ECG device, software, or population characteristics, including sex, age, and comorbidities. Conclusions Deep learning models can identify the origin of an ECG. Combining studies can introduce confounding to the model if the rate of the outcome varies between studies and the outcome is difficult to identify. In deep learning models with potential cohort confounding, we recommend training a baseline model to separate these cohorts to assess the strength of confounding.</p>
18 Authors
- Marie Mørch-Pedersen
- Malene Nørregaard
- Claus Graff
- Oliver B. Vad
- Pia R. Lundegaard
- Jesper Hastrup Svendsen
- Torben Hansen
- Allan Linneberg
- Peter G. Jørgensen
- Magnus T. Jensen
- Peter Rossing
- Christina Ellervik
- Dominik Linz
- Vajira Thambawita
- Mary M. Maleckar
- Michael A. Riegler
- Jørgen K. Kanters
- Jonas L. Isaksen
1 Application
| Application ID | Title |
|---|---|
| 43247 | Genetics of complex human traits with a main focus on cardiac disease. |
Enabling scientific discoveries that improve human health