Key result
Deep learning models predict age within ~8 years and identify sex from standard 12-lead ECGs.
Why the study?
Several deep learning models have been published to predict age and sex from 12-lead ECGs, but no direct comparisons exist.
How do different deep learning models compare in predicting age and sex from standard 12-lead ECGs?
How do different deep learning models compare in predicting age and sex from standard 12-lead ECGs?
Deep learning models can predict age with a mean absolute error of approximately 8 years and sex with high accuracy (AUROC 0.92) from standard 12-lead ECGs.
ECG age/sex prediction warrants further biomarker exploration; extends DL applications but leaves open prospective clinical validation.
Objective: The 12-lead electrocardiogram (ECG) is routine in clinical use and deep learning approaches have been shown to have the identify features not immediately apparent to human interpreters including age and sex. Several models have been published but no direct comparisons exist. Approach: We implemented three previously published models and one unpublished model to predict age and sex from a 12-lead ECG and then compared their performance on an open-access data set. Main results: All models converged and were evaluated on the holdout set. The best preforming age prediction model had a hold-out set mean absolute error of 8.06 years. The best preforming sex prediction model had a hold-out set area under the receiver operating curve of 0.92. Significance: We compared performance of four models on an open-access dataset.
No takes yet. Share an insight, caveat, or question.
Barros et al. (2024) studied General hospital population (n=800,000). Deep learning models (ResNet, CNN, MHA, ResNet+MHA) vs. Comparison between models was evaluated on Age prediction (Mean Absolute Error). Deep learning models can predict age and sex from standard 12-lead ECGs with reasonable accuracy, with the ResNet model achieving a mean absolute error of 8.06 years for age prediction and the ResNet+MHA model achieving an AUROC of 0.925 for sex prediction.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: