Artificial intelligence age prediction using electrocardiogram data: Exploring biological age differences
Authors' conclusion
There are significant between-group differences in AI-ECG-predicted biological age for patient subgroups. Biological age was greater than chronological age in young hospitalized patients and lower than chronological age in older hospitalized patients. Women and patients with a single ECG recorded were biologically younger than men and patients with multiple recorded ECGs.
Study context
- Population
- Cardiology inpatients aged 20–90 years
- Participants
- 63,246 inpatients353,704 electrocardiograms (ECGs)
- Outcome
- Performance of AI-ECG age prediction (correlation, mean absolute difference between chronological and AI-predicted age)
Methodological detailsArtificial intelligence (AI) model development and validation, convolutional neural network, 80:20 development/internal validation split, external validation in UK Biobank
- AI model development and validation
- Convolutional neural network
- 80:20 development/internal validation split
- External validation in UK Biobank
Main finding
As reported in the abstract
Correlation coefficient
0.72
Mean absolute difference, chronological vs AI-predicted age
9.1 years
Observational data: associations, not effects.
Further findings
| Outcome | Estimate | P value |
|---|---|---|
| Biological age, women vs menWomen biologically younger | mean difference 10.7 months | P = 0.023 |
| Biological age, single ECG vs multiple ECGsSingle ECG biologically younger | difference 1.0 years | P < 0.0001 |
Biological minus chronological age, patients aged 20-29 years
- AI-ECG-predicted age greater than chronological age by mean
- 14.3 ± 0.2 years
Biological minus chronological age, patients aged 80-89 years
- Biological age lower than chronological age by mean
- 10.5 ± 0.1 years
Original publication
Composed by AI from the published abstract and checked against it. How Briefs are made →