Key result
A regression forest model using standard lead II electrocardiogram features estimated living age with a mean absolute difference of 9.55 years compared to known age in the test set.
Why the study?
A straightforward bedside method for living age estimation from standard lead II ECG using explainable variables and machine learning algorithms was needed.
Can machine learning regression algorithms accurately estimate chronological age from a standard lead II ECG?
Cross-Sectional (n=729)
Can machine learning regression algorithms accurately estimate chronological age from a standard lead II ECG?
Machine learning applied to standard lead II ECG variables can estimate chronological age with reasonable accuracy, offering a potential bedside complementary tool for age estimation.
May support forensic age estimation from lead II ECG; leaves open prospective validation before any clinical adoption.
Objective: The objective of the study was to develop a straightforward bedside “medical age-test” for living age estimation from standard lead II electrocardiogram (ECG) using a limited number of explainable variables and to compare the performance of various linear and nonlinear machine learning (ML) regression algorithms. Materials and Methods: A nontruncated age dataset spanning 1–92 years was compiled, comprising 729 individuals (382 males and 347 females). Resting ECG tracings were recorded for all subjects in the supine position. A fiducial point-based method was adopted to detect the corresponding onset, offset, and peak points of the PQRST signal to compute the PR, RR, and QT intervals and the QRS waveform amplitudes and duration. The performance of five regression models was compared on unseen test data (20% of the original dataset; n = 145). Results: The feature importance plot ranked the RR interval, R amplitude, and sex as the top three features contributing to the model, whereas the PR interval had the lowest contribution. For the most parsimonious model (regression forest [RF]), the mean absolute difference between the known and estimated age was 8.99 years in the training set and 9.55 years in the test set. Conclusions: The RF model demonstrated superior accuracy in age estimation, supporting its use in living age estimation, particularly in adults aged <50 years. The standard single-lead II ECG represents an easy, accessible, and reliable biosignal for bedside evaluation complementary to radiological scanning.
No takes yet. Share an insight, caveat, or question.
Ghitani et al. (2026) reported a cross-sectional. Regression forest model using standard lead II ECG features was evaluated on Mean absolute difference between known and estimated age. A regression forest model using standard lead II electrocardiogram features estimated living age with a mean absolute difference of 9.55 years compared to known age in the test set.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: