Key result
Machine learning model using submaximal exercise predicts low exercise capacity in HF with 0.94 AUC.
Why the study?
Does a machine-learning model using pragmatic submaximal exercise parameters accurately predict exercise intolerance in patients with heart failure?
Cohort (n=343)
No
Does a machine-learning model using pragmatic submaximal exercise parameters accurately predict exercise intolerance in patients with heart failure?
A machine-learning model using five readily available clinical and submaximal exercise parameters can accurately predict severe exercise intolerance (peak VO2 ≤14 mL/min/kg) in patients with heart failure, potentially reducing the need for maximal exertion testing.
Answers are grounded in this paper and cite their sources.
May support submaximal prediction of exercise intolerance in HF; hypothesis-generating pending prospective validation.
Background: in HF patients using machine learning (ML). Methods and Results: at respiratory exchange ratio of 1.00. Using these 5 predictors, an optimized predictive model was validated on the testing dataset, yielding an accuracy of 85%, F1 score of 0.81, and area under the receiver operating curve of 0.94 (95% confidence interval: 0.89-1.00). Conclusions: in patients with HF.
No takes yet. Share an insight, caveat, or question.
Kato et al. (2025) conducted a cohort in Chronic heart failure with left ventricular ejection fraction <50% (n=343). Machine-learning predictive model (Support Vector Machine) was evaluated on Prediction of low peak oxygen uptake (≤14 mL/min/kg) on the testing dataset (AUC 0.94, 95% CI 0.89-1.00). A machine-learning model using five submaximal exercise parameters accurately predicted low exercise capacity in heart failure patients, achieving an AUC of 0.94.
Synapse has enriched 10 related papers. Consider them for comparative context: