Aims We aimed to compare performances of conventional survival models with machine learning (ML) survival models for incident heart failure (HF) in men and women without prevalent HF, cardiomyopathy (CM) or ischaemic heart disease (IHD), and to identify potential high-risk precursors overlooked by conventional survival models. Methods and results We predicted 10-year risk of incident HF in 266 306 women (2894 events) and 212 061 men (4213 events). We constructed multivariable Cox models, first using ∼ 400 baseline characteristics, and subsequently only those remaining after LASSO stability selection. We also used Random Survival Forest (RSF) and eXtreme Gradient Survival Boosting (XGBoost). Performances were assessed using internal cross validation and hold-out sets, with C-indices, calibration curves and net-benefit analyses. Model performances were comparable during internal validation: XGBoost (C-index ± SE) (men: 0.79 ± 0.0040, women: 0.83 ± 0.0023) showed similar performance to the multivariable Cox model (men: 0.80 ± 0.0031, women: 0.83 ± 0.0022) and Cox models after LASSO stability selection, while RSF showed numerically slightly lower performance (men: 0.78 ± 0.0025, women: 0.81 ± 0.0015). Findings were similar in the hold-out sets. Age, cystatin-C, lifetime treatments/medications, other heart disease, systolic blood pressure, and spirometry measures were identified as high-risk factors in both model types for both sexes. Additionally, sex-specific and model-specific risk factors were identified. Conclusion Machine learning models and Cox proportional hazard models performed well and similarly for 10-year incident HF risk prediction in the general population. However, sex-specific and model-specific risk predictors were found. Spirometry measures, rarely included in existing models, were identified as important risk factors. Our results suggest that ML models for HF prediction in the general population reveal insights that would otherwise remain unnoticed.

Background: Optimal timing of pulmonary valve replacement (PVR) in Tetralogy of Fallot (TOF) patients remains challenging. Ventricular wall stress is considered to be an early marker of right ventricular (RV) dysfunction.

Objectives: To investigate the association of RV wall stresses and their change over time with functional parameters in TOF patients.

Methods: Ten TOF patients after surgical repair with moderate/severe pulmonary regurgitation were included. At two timepoints (median follow-up time 7.2 years), patient-specific computational biventricular models for wall stress assessment were created using CMR short-axis cine images and echocardiography-based RV pressures. RV ejection fraction (RVEF), NT-proBNP and cardiopulmonary exercise tests were used as outcome measures reflecting RV function. Associations between regional RV diastolic wall stress and RV function were investigated using linear mixed models.

Results: Increased wall stress correlated with lower RV mass (rrm = −0.70, p = 0.017) and lower RV mass-to-volume (rrm = −0.80, p = 0.003) using repeated measures. Wall stress decreased significantly over time, especially in patients with a stable RVEF (p < 0.001). Higher wall stress was independently associated with lower RVEF, adjusted for left ventricular ejection fraction, RV end-diastolic volume and time since initial surgery (decrease of 1.27% RVEF per kPa increase in wall stress, p = 0.029) using repeated measurements. No association was found between wall stress, NT-proBNP, and exercise capacity.

Conclusions: Using a computational method to calculate wall stress locally in geometrically complex ventricles, we demonstrated that lower wall stress might be important to maintain ventricular function. RV wall stress assessment can be used in serial follow-up, and is potentially an early marker of impending RV dysfunction.