Heart failure & cardiomyopathiesCohort study
Deep Learning Echocardiographic Trajectories of Heart Failure With Preserved Ejection Fraction
- retrospective cohort
- single center
- machine and deep learning longitudinal clustering
- training and validation sets were randomly derived
Authors' conclusion
Longitudinal phenomapping resulted in echocardiographic trajectories of cardiac structure and function with different clinical, laboratory, and survival characteristics. HFpEF is a dynamic condition, and longitudinal phenomapping may improve trajectory characterization and guide treatment strategies.
Study context
- Population
- Heart failure with preserved ejection fraction (HFpEF) patients with follow-up echocardiogram at ≥ 1 yearHFpEF cases included
- Participants
- 1,626 patients
- Exposure
- Compared to Echo-Trajectories #2 and #3, Trajectory #1
- Outcome
- all-cause mortality
Main finding
As reported in the abstract
- Echo-Trajectory #1 vs Echo-Trajectories #2 and #3: P < 0.001
Observational data: associations, not effects.
Secondary outcomes
Baseline clinical characteristics among Echo-Trajectories
- varied significantly among Echo-Trajectories by sex, age, blood pressure, obesity, comorbidities, and left ventricular mass index
- P < 0.05
Trajectories of natriuretic peptides, sodium, and renal function
- Echo-Trajectory #1 presented favorable trajectories of natriuretic peptides, sodium, and renal function
- P < 0.05
Original publication
Deep Learning Echocardiographic Trajectories of Heart Failure With Preserved Ejection Fraction. JACC Adv. Published 18 March 2026. doi:10.1016/j.jacadv.2026.102659
Summarised from the published abstract; every number is checked against it