Heart failure & cardiomyopathiesDiagnostic or prognostic study
Noninvasive Pulmonary Capillary Wedge Pressure Estimation in Heart Failure Patients With the Use of Wearable Sensing and AI
- Diagnostic accuracy study
- multicentre
- prospective
- blinded core laboratory
- held-out test set
Authors' conclusion
The CardioTag and its ML algorithm estimate PCWP with accuracy approaching implantable hemodynamic sensors
Study context
- Population
- Inpatients and outpatients with heart failure with reduced ejection fraction (HFrEF) (EF ≤40%)
- Participants
- 310 inpatients
- Comparison
- Wearable sensor + machine learning (ML) model vs RHC for pulmonary capillary wedge pressure (PCWP) estimation
- Outcome
- Error of PCWP estimate vs RHC
Performance
As reported in the abstract
PCWP estimation error vs RHC (held-out test set)
Error of model-estimated PCWP
1.04 ± 5.57 mm Hg
Limits of agreement
−9.9 to 11.9 mm Hg
Original publication
Noninvasive Pulmonary Capillary Wedge Pressure Estimation in Heart Failure Patients With the Use of Wearable Sensing and AI. JACC Heart Fail. Published 20 June 2025. doi:10.1016/j.jchf.2025.102513
Summarised from the published abstract; every number is checked against it