Pulmonary vascular disease & thromboembolismDiagnostic or prognostic study
A Novel Computational Pipeline for Acquiring Pressure-Volume Hemodynamics of the Right Ventricle in Pulmonary Hypertension
- Method development and validation study
- multicentre
- 3 centres
Authors' conclusion
Its output correlates with single-beat methods and predicts clinical outcomes.
Study context
- Population
- Patients referred for right-sided heart catheterization for known or suspected pulmonary hypertension
- Participants
- 76 patients · 3 centres
- Comparison
- AI-based computational pipeline vs gold standard single-beat analysis
- Outcome
- Agreement with gold standard single-beat analysis for load-independent RV indices, and prognostic value
Performance
As reported in the abstract
Agreement between pipeline and single-beat analysis (validation and external cohorts)
- Ees: R
- 0.96
- Ees: CCC
- 0.58
- Ea: R
- 0.97
- Ea: CCC
- 0.88
- Eed: R
- 0.87
- Eed: CCC
- 0.47
- Ees/Ea: R
- 0.93
- Ees/Ea: CCC
- 0.71
Further findings
| Outcome | HR (95% CI) | |
|---|---|---|
| Prediction of clinical outcomes by calculated indices — Calculated Ea vs Single-beat analysis | 2.09 (1.04–4.20) | |
| Prediction of clinical outcomes by calculated indices — Calculated Ees/Ea vs Single-beat analysis | 0.27 (0.08–0.87) | |
| 0.050.21520Hazard ratio (log scale) |
Cluster analysis of single-beat indices
- 2 right ventricular (RV) subphenotypes with distinct hemodynamic features, more predictive of poor outcomes than standard clinical features
Original publication
A Novel Computational Pipeline for Acquiring Pressure-Volume Hemodynamics of the Right Ventricle in Pulmonary Hypertension. JACC Heart Fail. Published 31 October 2025. doi:10.1016/j.jchf.2025.102753
Summarised from the published abstract; every number is checked against it