85 posters,  3 sessions,  26 topics,  73 authors 

ePostersLive® by SciGen® Technologies S.A. All rights reserved.

Cardiopulmonary MRI as a diagnostic tool in Pulmonary Hypertension

Primary tabs

Poster Presenter


Average: 3.4 (5 votes)


385 reads
Cardiopulmonary MRI as a diagnostic tool in Pulmonary Hypertension

Pulmonary hypertension (PH) is defined invasively by a mean pulmonary artery pressure (mPAP) ≥ 25mmHg. The aim of this study was to derive and assess a non-invasive MRI regression model for estimation of mPAP.

Incident patients at a specialist PH centre with suspected PH from 2012 to 2016 were reviewed. Patients with left atrial volume index >41 were excluded as this is a marker of left heart disease. Biventricular volumes, mass and function and pulmonary artery flow measurements were recorded. The first half were used as a derivation cohort, the second validation. A regression equation was calculated in the derivation cohort, using any cardiac MRI metrics with statistically significant correlations with mPAP. The validation cohort was used to assess for the model’s diagnostic performance.

1272 incident suspected PH cases were identified. 957 had right heart catheter and MRI within 14 days and 603 had LA volume index ≤41. The first 300 were used as a derivation cohort and the second 303 as a validation cohort. 
The result of the regression analysis in the derivation cohort was:
mPAPmri = -178.7 + 42.730 * log10 septal angle + 7.569 * log10 Ventricular Mass Index + 3.393 * Black blood score
The accuracy of this model was assessed in the 303 patient validation cohort. There was strong correlation between RHC-mPAP and mPAPmri (r2=0.64 p<0.0001), with a small bias and close agreement on Bland-Altman. ROC showed an area under the curve of 0.95.

A newly derived parametric regression model for mPAP estimation shows strong correlation and high accuracy with RHC-mPAP in a large cohort of patients with suspected PH and normal left atrial volume, potentially allowing the avoidance of right heart catheter in selected patients.
Enter Poster ID (e.gGoNextPreviousCurrent