The place of MRI in the assessment of pulmonary hypertension remains controversial. Several studies proposed to use MRI to assess pulmonary pressure but the level of proof is low.

To evaluate the diagnostic power of cardiac MRI within a non-selected population of patients suspected of pulmonary hypertension after an echocardiography.

Fifty-six consecutive patients, suspected of pulmonary hypertension after an echocardiography, were assessed with right heart catheterization and cardiac MRI (including a high temporal resolution pulmonary flow curve). We extracted from the MR data the main parameters proposed by all precedent studies available in the literature. We looked for multivariate linear relations between those parameters and the mean pulmonary arterial pressure (mPAP), and eventually assessed with a logic regression the ability of those parameters to diagnose pulmonary hypertension in our population.

The multivariate model retained only two parameters: the right ventricle ejection fraction and the pulmonary trunk minimum area. The prediction of mPAP (r²=0.5) yielded limits of agreement of 15mmHg. However, the prediction of pulmonary hypertension within the population was feasible and the method yielded a specificity of 80% for a sensitivity of 100% (Fig. 1).

The performance of MRI to assess mPAP is too low to be used as a replacement for right heart catheterization but MRI could be used as second line examination after echocardiography to avoid right heart catheterization for normal patients.