December 5, 2024Journal of Cardiovascular Magnetic ResonanceOpen Access

Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model

View Full Paper

Ask AI

Bookmark

Why the study?

Pulmonary hypertension diagnosis requires invasive right heart catheterization, making a non-invasive alternative clinically valuable.

Does an explainable convolutional neural network using cardiac MR cine series accurately predict mean pulmonary artery pressure?

Structured PICO

Does an explainable convolutional neural network using cardiac MR cine series accurately predict mean pulmonary artery pressure?

Population

1646 examinations for estimating right heart catheterization parameters non-invasively from cardiac magnetic resonance data

Intervention

Explainable convolutional neural network (CNN) taking cardiac MR cine series from four different views as input

Comparator

Invasive right heart catheterization (RHC) as reference standard

Outcome

Mean pulmonary artery pressure (mPAP)surrogate

A deep learning model using cardiac MR cine series can non-invasively estimate mean pulmonary artery pressure, highlighting the right ventricle on short-axis view as a key feature.

Implication

May enable non-invasive RHC estimation from CMR in PH; leaves open prospective validation before clinical use.

Abstract

BACKGROUND: Pulmonary hypertension (PH) is a heterogeneous condition and regardless of etiology impacts negatively on survival. Diagnosis of PH is based on hemodynamic parameters measured invasively at right heart catheterization (RHC); however, a non-invasive alternative would be clinically valuable. Our aim was to estimate RHC parameters non-invasively from cardiac magnetic resonance (MR) data using deep learning models and to identify key contributing imaging features. METHODS: We constructed an explainable convolutional neural network (CNN) taking cardiac MR cine series from four different views as input to predict mean pulmonary artery pressure (mPAP). The model was trained and evaluated on 1646 examinations. The model's attention weight and predictive performance associated with each frame, view, or phase were used to judge its importance. Additionally, the importance of each cardiac chamber was inferred by perturbing part of the input pixels. RESULTS: of 0.64 in predicting mPAP and identified the right ventricle region on short-axis view to be especially informative. CONCLUSION: Hemodynamic parameters can be estimated non-invasively with a CNN, using MR cine series from four views, revealing key contributing features at the same time.