Highly accelerated compressed sensing 4D flow MRI in congenital and acquired heart disease: comparison of aorta and main pulmonary artery flow parameters with conventional 4D flow MRI in children and young adults.

Abstract

Four-dimensional flow (4D flow) MRI has become a clinically utilized cardiovascular flow assessment tool. However, scans can be lengthy and may require anesthesia in younger children. Adding compressed sensing can decrease scan time, but its impact on hemodynamic data accuracy needs additional assessment. To compare 4D flow hemodynamics acquired with and without compressed sensing. Twenty-seven patients (median age: 13 [IQR: 9.5] years) underwent conventional and compressed sensing cardiovascular 4D flow following informed consent. Conventional 4D flow was performed using parallel imaging and an acceleration factor of 2. Compressed sensing 4D flow was performed with an acceleration factor of 7.7. Regions of interest were placed to compare flow parameters in the ascending aorta and main pulmonary artery. Paired Student's t-tests, Wilcoxon signed-rank tests, Bland-Altman plots, and intraclass correlation coefficients were conducted. A P-value of < 0.05 was considered statistically significant. Mean scan acquisition time was reduced by 59% using compressed sensing (3.4 vs. 8.2min, P < 0.001). Flow quantification was similar for compressed sensing and conventional 4D flow for the ascending aorta net flow: 47 vs. 49ml/beat (P = 0.28); forward flow: 49 vs. 50ml/beat (P = 0.07), and main pulmonary artery net flow: 49 vs. 51ml/beat (P = 0.18); forward flow: 50 vs. 55ml/beat (P = 0.07). Peak systolic velocity was significantly underestimated by compressed sensing 4D flow in the ascending aorta: 114 vs. 128cm/s (P < 0.001) and main pulmonary artery: 106 vs. 112cm/s (P = 0.02). For both the aorta and main pulmonary artery, compressed sensing 4D flow provided equivalent net and forward flow values compared to conventional 4D flow but underestimated peak systolic velocity. By reducing scan time, compressed sensing 4D flow may decrease the need for anesthesia and increase scanner output without significantly compromising data integrity.