Hemodynamic forces in the left and right ventricles of the human heart using 4D flow magnetic resonance imaging: Phantom validation, reproducibility, sensitivity to respiratory gating and free analysis software

Purpose To investigate the accuracy, reproducibility and sensitivity to respiratory gating, field strength and ventricle segmentation of hemodynamic force quantification in the left and right ventricles of the heart (LV and RV) using 4D-flow magnetic resonance imaging (MRI), and to provide free hemodynamic force analysis software. Materials and methods A pulsatile flow phantom was imaged using 4D flow MRI and laser-based particle image velocimetry (PIV). Cardiac 4D flow MRI was performed in healthy volunteers at 1.5T (n = 23). Reproducibility was investigated using MR scanners from two different vendors on the same day (n = 8). Subsets of volunteers were also imaged without respiratory gating (n = 17), at 3T on the same day (n = 6), and 1-12 days later on the same scanner (n = 9, median 6 days). Agreement was measured using the intraclass correlation coefficient (ICC). Results Phantom validation showed good accuracy for both scanners (Scanner 1: bias -14 +/- 9%, y = 0.82x+0.08, R-2 = 0.96, Scanner 2: bias -12 +/- 8%, y = 0.99x-0.08, R-2 = 1.00). Force reproducibility was strong in the LV (0.09 +/- 0.07 vs 0.09 +/- 0.07 N, bias 0.00 +/- 0.04 N, ICC = 0.87) and RV (0.09 +/- 0.06 vs 0.09 +/- 0.05 N, bias 0.00 +/- 0.03, ICC = 0.83). Strong to very strong agreement was found for scans with and without respiratory gating (LV/RV: ICC = 0.94/0.95), scans on different days (ICC = 0.92/0.87), and 1.5T and 3T scans (ICC = 0.93/0.94). Conclusion Software for quantification of hemodynamic forces in 4D-flow MRI was developed, and results show high accuracy and strong to very strong reproducibility for both the LV and RV, supporting its use for research and clinical investigations. The software including source code is released freely for research.