Comparison of 3D UTE free-breathing lung MRI with hyperpolarized 129Xe MRI in pediatric cystic fibrosis

Purpose: To compare phase-resolved functional lung (PREFUL) regional ventilation derived from a free breathing 3D UTE radial MRI acquisition to hyperpolarized Xe-129-MRI (Xe-MRI), conventional 2D multi-slice PREFUL MRI, and pulmonary function tests in pediatric cystic fibrosis (CF) lung disease. Methods: Free-breathing 3D UTE and 2D multi-slice H-1 MRI as well as Xe-MRI were acquired in 12 stable pediatric CF patients. Using PREFUL, regional ventilation (RVent) maps were calculated from the free-breathing data. Ventilation defect percentage (VDP) was determined from 3D and 2D RVent maps (2D VDPRVent and 3D VDPRVent, respectively) and Xe-MRI ventilation (VDPXe). VDP was calculated for the whole lung and for eight regions based on left/right, anterior/posterior, and superior/inferior divisions of the lung. Global and regional VDP was compared between the three methods using Bland-Altman analysis, linear mixed model-based correlation, and one-way analysis of variance and multiple comparisons tests. Results: Global 3D VDPRVent, VDPXe, and 2D VDPRVent were all strongly correlated (all R-2 > 0.62, p < 0.0001) and showed minimal, non-significant bias (all <2%, p > 0.05). Three dimensional and 2D VDPRVent significantly correlated to VDPXe in most of the separate lung regions (R-2=0.18-0.74, p < 0.04), but showed lower inter-agreement. The superior/anterior lung regions showed the least agreement between all three methods (all p > 0.12). Conclusion: Absolute VDP assessed by 3D UTE PREFUL MRI showed good global agreement with Xe-MRI and 2D multi-slice PREFUL MRI in pediatric CF lung disease. Therefore, 3D UTE PREFUL MRI offers a sensitive and potentially more accessible alternative to Xe-MRI for regional volumetric evaluation of ventilation.