Project presentation 2024

3D PREFUL-MRI for visualization and quantification of regional lung function

In recent years, a novel 2D MRI method (PREFUL: Phase-Resolved Functional LungMRI) for dynamic visualization and quantification of regional lung ventilation and perfusion without contrast agent has been invented, validated and applied in clinical studies in the working group of Prof. Dr. Jens Vogel-Claussen at the Institute of Diagnostic and Interventional Radiology with funding from the German Center for Lung Research (DZL ). PREFUL uses the intrinsic signal alterations caused by breathing and the heartbeat during data processing after the MRI examination. To date, PREFUL-MRI has been shown to be sensitive for early assessment of treatment response in patients with chronic obstructive pulmonary disease (COPD), cystic fibrosis (CF), asthma and COVID-19 and early prediction of chronic lung transplant failure. The 3D MRI-PREFUL method was developed to capture the entire lung volume without gaps. Commercially available MRI sequences are used to enable the scalability of the PREFUL method.

In this validation study, the ventilation maps obtained from the 3D-PREFUL MRI are compared with a direct ventilation measurement from the 129Xe MRI (gold standard) in patients with COPD and healthy volunteers. This study was conducted in collaboration with Fraunhofer ITEM.

Thirty-one COPD patients and 12 healthy controls underwent free-breathing 3D-PREFUL MRI and 129Xe MRI. Ventilation defect (VD) maps were created for both MRI techniques and the corresponding values for the percentage of ventilation defects (VDP) were calculated. All parameters of both techniques were compared using the Spearman correlation coefficient (r). The differences between the VDP values were quantified using Bland-Altman analysis and tested using the Wilcoxon signed rank test. In a regional comparison of the VD maps, the spatial overlap and Sørensen-Dice coefficients of healthy and defective areas were calculated.

On a global scale, both static (RVent) and 3D PREFUL VDP values derived from dynamic flow-volume curves (FVL-CM) correlated significantly with the VDP value derived from 129Xe ventilation imaging (all r > 0.65; all p < 0.0001). 129Xe VDP was significantly greater than 3D PREFUL-derived VDPRVent (mean deviation = 10.5%, p < 0.001) and VDPFVL-CM (mean deviation = 11.3%, p < 0.0001), but not for VDPCombined - a combination of RVent and FVL-CM (mean deviation = 1.7%, p = 0.70). The overall regional agreement of the 129Xe and 3D-PREFUL-VD maps was between 60 % and 63 %.

Free-breathing 3D-PREFUL MRI showed a strong correlation with breath-hold hyperpolarized 129Xe MRI in terms of VDP values and slight differences in the detection of VDs at the regional level. This work is an important step for the clinical translation of 3D PREFUL MRI.