Publications
Take a look

Marta Tibiletti, Josephine H Naish, Katharina Martini, Thomas Frauenfelder, Geoff JM Parker

British Chapter of ISMRM Meeting 2019

Abstract

Introduction:

Oxygen Enhanced MRI (OE-MRI) exploits the paramagnetic properties of molecular oxygen to modify local T1 values to explore local lung function. During a dynamic OE-MRI experiment, the subject breaths varying concentrations of O₂. Various parameters can be extracted, such as the ventilated volume fraction (VVF), defined as is the fraction of lung tissue showing O₂ enhancement, and the oxygen wash-in time (T_up) [1]. In this work, we calculated τup, VVF, VVF-masked T_up and compared them with pulmonary function tests in a population of healthy volunteers (HV) and cystic fibrosis (CF) subjects.

Methods:

Analysis was applied retrospectively to OE-MRI data acquired from a previously published study [1]. 20 patients with CF (20 - 40 years, 13 male) and 4 HV (27-37 y, 2 male) underwent dynamic OE-MRI on a 1.5 T Philips Achieva MRI scanner. A free-breathing protocol based on an inversion-prepared centric ordered single shot 3D-turbo field echo sequence was used. The dynamic acquisition lasted 15 min (90 volumes), during which gas was delivered at 15 l/min via a disposable non-rebreathing mask and switched at minute 2 from medical air to 100% O₂, and back to air at minute 10. All images were registered to correct for breathing motion using a non-linear registration algorithm [4]. Maps of T_up were derived from the signal intensity curves for each pixel by fitting to a mono-exponential recovery function. VVF was considered positive where the Akaike information criterion favours an exponential fit over a constant function. VVF fraction, median T_u and median T_up calculated only within VVF positive pixels (masked T_up) were extracted and compared to the results from conventional spirometry (FEV1, FVC and FEV1/FVC corrected for age, height and gender). Relationships between variables were evaluated with Pearson correlation with p < 0.05 considered to indicate the presence of a statistically significant correlations.

Results:

The table present the R2 and p-values of correlation between calculated parameters and corrected spirometry results.

                         FVC pred [%]     FEV1 pred [%]   FEV1/FVC pred [%]

VVF                   0.54 (p<0.001)   0.70 (p<0.001)    0.24 (p=0.02)

T_up [s]                0.00 (p=0.89)     0.01 (p=0.59)      0.12 (p=0.12)

Masked T_up [s]  0.03 (p=0.38)     0.11 (p=0.11)       0.27 (p=0.01)

Discussion:

VVF highly correlates with measurements of lung function derived from spirometry, particularly with FEV1. Oxygen wash-in time correlated significantly with FEV1/FVC pred [%] only when non-enhancing voxels are excluded.

Conclusion:

Parameters calculated from dynamic OE-MRI are highly correlated with measurement of lung function derived from spirometry.

Reference: [1] Martini K, et al. European Radiology:1-11 (2018)