Fourier Decomposition MRI provides site-resolved functional lung imaging without application of contrast media (1). A Fourier analysis of an extended non-triggered time series of morphologic lung images yields perfusion and ventilation-weighted images. It has recently been demonstrated that the perfusion-weighted data also carries valuable information regarding the delay of maximum signal increase in the lung parenchyma during a cardiac cycle in relation to the maximum increase in a central reference vessel (e.g. pulmonary trunk). The ratio of this delay to the duration of a cardiac cycle is expressed by the pulmonary phase, which might be directly associated with the delay of the maximum inflow into the lung parenchyma (2,3). In the present work we compare the pulmonary phase of patients with cystic fibrosis (CF) and healthy controls to evaluate its diagnostic potential.Perfusion measurements were performed using the SENCEFUL approach, which adds cardiac and respiratory self-navigation of quasi randomly sampled data to the classic Fourier Decomposition technique (4). Within perfusion-weighted data, the pulmonary phase can be illustrated as a separate contrast in functional lung maps. Pulmonary phase measurements of 15 patients with cystic fibrosis and 15 age-matched healthy controls were performed using a 1.5 T system (Magnetom Aera, Siemens Healthcare, Erlangen, Germany) and a 2D FLASH sequence with DC signal acquisition for self navigation (4). To generate the phase maps, 40 time frames in end-expiration were reconstructed and Fourier decomposed for each 10 mm coronal slice. Further technical details of data acquisition and image reconstruction have been described before (3,4). The lung parenchyma was segmented manually and normalized histograms were generated from the phase values obtained in all parenchyma voxels. A peak-to-offset ratio was subsequently calculated taking into account the average number of counts within the outer 18 degrees of the distribution. Results for patients with cystic fibrosis and healthy controls were compared using a Mann-Whitney-U test. Regarding the patients, the peak-to-offset ratios were also compared to the forced expiratory volume in one second (FEV₁) using Spearman correlation analysis. SENCEFUL MRI using the 2D FLASH sequence was successfully performed in all patients and volunteers without periprocedural complications. In general, the functional maps of the healthy volunteers indicated a similar phase among the entire lung parenchyma. Only few areas with higher phase dispersion could be found, e.g. corresponding to pulmonary veins or being unspecific. In contrast, maps of the patients with cystic fibrosis showed inhomogeneous pulmonary phases leading to a confetti-like appearance of the functional maps (Fig.1). In accordance to the visibly higher dispersion in the maps, the peak-to-offset ratio of the patients was significantly lower when compared with the healthy controls (CF: mean 15.9±17.5, median 7.8, min. 5.2, max. 90.0; healthy controls: mean 38.7±27.9, median 29.3, min. 5.2, max. 90.0; p=0.005). The histograms in Fig. 2 show the mean phase values among all cystic fibrosis patients and healthy controls, respectively. Spearman correlation analysis revealed a moderate correlation of the peak-to-offset ratio and the FEV₁ values of the patients with cystic fibrosis with r_s=0.72 (p=0.001).First measurements revealed that the pulmonary phase dispersion of patients with cystic fibrosis significantly differs from those of healthy subjects. A balanced pulmonary phase in healthy volunteers might indicate a homogeneous pulse wave velocity throughout the lungs. As patients with cystic fibrosis show regionally varying delays, this may be caused by different pathophysiologic mechanisms: Ventilation inhomogeneities resulting from acute or chronic inflammatory impairment of the lung parenchyma or postinflammatory fibrotic changes are supposed to cause alterations of the vascular resistance and thus of the perfusion pattern due to hypoxic vasoconstriction within the Euler-Liljestrand mechanism. This would influence the pulse wave velocity and hence the pulmonary phase as observed. Furthermore, non-perfusion-related signal effects are also conceivable e.g. due to increased pulmonary stiffness in CF patients. This could lead to an increase of cardiac or perfusion-related parenchyma motion and consecutively to alterations in signal intensity. To evaluate whether the pulmonary phase maps or the peak-to-offset ratios offer a prognostic value or correlate with clinical parameters used for disease monitoring other than FEV₁, is of high interest and will be subject to future work.