Currently, only self-gated or ultra-fast sequences allow an adequate temporal resolution to resolve different cardiac phases during lung perfusion. First studies show the benefit of this additional information. Nevertheless such techniques are not widely available. Therefore, in this study a post processing method is assessed, which can increase the temporal resolution by sorting images according to their cardiac phase using a piecewise cosine fit. The feasibility is demonstrated in 6 healthy volunteers and two patients with chronic thromboembolic hypertension (CTEPH). The easy implementation and possibility of retrospective evaluation of existing Fourier Decomposition acquisitions are the advantages of this method.
Six healthy volunteers (age 24-51, 3 female) were enrolled in this study. For each subject a sagittal acquisition of the right lung was performed on a 1.5T scanner during free breathing in prone and supine position using a spoiled gradient sequence (FOV=50x50 cm2, matrix size=128x128, slice thickness=15mm, TE=0.82, TR=3ms, flip angle=5° and GRAPPA=2) over a period of 48s at a temporal resolution of 192ms. Additionally, three coronal acquisitions were performed in supine position only. A retrospective analysis was performed on FD data of two patients with chronic thromboembolic hypertension (CTEPH) acquired with a similar sequence protocol as described above and compared with dynamic contrast enhanced (DCE) MRI.
PRELP analysis was as follows: After image registration (ANTS6) an edge-preserving filter (guided image filtering7) was applied to all images using the averaged image of the registered image time series as the guiding image. This filtering is especially suited for registered image time series, since it enables to use the sharp edge information of the averaged image, which has a higher signal to noise ratio, and at the same time preserves the contrast information of the individual images. To remove signal variations due to respiration a high-pass filter at 0.8 Hz was applied.
Next, a group of vessels were segmented and used as a region of interest (ROI) for spatial averaging to obtain a signal time series (TS) for phase sorting. The local maxima were used to subdivide the TS into smaller sections. Subsequently, piecewise cosines were fitted (fit parameters: amplitude, phase offset, frequency, Figure 1a). Since a cosine has a definite phase at each time point, phase can be assigned to each measurement. Using this information the images were sorted into one cardiac cycle (Figure 1b). Using a Gaussian kernel (sigma = 0.1) 30 images were interpolated at an equidistant time grid (Figure 2) covering one cardiac cycle, achieving a nominal temporal resolution of 33ms depending on the heart rate. The phase of each voxel was determined in relation to the vessels ROI in relative units of a full cardiac cycle. Using this information time-to-peak (TTP) maps were calculated.
1. Bauman G, Puderbach M, Deimling M, et al. Non-contrast-enhanced perfusion and ventilation assessment of the human lung by means of fourier decomposition in proton MRI. Magnetic resonance in medicine 2009;62(3):656-664.
2. Bauman G, Eichinger M, Uecker M. High temporal resolution radial bSSFP sequence with nonlinear inverse reconstruction for the measurement of the pulmonary blood inflow time using Fourier decomposition MRI. Proc Intl Soc Mag Reson Med 2012;20:1340.
3. Fischer A, Weick S, Ritter CO, et al. SElf-gated Non-Contrast-Enhanced FUnctional Lung imaging (SENCEFUL) using a quasi-random fast low-angle shot (FLASH) sequence and proton MRI. NMR in Biomedicine 2014;27(8):907-917.
4. Veldhoen S, Weng AM, Wirth C, et al. Pulmonary Phase Imaging using Self-Gated Fourier Decomposition MRI in Patients with Cystic Fibrosis. Proc Intl Soc Mag Reson Med 2016;24:1144.
5. Henderson AC, Sá RC, Theilmann RJ, et al. The gravitational distribution of ventilation-perfusion ratio is more uniform in prone than supine posture in the normal human lung. Journal of Applied Physiology 2013;115(3):313-324.
6. Avants BB, Tustison NJ, Song G, et al. A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 2011;54(3):2033-2044.
7. He K, Sun J, Tang X. Guided Image Filtering. IEEE Transactions on Pattern Analysis and Machine Intelligence 2013;35(6):1397-1409.