Purpose: To compare low-pass imaging dynamic acquisitions (LIDA) approach in combination with a group-oriented registration (GOREG) scheme with conventional Fourier decomposition (FD). Methods: Dynamic images of six healthy volunteers and one chronic obstructive pulmonary disease (COPD) patient were acquired on a 1.5 Tesla scanner using a spoiled gradient echo sequence. For the volunteers each slice was acquired during regular (RB) and irregular breathing (IB). For conventional registration (CR), all images were directly registered to the intermediate respiratory position. For the group-oriented registration (GOREG) a group-based step-by-step scheme was used. The registered datasets were evaluated with FD and LIDA. The image quality was assessed by image sharpness metric and parenchyma-vessel contrast. The GOREG+LIDA and CR+FD results were compared with computed tomography (CT) in a patient with COPD. To show feasibility of the presented method at 3 Tesla (T), a healthy volunteer was included, who was scanned at 1.5T and 3T on the same day. Results: Improved image sharpness (predominantly by GOREG, P < 0.0125) and contrast (predominantly by LIDA, P < 0.0125) were obtained for RB and IB. Synergetic improvements were found for GOREG+LIDA. Hypoventilated regions in the COPD patient calculated with GOREG+LIDA showed a good agreement with emphysema identified by CT. Conclusion: GOREG+LIDA can be used to calculate quantitative ventilation images with improved image quality, which is an important step toward clinical translation of this method. The presented method is also feasible at 3T. Magn Reson Med 78:1496–1505, 2017.
- Fourier decomposition
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging