Guillaume Chassagnon1, Charlotte Martin2, Wadie Ben Hassen3, Gael Freche2, Souhail Bennani2, Baptiste Morel4, Marie-Pierre Revel5. 1. Radiology Department, Groupe Hospitalier Cochin-Hotel Dieu, AP-HP, Université Paris Descartes, 27 Rue du Faubourg Saint-Jacques, 75014 Paris, France; Center for Visual Computing, Ecole CentraleSupelec, 3 Rue Joliot Curie, 91190, Gif-sur-Yvette, France. 2. Radiology Department, Groupe Hospitalier Cochin-Hotel Dieu, AP-HP, Université Paris Descartes, 27 Rue du Faubourg Saint-Jacques, 75014 Paris, France. 3. Siemens Healthineers France, 40 avenue des fruitiers, 93210 Saint-Denis, France. 4. Radiology Department, Hopital Clocheville, CHU Tours, Université François Rabelais, 49 Boulevard Béranger, 37000 Tours, France. 5. Radiology Department, Groupe Hospitalier Cochin-Hotel Dieu, AP-HP, Université Paris Descartes, 27 Rue du Faubourg Saint-Jacques, 75014 Paris, France. Electronic address: marie-pierre.revel@aphp.fr.
Abstract
PURPOSE: To assess the influence of magnetic field strength and additionally of acquisition and reconstruction parameters on the quality of high-resolution lung MRI, using a prototype Ultrashort-TE (UTE) sequence. MATERIALS AND METHODS: This prospective study received ethical approval and all participants provided written informed consent. From January to February 2018, images were obtained in 10 healthy volunteers at 1.5 T and 3 T with a prototypical free-breathing UTE spiral 3D-GRE sequence with volumetric interpolation (VIBE) sequence and near-millimeter resolution. Five sequences were acquired to assess the effects of magnetic field strength (1.5 vs 3 T), voxel resolution (1.2 vs 1.0mm3), number of spiral interleaves (464 vs 264) and iterative reconstruction (iterative self-consistent parallel imaging reconstruction [SPIRiT] versus Non-Uniform Fourier Transform [NUFFT]) on image quality. Image quality was assessed by two independent observers. They evaluated the proportion of detected airways from the trachea down to the subsegmental level and placed ROI in the lung parenchyma, airways and vessels to calculate signal-to noise (SNR) and contrast-to-noise (CNR) ratios. Continuous variables were expressed as mean ± standard deviation and were compared by t-test. RESULTS: Nearly complete visualization of the segmental bronchi (94 ± 12 to 99 ± 3%) was obtained with all sequences. Acquisition at 3 T (p < 0.001), use of a fewer spiral interleaves (p < 0.001) and NUFFT reconstruction (p < 0.001) all resulted in a significantly lower visibility of the subsegmental bronchi, while a smaller voxel size improved their visibility (p = 0.001). SNR and CNR were significantly lower at 3 T (140.2 ± 19.9 vs 190.2 ± 34.8, p < 0.001; and 5.7 ± 2.4vs 10.8 ± 2.8, p < 0.001, respectively). CONCLUSIONS: Using equivalent acquisition and reconstruction parameters, image quality was lower at 3 T than at 1.5 T with decreased visibility of the subsegmental bronchi and lower SNR and CNR values.
PURPOSE: To assess the influence of magnetic field strength and additionally of acquisition and reconstruction parameters on the quality of high-resolution lung MRI, using a prototype Ultrashort-TE (UTE) sequence. MATERIALS AND METHODS: This prospective study received ethical approval and all participants provided written informed consent. From January to February 2018, images were obtained in 10 healthy volunteers at 1.5 T and 3 T with a prototypical free-breathing UTE spiral 3D-GRE sequence with volumetric interpolation (VIBE) sequence and near-millimeter resolution. Five sequences were acquired to assess the effects of magnetic field strength (1.5 vs 3 T), voxel resolution (1.2 vs 1.0mm3), number of spiral interleaves (464 vs 264) and iterative reconstruction (iterative self-consistent parallel imaging reconstruction [SPIRiT] versus Non-Uniform Fourier Transform [NUFFT]) on image quality. Image quality was assessed by two independent observers. They evaluated the proportion of detected airways from the trachea down to the subsegmental level and placed ROI in the lung parenchyma, airways and vessels to calculate signal-to noise (SNR) and contrast-to-noise (CNR) ratios. Continuous variables were expressed as mean ± standard deviation and were compared by t-test. RESULTS: Nearly complete visualization of the segmental bronchi (94 ± 12 to 99 ± 3%) was obtained with all sequences. Acquisition at 3 T (p < 0.001), use of a fewer spiral interleaves (p < 0.001) and NUFFT reconstruction (p < 0.001) all resulted in a significantly lower visibility of the subsegmental bronchi, while a smaller voxel size improved their visibility (p = 0.001). SNR and CNR were significantly lower at 3 T (140.2 ± 19.9 vs 190.2 ± 34.8, p < 0.001; and 5.7 ± 2.4vs 10.8 ± 2.8, p < 0.001, respectively). CONCLUSIONS: Using equivalent acquisition and reconstruction parameters, image quality was lower at 3 T than at 1.5 T with decreased visibility of the subsegmental bronchi and lower SNR and CNR values.
Authors: Ipshita Bhattacharya; Rajiv Ramasawmy; Ahsan Javed; Marcus Y Chen; Thomas Benkert; Waqas Majeed; Robert J Lederman; Joel Moss; Robert S Balaban; Adrienne E Campbell-Washburn Journal: NMR Biomed Date: 2021-06-02 Impact factor: 4.478