Charlie C Park1, Gavin Hamilton1, Ajinkya Desai2, Kevin A Zand1, Tanya Wolfson3, Jonathan C Hooker1, Eduardo Costa1, Elhamy Heba1, Lisa Clark1, Anthony Gamst3, Rohit Loomba4,5, Michael S Middleton1, Claude B Sirlin6. 1. MR3T Bydder Laboratory, Liver Imaging Group, Department of Radiology, University of California, San Diego, 408 Dickinson Street, MC 8226, San Diego, CA, 92103-8226, USA. 2. Department of Diagnostic and Interventional Radiology, Rochester General Hospital, Rochester, NY, USA. 3. Computational and Applied Statistics Laboratory (CASL), San Diego Supercomputer Center (SDSC), University of California, San Diego, La Jolla, CA, USA. 4. Division of Gastroenterology, Department of Medicine, University of California, San Diego, La Jolla, CA, USA. 5. Division of Epidemiology, Department of Family Medicine and Preventive Medicine, University of California, San Diego, La Jolla, CA, USA. 6. MR3T Bydder Laboratory, Liver Imaging Group, Department of Radiology, University of California, San Diego, 408 Dickinson Street, MC 8226, San Diego, CA, 92103-8226, USA. csirlin@ucsd.edu.
Abstract
PURPOSE: The aim of the study was to determine in patients undergoing gadoxetate disodium (Gx)-enhanced MR exams whether proton density fat fraction (PDFF) estimation accuracy of magnitude-based multi-gradient-echo MRI (MRI-M) could be improved by using high flip angle (FA) on post-contrast images. MATERIALS AND METHODS: Thirty-one adults with known or suspected hepatic steatosis undergoing 3T clinical Gx-enhanced liver MRI were enrolled prospectively. MR spectroscopy (MRS), the reference standard, was performed before Gx to measure MRS-PDFF. Low (10°)- and high (50°)-flip angle (FA) MRI-M sequences were acquired before and during the hepatobiliary phase after Gx administration; MRI-PDFF was estimated in the MRS-PDFF voxel location. Linear regression parameters (slope, intercept, average bias, R 2) were calculated for MRS-PDFF as a function of MRI-PDFF for each MRI-M sequence (pre-Gx low-FA, pre-Gx high-FA, post-Gx low-FA, post-Gx high-FA) for all patients and for patients with MRS-PDFF <10%. Regression parameters were compared (Bonferroni-adjusted bootstrap-based tests). RESULTS: Three of the four MRI-M sequences (pre-Gx low-FA, post-Gx low-FA, post-Gx high-FA) provided relatively unbiased PDFF estimates overall and in the low-PDFF range, with regression slopes close to 1 and intercepts and biases close to zero. Pre-Gx high-FA MRI overestimated PDFF in proportion to MRS-PDFF, with slopes of 0.72 (overall) and 0.63 (low-PDFF range). Based on regression bias closest to 0, the post-Gx high-FA sequence was the most accurate overall and in the low-PDFF range. This sequence provided statistically significant improvements in at least two regression parameters compared to every other sequence. CONCLUSION: In patients undergoing Gx-enhanced MR exams, PDFF estimation accuracy of MRI-M can be improved by using high-FA on post-contrast images.
PURPOSE: The aim of the study was to determine in patients undergoing gadoxetate disodium (Gx)-enhanced MR exams whether proton density fat fraction (PDFF) estimation accuracy of magnitude-based multi-gradient-echo MRI (MRI-M) could be improved by using high flip angle (FA) on post-contrast images. MATERIALS AND METHODS: Thirty-one adults with known or suspected hepatic steatosis undergoing 3T clinical Gx-enhanced liver MRI were enrolled prospectively. MR spectroscopy (MRS), the reference standard, was performed before Gx to measure MRS-PDFF. Low (10°)- and high (50°)-flip angle (FA) MRI-M sequences were acquired before and during the hepatobiliary phase after Gx administration; MRI-PDFF was estimated in the MRS-PDFF voxel location. Linear regression parameters (slope, intercept, average bias, R 2) were calculated for MRS-PDFF as a function of MRI-PDFF for each MRI-M sequence (pre-Gx low-FA, pre-Gx high-FA, post-Gx low-FA, post-Gx high-FA) for all patients and for patients with MRS-PDFF <10%. Regression parameters were compared (Bonferroni-adjusted bootstrap-based tests). RESULTS: Three of the four MRI-M sequences (pre-Gx low-FA, post-Gx low-FA, post-Gx high-FA) provided relatively unbiased PDFF estimates overall and in the low-PDFF range, with regression slopes close to 1 and intercepts and biases close to zero. Pre-Gx high-FA MRI overestimated PDFF in proportion to MRS-PDFF, with slopes of 0.72 (overall) and 0.63 (low-PDFF range). Based on regression bias closest to 0, the post-Gx high-FA sequence was the most accurate overall and in the low-PDFF range. This sequence provided statistically significant improvements in at least two regression parameters compared to every other sequence. CONCLUSION: In patients undergoing Gx-enhanced MR exams, PDFF estimation accuracy of MRI-M can be improved by using high-FA on post-contrast images.
Authors: Charlie C Park; Catherine Hooker; Jonathan C Hooker; Emily Bass; William Haufe; Alexandra Schlein; Yesenia Covarrubias; Elhamy Heba; Mark Bydder; Tanya Wolfson; Anthony Gamst; Rohit Loomba; Jeffrey Schwimmer; Diego Hernando; Scott B Reeder; Michael Middleton; Claude B Sirlin; Gavin Hamilton Journal: J Magn Reson Imaging Date: 2018-04-29 Impact factor: 4.813
Authors: Takeshi Yokoo; Suraj D Serai; Ali Pirasteh; Mustafa R Bashir; Gavin Hamilton; Diego Hernando; Houchun H Hu; Holger Hetterich; Jens-Peter Kühn; Guido M Kukuk; Rohit Loomba; Michael S Middleton; Nancy A Obuchowski; Ji Soo Song; An Tang; Xinhuai Wu; Scott B Reeder; Claude B Sirlin Journal: Radiology Date: 2017-09-11 Impact factor: 11.105
Authors: Xiaoke Wang; Timothy J Colgan; Louis A Hinshaw; Nathan T Roberts; Leah C Henze Bancroft; Gavin Hamilton; Diego Hernando; Scott B Reeder Journal: Magn Reson Med Date: 2019-11-14 Impact factor: 4.668