Shigeo Okuda1, Yoshitake Yamada2, Akihiro Tanimoto3, Jun Fujita4, Motoaki Sano5, Keiichi Fukuda6, Sachio Kuribayashi7, Masahiro Jinzaki8, Atsushi Nozaki9, Peng Lai10. 1. Department of Diagnostic Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: okuda@rad.med.keio.ac.jp. 2. Department of Diagnostic Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: yamada@rad.med.keio.ac.jp. 3. Department of Diagnostic Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: t-mri@tt.rim.or.jp. 4. Department of Cardiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: jfujita@a6.keio.jp. 5. Department of Cardiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: msano@a8.keio.jp. 6. Department of Cardiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: kfukuda@a2.keio.jp. 7. Department of Diagnostic Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: skuribay@med.keio.ac.jp. 8. Department of Diagnostic Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan. Electronic address: Jinzaki@rad.med.keio.ac.jp. 9. MR Applications & Workflow, GE Healthcare, 4-7-127 Asahigaoka Hino-shi, Tokyo 191-8503, Japan. Electronic address: atsushi.nozaki@ge.com. 10. MR Applications & Workflow, GE Healthcare, 333 Ravenswood Ave, Bldg 307, Menlo Park, CA 94025, USA. Electronic address: peng.lai@ge.com.
Abstract
OBJECTIVE: Three-dimensional cardiac cine imaging has demonstrated promising clinical 1.5-Tesla results; however, its application to 3T scanners has been limited because of the higher sensitivity to off-resonance artifacts. The aim of this study was to apply 3D cardiac cine imaging during a single breath hold in clinical patients on a 3T scanner using the kat ARC (k- and adaptive-t auto-calibrating reconstruction for Cartesian sampling) technique and to evaluate the interchangeability between 2D and 3D cine imaging for cardiac functional analysis and detection of abnormalities in regional wall motion. METHODS: Following institutional review board approval, we obtained 2D cine images with an acceleration factor of two during multiple breath holds and 3D cine images with a net scan acceleration factor of 7.7 during a single breath hold in 20 patients using a 3T unit. Two readers independently evaluated the wall motion of the left ventricle (LV) using a 5-point scale, and the consistency in the detection of regional wall motion abnormality between 2D and 3D cine was analyzed by Cohen's kappa test. The LV volume was calculated at end-diastole and end-systole (LVEDV, LVESV); the ejection fraction (LVEF) and myocardial weight (LVmass) were also calculated. The relationship between functional parameters calculated for 2D and 3D cine images was analyzed using Pearson's correlation analysis. The bias and 95% limit of agreement (LA) were calculated using Bland-Altman plots. In addition, a qualitative evaluation of image quality was performed with regard to the myocardium-blood contrast, noise level and boundary definition. RESULTS: Despite slight degradation in image quality for 3D cine, excellent agreement was obtained for the detection of wall motion abnormalities between 2D and 3D cine images (κ=0.84 and 0.94 for each reader). Excellent correlations between the two imaging methods were shown for the evaluation of functional parameters (r>0.97). Slight differences in LVEDV, LVESV, LVEF and LVmass were observed, with average values of 1.6±8.9mL, -0.6±5.9mL, 1.4±3.6%, and 1.3±8.7g, respectively. CONCLUSIONS: Images obtained using the kat ARC 3D and conventional 2D cine techniques were equivalent in the detection of regional wall motion abnormalities and the evaluation of cardiac functional parameters.
OBJECTIVE: Three-dimensional cardiac cine imaging has demonstrated promising clinical 1.5-Tesla results; however, its application to 3T scanners has been limited because of the higher sensitivity to off-resonance artifacts. The aim of this study was to apply 3D cardiac cine imaging during a single breath hold in clinical patients on a 3T scanner using the kat ARC (k- and adaptive-t auto-calibrating reconstruction for Cartesian sampling) technique and to evaluate the interchangeability between 2D and 3D cine imaging for cardiac functional analysis and detection of abnormalities in regional wall motion. METHODS: Following institutional review board approval, we obtained 2D cine images with an acceleration factor of two during multiple breath holds and 3D cine images with a net scan acceleration factor of 7.7 during a single breath hold in 20 patients using a 3T unit. Two readers independently evaluated the wall motion of the left ventricle (LV) using a 5-point scale, and the consistency in the detection of regional wall motion abnormality between 2D and 3D cine was analyzed by Cohen's kappa test. The LV volume was calculated at end-diastole and end-systole (LVEDV, LVESV); the ejection fraction (LVEF) and myocardial weight (LVmass) were also calculated. The relationship between functional parameters calculated for 2D and 3D cine images was analyzed using Pearson's correlation analysis. The bias and 95% limit of agreement (LA) were calculated using Bland-Altman plots. In addition, a qualitative evaluation of image quality was performed with regard to the myocardium-blood contrast, noise level and boundary definition. RESULTS: Despite slight degradation in image quality for 3D cine, excellent agreement was obtained for the detection of wall motion abnormalities between 2D and 3D cine images (κ=0.84 and 0.94 for each reader). Excellent correlations between the two imaging methods were shown for the evaluation of functional parameters (r>0.97). Slight differences in LVEDV, LVESV, LVEF and LVmass were observed, with average values of 1.6±8.9mL, -0.6±5.9mL, 1.4±3.6%, and 1.3±8.7g, respectively. CONCLUSIONS: Images obtained using the kat ARC 3D and conventional 2D cine techniques were equivalent in the detection of regional wall motion abnormalities and the evaluation of cardiac functional parameters.