Eve Piekarski1,2, Teodora Chitiboi1,2, Rebecca Ramb1,2, Larry A Latson3, Puneet Bhatla3, Li Feng1,2, Leon Axel4,5,6. 1. Center for Advanced Imaging Innovation and Research (CAI2R), Department of Radiology, New York University School of Medicine, 660 First Ave., New York, NY, USA. 2. Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University School of Medicine, 660 First Ave., New York, NY, USA. 3. Department of Radiology, New York University Langone Medical Center, 650 First Ave., New York, NY, USA. 4. Center for Advanced Imaging Innovation and Research (CAI2R), Department of Radiology, New York University School of Medicine, 660 First Ave., New York, NY, USA. leon.axel@nyumc.org. 5. Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University School of Medicine, 660 First Ave., New York, NY, USA. leon.axel@nyumc.org. 6. Department of Radiology, New York University Langone Medical Center, 650 First Ave., New York, NY, USA. leon.axel@nyumc.org.
Abstract
OBJECTIVES: Residual respiratory motion degrades image quality in conventional cardiac cine MRI (CCMRI). We evaluated whether a free-breathing (FB) radial imaging CCMRI sequence with compressed sensing reconstruction [extradimensional (e.g. cardiac and respiratory phases) golden-angle radial sparse parallel, or XD-GRASP] could provide better image quality than a conventional Cartesian breath-held (BH) sequence in an unselected population of patients undergoing clinical CCMRI. MATERIALS AND METHODS: One hundred one patients who underwent BH and FB imaging in a midventricular short-axis plane at a matching location were included. Visual and quantitative image analysis was performed by two blinded experienced readers, using a five-point qualitative scale to score overall image quality and visual signal-to-noise ratio (SNR) grade, with measures of noise and sharpness. End-diastolic and end-systolic left ventricular areas were also measured and compared for both BH and FB images. RESULTS: Image quality was generally better with the BH cines (overall quality grade for BH vs FB images 4 vs 2.9, p < 0.001; noise 0.06 vs 0.08 p < 0.001; SNR grade 4.1 vs 3, p < 0.001), except for sharpness (p = 0.48). There were no significant differences between BH and FB images regarding end-diastolic or end-systolic areas (p = 0.35 and p = 0.12). Eighteen of the 101 patients had poor BH image quality (grade 1 or 2). In this subgroup, the quality of the FB images was better (p = 0.0032), as was the SNR grade (p = 0.003), but there were no significant differences regarding noise and sharpness (p = 0.45 and p = 0.47). CONCLUSION: Although FB XD-GRASP CCMRI was visually inferior to conventional BH CCMRI in general, it provided improved image quality in the subgroup of patients with respiratory-motion-induced artifacts on BH images.
OBJECTIVES: Residual respiratory motion degrades image quality in conventional cardiac cine MRI (CCMRI). We evaluated whether a free-breathing (FB) radial imaging CCMRI sequence with compressed sensing reconstruction [extradimensional (e.g. cardiac and respiratory phases) golden-angle radial sparse parallel, or XD-GRASP] could provide better image quality than a conventional Cartesian breath-held (BH) sequence in an unselected population of patients undergoing clinical CCMRI. MATERIALS AND METHODS: One hundred one patients who underwent BH and FB imaging in a midventricular short-axis plane at a matching location were included. Visual and quantitative image analysis was performed by two blinded experienced readers, using a five-point qualitative scale to score overall image quality and visual signal-to-noise ratio (SNR) grade, with measures of noise and sharpness. End-diastolic and end-systolic left ventricular areas were also measured and compared for both BH and FB images. RESULTS: Image quality was generally better with the BH cines (overall quality grade for BH vs FB images 4 vs 2.9, p < 0.001; noise 0.06 vs 0.08 p < 0.001; SNR grade 4.1 vs 3, p < 0.001), except for sharpness (p = 0.48). There were no significant differences between BH and FB images regarding end-diastolic or end-systolic areas (p = 0.35 and p = 0.12). Eighteen of the 101 patients had poor BH image quality (grade 1 or 2). In this subgroup, the quality of the FB images was better (p = 0.0032), as was the SNR grade (p = 0.003), but there were no significant differences regarding noise and sharpness (p = 0.45 and p = 0.47). CONCLUSION: Although FB XD-GRASP CCMRI was visually inferior to conventional BH CCMRI in general, it provided improved image quality in the subgroup of patients with respiratory-motion-induced artifacts on BH images.
Entities:
Keywords:
Cardiac imaging techniques; Cardiac-gated imaging techniques; Cine magnetic resonance imaging; Image reconstruction; Magnetic resonance imaging
Authors: Yiqun Xue; Jiangsheng Yu; Hyun Seon Kang; Sarah Englander; Mark A Rosen; Hee Kwon Song Journal: Magn Reson Med Date: 2011-06-07 Impact factor: 4.668