BACKGROUND: Left ventricular (LV), right ventricular (RV), and left atrial (LA) volumes and functions contain important prognostic information in ischemic heart disease. Because multidetector computed tomography (MDCT) has high spatial resolution, this method may be optimal to obtain this information. OBJECTIVE: We evaluated automated assessment for MDCT, by comparing it with cardiac magnetic resonance (CMR). METHODS: Fifty-three patients with previous myocardial infarction were scanned with 1.5 Tesla CMR and 64-slice MDCT. End-diastolic volume, end-systolic volume, stroke volume, and ejection fraction (EF) were assessed for the left and right ventricle with automatic MDCT software and manual CMR software. LV myocardial mass and cyclic changes in LA volume were derived. RESULTS: The mean age of patients was 61 ± 10 years, 40 (75%) were men. Automated MDCT segmentation was possible in all but 2 patients. The average duration of image processing was 21 ± 4 minutes by CMR and 11 ± 4 minutes by MDCT. Bland-Altman plots showed good agreement between MDCT and CMR with only small bias. LVEF by CMR was 56% ± 10% and by MDCT 61% ± 11%, mean difference of -5% (limits of agreement, -18% to 8%), and P < 0.001. RVEF by CMR was 60% ± 5% and by MDCT 56% ± 8%, mean difference of 5% (limits of agreement, -10% to 20%), and P < 0.001. LA fractional change by CMR was 49% ± 9% and by MDCT 45% ± 9%, mean difference of 4% (limits of agreement, -12% to 20%), and P ≤ 0.001. CONCLUSION: LV, RV, and LA volumes and functions may be evaluated fast and reliably with the use of automated assessment and cardiac MDCT, with good agreement to CMR. Accurate assessment of cardiac chambers with MDCT appears possible in clinical practice.
BACKGROUND: Left ventricular (LV), right ventricular (RV), and left atrial (LA) volumes and functions contain important prognostic information in ischemic heart disease. Because multidetector computed tomography (MDCT) has high spatial resolution, this method may be optimal to obtain this information. OBJECTIVE: We evaluated automated assessment for MDCT, by comparing it with cardiac magnetic resonance (CMR). METHODS: Fifty-three patients with previous myocardial infarction were scanned with 1.5 Tesla CMR and 64-slice MDCT. End-diastolic volume, end-systolic volume, stroke volume, and ejection fraction (EF) were assessed for the left and right ventricle with automatic MDCT software and manual CMR software. LV myocardial mass and cyclic changes in LA volume were derived. RESULTS: The mean age of patients was 61 ± 10 years, 40 (75%) were men. Automated MDCT segmentation was possible in all but 2 patients. The average duration of image processing was 21 ± 4 minutes by CMR and 11 ± 4 minutes by MDCT. Bland-Altman plots showed good agreement between MDCT and CMR with only small bias. LVEF by CMR was 56% ± 10% and by MDCT 61% ± 11%, mean difference of -5% (limits of agreement, -18% to 8%), and P < 0.001. RVEF by CMR was 60% ± 5% and by MDCT 56% ± 8%, mean difference of 5% (limits of agreement, -10% to 20%), and P < 0.001. LA fractional change by CMR was 49% ± 9% and by MDCT 45% ± 9%, mean difference of 4% (limits of agreement, -12% to 20%), and P ≤ 0.001. CONCLUSION: LV, RV, and LA volumes and functions may be evaluated fast and reliably with the use of automated assessment and cardiac MDCT, with good agreement to CMR. Accurate assessment of cardiac chambers with MDCT appears possible in clinical practice.
Authors: Louise Hindsø; Andreas Fuchs; Jørgen Tobias Kühl; Emma Julia P Nilsson; Per Ejlstrup Sigvardsen; Lars Køber; Børge G Nordestgaard; Klaus Fuglsang Kofoed Journal: Int J Cardiovasc Imaging Date: 2016-11-14 Impact factor: 2.357
Authors: Henrik Øder Hjortkjær; Tonny Jensen; Klaus F Kofoed; Ulrik M Mogensen; Per Ejlstrup Sigvardsen; Lars Køber; Karen Lisa Hilsted; Helle Corinth; Simone Theilade; Jannik Hilsted Journal: BMJ Open Date: 2016-12-05 Impact factor: 2.692