Evaluation of accordance of magnetic resonance volumetric and flow measurements in determining ventricular stroke volume in cardiac patients

• Published in: Acta radiologica (1987) Vol. 49; no. 5; p. 530
• Main Authors: Jeltsch, M, Ranft, S, Klass, O, Aschoff, A J, Hoffmann, M H K
• Format: Journal Article
• Language: English
• Published: England 01.06.2008
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Blood Flow Velocity; Child; Cohort Studies; Female; Heart Diseases - physiopathology; Heart Ventricles - physiopathology; Humans; Image Processing, Computer-Assisted; Imaging, Three-Dimensional - methods; Magnetic Resonance Imaging; Male; Middle Aged; Reproducibility of Results; Stroke Volume; Time Factors
• ISSN: 1600-0455
• DOI: 10.1080/02841850801998847

Cardiovascular magnetic resonance imaging (CMR) has become an established noninvasive method for evaluating ventricular function utilizing three-dimensional volumetry. Postprocessing of volumetric measurements is still tedious and time consuming. Stroke volumes obtained by flow quantification across the aortic root or pulmonary trunk could be utilized to increase both speed of workflow and accuracy. To assess accuracy of stroke volume quantification using MR volumetric imaging compared to flow quantification in patients with various cardiac diseases. Strategies for the augmentation of accuracy in clinical routine were deduced. 78 patients with various cardiac diseases-excluding intra- or extracardiac shunts, regurgitant valvular defects, or heart rhythm disturbance-underwent cardiac function analysis with flow measurements across the aortic root and cine imaging of the left ventricle. Forty-six patients additionally underwent flow measurements in the pulmonary trunk and cine imaging of the right ventricle. Left ventricular stroke volume (LVSV) and stroke volume of the aortic root (SVAo) correlated with r=0.97, and Bland-Altman analysis showed a mean difference of 0.11 ml and a standard error of estimation (SEE) of 4.31 ml. Ninety-two percent of the data were within the 95% limits of agreement. Right ventricular stroke volume (RVSV) and stroke volume of the pulmonary trunk (SVP) correlated with a factor of r=0.86, and mean difference in the Bland-Altman analysis was fixed at -2.62 ml (SEE 8.47 ml). For RVSV and SVP, we calculated r=0.82, and Bland-Altman analysis revealed a mean difference of 1.27 ml (SEE 9.89 ml). LVSV and RVSV correlated closely, with r=0.91 and a mean difference of 2.79 ml (SEE 7.17 ml). SVAo and SVP correlated with r=0.95 and a mean difference of 0.50 ml (SEE 5.56 ml). Flow quantification can be used as a guidance tool, providing accurate and reproducible stroke volumes of both ventricles. Combining both offers a highly accurate tool to gauge ventricular function in a routine clinical setting, increasing workflow speed.