Accuracy of Cardiac Magnetic Resonance of Absolute Myocardial Blood Flow With a High-Field SystemComparison With Conventional Field Strength
Timothy F. Christian, MD*,*,
Stephen P. Bell, BS*,
Lawrence Whitesell, BS ,
Michael Jerosch-Herold, PhD
* University of Vermont College of Medicine, Burlington, Vermont
University of Wisconsin College of Medicine, Madison, Wisconsin
Brigham and Women's Hospital, Department of Radiology, Harvard University, Boston, Massachusetts

View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1 Comparison of Perfusion Images by Field Strength
Top panels: short-axis images in 2 animals with similar regional hyperemia and control myocardial blood flow (MBF) values at 1.5-T (left) and 3.0-T (right). White arrows depict the adenosine zone; control zones are indicated by black arrows. Bottom panels: time intensity curves for the left ventricular cavity (arterial input function, left panel), myocardial control zone (center panel), and adenosine zone (right panel); zones from the images above. RV = right ventricle; SI = signal intensity.
|
|

View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2 Comparison of Agreement of Perfusion Measures by Field Strength
Scatter-plots and linear correlations between absolute myocardial blood flow (MBF) by fluorescent colored microspheres and Fermi-function deconvolution values derived from cardiac magnetic resonance (CMR) acquired at 1.5-T (left) and 3.0-T (right). Both field strengths produced strong correlations, but there was a small degree of overestimation of values at 1.5-T. Both correlations coefficients exceeded 0.90, and the agreements for both were highly significant. The yellow line represents identity.
|
|

View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3 Perfusion Measures by Field Strength in Ischemic Zones
Scatter-plots with linear fit for myocardial blood flow (MBF) values by microsphere technique <0.50 ml/min/g (ischemic flow). There was a stronger correlation at 3.0-T for ischemic myocardium, perhaps as a result of the improved signal, but the number of data points were limited for formal statistical comparison. The truncated range of the x-axis, in comparison with the overall group, should be noted when evaluating this analysis. The yellow line represents identity.
|
|

View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4 Bland-Altman Analysis by Field Strength
Bland Altman plots from the correlations in Figure 2. This type of analysis plots the average between the microsphere value and the cardiac magnetic resonance (CMR) value on the x-axis against the difference between these measures on the y-axis. The dotted yellow lines denote 0 difference between measures, and the dashed brown lines reflect the mean of the difference of the measures, which are slightly overestimated by CMR at 1.5-T. Solid black lines depict the 95% confidence intervals of the difference, and these are wider at 1.5-T (0.49 vs. 0.68 ml/min/g for 3.0- and 1.5-T, respectively), suggesting a slightly less close agreement at 1.5-T. MBF = myocardial blood flow.
|
|

View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5 Difference Between CMR and Microsphere Measures of Absolute MBF by Magnet Strength
Boxplots examining the difference between MR-derived measures of absolute myocardial blood flow (MBF) and those measured by the microsphere technique as a function of magnet strength. Mean (dotted line), SEM (boxes), and SD (whiskers) of the difference between cardiac magnetic resonance (CMR) Fermi-function deconvolution estimates of MBF and absolute MBF by field strength are shown. These values show a small but significant difference in the mean of the magnitude of agreement, but the spread of the differences are very similar.
|
|
|