Accurate and objective infarct sizing by contrast-enhanced magnetic resonance imaging in a canine myocardial infarction model
Luciano C. Amado, MD*,
Bernhard L. Gerber, MD*,
Sandeep N. Gupta, PhD ,
Dan W. Rettmann, BS,
Gilberto Szarf, MD ,
Robert Schock, PhD ,
Khurram Nasir, MD, MPH*,
Dara L. Kraitchman, VMD, PhD and
João A.C. Lima, MD*,*
* Cardiology
Radiology, Johns Hopkins University, School of Medicine, Baltimore, Maryland
GE Medical Systems, Waukesha, Wisconsin
Datascope Corporation, Mahwah, New Jersey
View larger version (50K):
[in a new window]
|
Figure 1 Delayed enhancement images and semi-automatic techniques using different criteria to quantify myocardial infarct (MI) extension: 1) the full-width at half-maximum (FWHM) criterion—an initial region is determined to grow to include all pixels with signal intensity (SI) >50% of a user selected point. The maximum signal intensity (MX) inside this initial region is then determined, and the final MI extent is defined as the area presenting with a signal intensity 50% above the maximum of the initial region (MI = MX * 0.5). 2) n-standard deviation (SD) criteria—the mean ± SD of remote signal intensity (RI) under the yellow circle was determined. The extent of MI was defined as areas presenting with signal intensities above the mean value from the remote (MI = RI + SD). Measurements were done using 1, 2, 3, 4, 5, and 6 SD above the mean value.
|
|
View larger version (20K):
[in a new window]
|
Figure 2 Infarct size measured by delayed contrast-enhanced magnetic resonance imaging using different criteria. (A) The best full-width at half-maximum (FWHM) and worst (6 SD) correlations against triphenyltetrazolium chloride (TTC) pathology. (B) Bland-Altman analyses show the level of agreement between infarct size measured by the different criteria against TTC (FWHM and 1 SD for best and worst, respectively). (C) Bland-Altman analyses of interobserver reproducibility studies obtained from visual and semi-automatic criteria methods.
|
|
View larger version (14K):
[in a new window]
|
Figure 3 Comparison of infarct size measured from short-axis images by delayed contrast-enhanced magnetic resonance imaging (MRI) using different techniques. The arrow demonstrates the border zone, areas with partial enhancement on the edge of the infarct, which may lead to visual ambiguity in assessing the delayed hyperenhanced area. The infarcted region was delineated by the different techniques using different criteria, as shown by contours in red: 1 = visual; 2 = standard deviation (1 SD); and 3 = full-width at half-maximum (FWHM). Significant variability among segmentation methods is noted, especially related to the inclusion of the border zone on the analysis. The MRI infarct size was compared with postmortem measurements (triphenyltetrazolium chloride [TTC] pathology).
|
|
View larger version (111K):
[in a new window]
|
Figure 4 Differences in infarct size over time. This figure depicts delayed enhancement short-axis images from three different animals. Images were obtained every 6 min after contrast injection until 30 min after contrast. The presence, location, and size of the hyperenhanced region were similar for all time points. There were high levels of agreement in measurement of myocardial infarct size between the first set of images at 6 min and all others up to 30 min after contrast injection.
|
|
View larger version (13K):
[in a new window]
|
Figure 5 Bland-Altman analysis demonstrates no important differences in infarct size measurements over time (bias of –0.2% of left ventricle).
|
|
|
