Echo-planar magnetic resonance myocardial perfusion imaging: Parametric map analysis and comparison with thallium SPECT

• Published in: Journal of magnetic resonance imaging 2001-02, Vol.13 (2), p.192-200
• Main Authors: Panting, Jonathan R., Gatehouse, Peter D., Yang, Guang Z., Jerosch-Herold, Michael, Wilke, Norbert, Firmin, David N., Pennell, Dudley J.
• Format: Article
• Language: English
• Subjects: Aged; Coronary Angiography; Coronary Circulation - physiology; coronary disease; Coronary Disease - diagnosis; Coronary Disease - physiopathology; Echo-Planar Imaging; Exercise Test; Female; Humans; Image Enhancement; Image Processing, Computer-Assisted; magnetic resonance imaging; Male; Middle Aged; perfusion; Sensitivity and Specificity; Thallium Radioisotopes; Tomography, Emission-Computed, Single-Photon
• ISSN: 1053-1807; 1522-2586
• DOI: 10.1002/1522-2586(200102)13:2<192::AID-JMRI1029>3.0.CO;2-N

Magnetic resonance (MR) perfusion FLASH imaging has been used for assessing coronary artery disease (CAD). Echo‐planar MR techniques have advantages in speed and in making MR perfusion imaging results more clinically accessible through parametric maps, but have not been previously assessed. We implemented a spin‐echo, echo‐planar MR technique and applied it at rest and during adenosine stress in 26 patients with CAD and abnormal thallium single‐photon‐emission computed tomography (SPECT), and analyzed the results by using a newly developed parametric map analysis of time to peak, peak intensity, and slope of contrast washin. The results were compared with the results of conventional visual analysis of the perfusion cine series. For detecting abnormal coronary territories, MR and SPECT were comparable for sensitivity, specificity, and accuracy (thallium, 70%, 78%, and 73%; MR, 79% 83%, and 80%; P = NS). There was good agreement between thallium and MR during stress (kappa = 0.49), but defects were larger by MR (2.4 vs. 3.1 segments for slope; P < 0.01). Additional segments were detected at rest by MR (58 for slope vs. 25 for thallium), which correlated with areas that became abnormal with stress in the thallium (sensitivity, 100%; specificity, 63%). The parametric maps were easier and faster to interpret than review of the original first‐pass series of images (χ2 = 10.8; P < 0.04). The diagnostic performance of echo‐planar perfusion MR and SPECT was similar, and combining the results with parametric mapping was useful for interpretation and considerably improved data display for clinical interpretation. MR, however, was faster and yielded images of higher resolution with no radiation burden. In multislice mode, these new MR techniques may have clinical value. J. Magn. Reson. Imaging 2001;13:192–200. © 2001 Wiley‐Liss, Inc.