Article Text


087 Incremental value of high-resolution cardiovascular magnetic resonance myocardial perfusion imaging in suspected coronary artery disease
  1. M Motwani1,
  2. N Maredia1,
  3. T A Fairbairn1,
  4. S Kozerke2,
  5. A Radjenovic1,
  6. J P Greenwood1,
  7. S Plein1
  1. 1University of Leeds, Leeds, UK
  2. 2University and ETH Zurich, Zurich, Switzerland


Introduction Although accelerated high-spatial-resolution cardiovascular magnetic resonance (CMR) perfusion imaging has recently been shown to be clinically feasible, there has not yet been a direct comparison with standard-resolution methods. We hypothesised that higher spatial resolution detects more subendocardial ischaemia and leads to greater diagnostic accuracy for the detection of angiographically defined CAD. This study compared the diagnostic accuracy of high-resolution and standard-resolution CMR perfusion imaging in patients with suspected coronary artery disease (CAD).

Methods A total of 111 patients with suspected CAD were prospectively recruited. All patients underwent two separate perfusion CMR studies on a 1.5 Tesla CMR scanner (Intera CV, Philips Healthcare, Best, the Netherlands), one with standard-resolution (2.5×2.5 mm in-plane resolution) and one with high-resolution (1.6×1.6 mm in-plane resolution) acquisition. High-resolution acquisition was facilitated by eightfold k-t broad linear speed up technique (BLAST) acceleration. Two observers visually graded perfusion in each myocardial segment on a 4-point scale. Segmental scores were summed to produce a perfusion score for each patient. All patients underwent invasive coronary angiography. Significant CAD was defined as a coronary artery stenosis of ≥50% diameter on quantitative coronary angiography.

Results CMR data were successfully obtained in 100 patients. A typical example is shown in Abstract 087 figure 1. In patients with CAD (n=70), more segments were determined to have subendocardial ischaemia with high-resolution acquisition than with standard-resolution acquisition (279 vs 108; p<0.001). High-resolution acquisition had a greater diagnostic accuracy than standard-resolution acquisition for identifying single-vessel disease (area under the curve [AUC]: 0.88 vs 0.73; p<0.001) or multi-vessel disease (AUC: 0.98 vs 0.91; p=0.002) and overall (AUC: 0.93 vs 0.83; p<0.001) (Abstract 087 figure 2).

Abstract 087 Figure 1

Case example: Proximal left anterior descending and right coronary artery disease standard-resolution shows possible perfusion defects (white arrows) in the basal anteroseptal, inferoseptal and inferior segments which are difficult to distinguish from dark-rim artifacts (A); and a more convincing perfusion defect in the apical septal segment (C). High-resoultion more clearly delineates perfusion defects in the basal infero-septal, basal inferior (D), mid-anteroseptal (E) and apical septal (F) segments. At high-resloution, the transmural extent of ischaemia can be clearly assessed (the basal and mid-ventricular defects appear subendocardial) and the perfusion defects are better delineated.

Abstract 087 Figure 2

Receiver-Operator Characteristic Curves. Standard and high-resolution perfusion CMR both had a high diagnostic accuracy for the detection of coronary artery disease but the high-resolution technique was superior. The arears under the curve were 0.83 (95% CI 0.75 to 0.91) for standard-resolution and 0.93 (95% CI 0.88 to 0.98) for high-resolution (p<0.001).

Conclusions Our study shows that high-resolution CMR perfusion imaging has greater diagnostic accuracy than standard-resolution acquisition for the detection of CAD in both single and multi-vessel disease patients and detects more subendocardial ischaemia.

  • Cardiovascular magnetic resonance
  • high-resolution
  • myocardial perfusion imaging

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.