Related Experiment Video
Updated: Apr 19, 2026
Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla
Published on: February 19, 2021
Impact of reconstruction type on late gadolinium enhancement scar quantification at 3-Tesla: Magnitude vs.
Aida Moafi1, Simran Shergill1, David Adlam2
1Department of Cardiovascular Sciences, University of Leicester, the National Institute for Health and Care Research Leicester Biomedical Research Centre and British Heart Foundation Centre of Research Excellence, Glenfield Hospital, Leicester, UK.
Background:
The identification of myocardial scar using late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) is well established, providing diagnostic and prognostic value in both ischemic and non-ischemic cardiomyopathies. Phase-sensitive inversion recovery (PSIR) provides more consistent image contrast compared with magnitude (MAG) reconstructions. However, the impact of reconstruction method on scar quantification remains uncertain.
Methods:
One-hundred and thirty-six patients from three prospectively enrolled cohorts with chronic myocardial infarction (>3 months post-infarction), representing distinct infarct phenotypes, underwent contrast-enhanced research CMR. LGE images were acquired using either a breath-hold, segmented spoiled gradient-echo or free-breathing, motion-corrected, balanced steady-state free precession sequence. Infarction was quantified on matched, short-axis MAG and PSIR reconstructions using the full-width at half-maximum method, with total infarct size expressed in grams (g) and as percentage of left ventricular mass. The primary analysis compared scar quantification between the two reconstruction methods.
Results:
Compared with MAG, PSIR demonstrated significantly higher median scar mass (13.7 g [6.0-25.7] vs. 8.9 g [4.4-16.6], p<0.001) and percentage of left ventricular mass (14.2% [7.5-21.9] vs. 10.6% [5.4-14.2], p<0.001). Inter-method agreement was moderate for both absolute and relative scar mass (intraclass correlation coefficients: 0.76 [95% CI: 0.33, 0.89] and 0.74 [95% CI: 0.20, 0.89], respectively), with a systematic, proportional difference on Bland-Altman analysis, with consistently higher scar mass by PSIR (mean bias: +5.01 ± 5.87 g; limits of agreement: -6.50 to +16.51 g and +4.54 ± 4.62%; limits of agreement: -4.51 to +13.60%).
Conclusion:
Quantification using PSIR yields higher infarct sizes compared with MAG LGE reconstructions. This emphasizes the importance of using a consistent reconstruction method when quantifying scar burden.
