Cardiac magnetic resonance imaging detection of intramyocardial hemorrhage in patients with ST-elevated myocardial infarction: comparison between susceptibility-weighted imaging and T1/T2 mapping techniques

• Published in: Quantitative imaging in medicine and surgery Vol. 14; no. 1; pp. 476 - 488
• Main Authors: Wen, Jinyang, Qiao, Jinhan, Tang, Yuanyuan, Zhao, Yun, Yang, Zhaoxia, Wang, Luyun, Tao, Xinwei, Zhou, Xiaoyue, Xia, Liming, Tang, Dazhong, Huang, Lu
• Format: Journal Article
• Language: English
• Published: China AME Publishing Company 03.01.2024
• Subjects: Original; intramyocardial hemorrhage (IMH); T2 mapping; cardiac magnetic resonance (CMR); Susceptibility-weighted imaging (SWI); T1 mapping
• ISSN: 2223-4292; 2223-4306
• DOI: 10.21037/qims-23-591

Susceptibility-weighted imaging (SWI) and T1/T2 mapping can be used to detect reperfusion intramyocardial hemorrhage (IMH) in ST-segment elevation myocardial infarction (STEMI) patients. However, the sensitivity and accuracy of the SWI and T1/T2 mapping sequences were not systematically compared. The study aimed to evaluate image quality and diagnostic performance of SWI in patients with IMH, compared with T1/T2 mapping. A prospective study was conducted on consecutive acute STEMI patients who were recruited from January to July 2022. Within 2-6 days after reperfusion treatment, all patients underwent a 3T cardiac magnetic resonance (CMR) examination, including T2-weighted short-tau inversion recovery (T2W-STIR), T1/T2 mapping, and SWI. A total of 36 patients [age, 56.50±17.25 years; males, 83.33% (30/36)] were enrolled. The relative infarct-remote myocardium signal intensity ratio (SIinfarct-remote) and contrast-to-noise ratio (CNR) were calculated for each patient on T1/T2 mapping and SWI, and the difference between relative signal intensity-to-noise ratio (rSNR) in the IMH (rSNRIMH) was measured for IMH patients on T1/T2 mapping and SWI. SIinfarct-remote, CNR, and rSNRIMH were compared among the three sequences. Receiver operating characteristic (ROC) analyses were used to evaluate the diagnostic performance of three sequences by SIinfarct-remote and visual assessment. A total of 26 (72.22%) patients had IMH. Quantitatively, the SIinfarct-remote of three sequences had excellent diagnostic performance for detecting IMH [SWI area under the curve (AUC) =1.000, 95% confidence interval (CI): 1.000-1.000 . T1 mapping AUC =0.954, 95% CI: 0.885-1.000 . T2 mapping AUC =0.985, 95% CI: 0.955-1.000; SWI . T1 mapping, P=0.300; SWI . T2 mapping, P=0.188; T1 mapping . T2 mapping, P=0.302). Qualitatively, three sequences had similar performance on detecting IMH (SWI AUC =0.895, 95% CI: 0.784-1.000; T1 mapping AUC =0.835, 95% CI: 0.711-0.958; and T2 mapping AUC =0.855, 95% CI: 0.735-0.974; SWI . T1 mapping, P=0.172; SWI . T2 mapping, P=0.317; T1 mapping . T2 mapping, P=0.710). The rSNRIMH was highest in T1 mapping, followed by T2 mapping and SWI, but SWI had the highest CNR. SWI, as well as T1/T2 mapping, is a feasible and accurate approach for clinical diagnosis of IMH with excellent performance.