Agreement between a novel echocardiographic “Scar Imaging” technique and cardiac magnetic resonance for quantifying infarct size in acute myocardial infarction

Background Infarct size is a key determinant of left ventricular remodeling and prognosis after acute myocardial infarction (AMI). Cardiac magnetic resonance (CMR) with late gadolinium enhancement (LGE) is the reference standard for infarct size quantification; however, its availability and contraindications limit routine clinical use. We developed a novel echocardiographic technique based on three-dimensional speckle-tracking area strain, Scar Imaging, which enables bedside visualization of infarct myocardium and estimation of its size. Purpose To evaluate the inter-modality agreement between Scar Imaging and CMR for quantifying infarct size in AMI. Methods We retrospectively analyzed AMI patients who underwent percutaneous coronary intervention (PCI) and received both Scar Imaging and CMR during the index hospitalization between June 2022 and December 2024. On CMR, infarct size was quantified from LGE as the percentage of left ventricular (LV) myocardium, using automated software-based analysis. Scar Imaging identifies LV regions with markedly reduced deformation as infarct according to a pre-specified area strain threshold and reports infarct size (Figure 1). Inter-modality agreement was assessed using Pearson correlation, Bland–Altman analysis, mean absolute error (MAE), intraclass correlation coefficient (ICC; two-way random-effects model, absolute agreement), and Deming regression. Results In 40 patients (median age 72 years; 33 men [82.5%]), the median time from PCI to CMR and to echocardiography was 7 and 8 days, respectively. Median CMR-LGE infarct size was 18.0% (IQR 8.69–25.46). Infarct size by Scar Imaging showed a strong correlation with CMR (r=0.755, p<0.01). Bland–Altman analysis showed a small mean bias of −2.19%, with 95% limits of agreement from −19.91% to 15.52% (Figure 2). MAE was 7.56%, and ICC was 0.749, indicating good absolute agreement. Deming regression showed near 1:1 agreement (slope 0.999; intercept −2.18) and suggested no proportional bias. Conclusion Scar Imaging demonstrated good agreement with CMR for infarct size quantification in AMI, with minimal systematic bias and acceptable limits of agreement. It may serve as a practical alternative for infarct size assessment when CMR is unavailable or contraindicated.