Comparison of semi-automated methods to quantify infarct size and area at risk by cardiovascular magnetic resonance imaging at 1.5T and 3.0T field strengths.

Khan, Jamal N ORCID: 0000-0002-3648-3548, Nazir, Sheraz A, Horsfield, Mark A, Singh, Anvesha ORCID: 0000-0003-1112-3973, Kanagala, Prathap ORCID: 0000-0001-6739-5607, Greenwood, John P, Gershlick, Anthony H and McCann, Gerry P ORCID: 0000-0002-5542-8448 (2015) Comparison of semi-automated methods to quantify infarct size and area at risk by cardiovascular magnetic resonance imaging at 1.5T and 3.0T field strengths. BMC research notes, 8 (1). 52-.

Text Comparison of infarct size.pdf - Published version Download (2MB) | Preview

Abstract

<h4>Background</h4>There is currently no gold standard technique for quantifying infarct size (IS) and ischaemic area-at-risk (AAR [oedema]) on late gadolinium enhancement imaging (LGE) and T2-weighted short tau inversion recovery imaging (T2w-STIR) respectively. This study aimed to compare the accuracy and reproducibility of IS and AAR quantification on LGE and T2w-STIR imaging using Otsu's Automated Technique (OAT) with currently used methods at 1.5T and 3.0T post acute ST-segment elevation myocardial infarction (STEMI).<h4>Methods</h4>Ten patients were assessed at 1.5T and 10 at 3.0T. IS was assessed on LGE using 5-8 standard-deviation thresholding (5-8SD), full-width half-maximum (FWHM) quantification and OAT. AAR was assessed on T2w-STIR using 2SD and OAT. Accuracy was assessed by comparison with manual quantification. Interobserver and intraobserver variabilities were assessed using Intraclass Correlation Coefficients and Bland-Altman analysis. IS using each technique was correlated with left ventricular ejection fraction (LVEF).<h4>Results</h4>FWHM and 8SD-derived IS closely correlated with manual assessment at both field strengths (1.5T: 18.3 ± 10.7% LV Mass [LVM] with FWHM, 17.7 ± 14.4% LVM with 8SD, 16.5 ± 10.3% LVM with manual quantification; 3.0T: 10.8 ± 8.2% LVM with FWHM, 11.4 ± 9.0% LVM with 8SD, 11.5 ± 9.0% LVM with manual quantification). 5SD and OAT overestimated IS at both field strengths. OAT, 2SD and manually quantified AAR closely correlated at 1.5T, but OAT overestimated AAR compared with manual assessment at 3.0T. IS and AAR derived by FWHM and OAT respectively had better reproducibility compared with manual and SD-based quantification. FWHM IS correlated strongest with LVEF.<h4>Conclusions</h4>FWHM quantification of IS is accurate, reproducible and correlates strongly with LVEF, whereas 5SD and OAT overestimate IS. OAT accurately assesses AAR at 1.5T and with excellent reproducibility. OAT overestimated AAR at 3.0T and thus cannot be recommended as the preferred method for AAR quantification at 3.0T.

Item Type:	Article
Uncontrolled Keywords:	Myocardium, Humans, Myocardial Infarction, Gadolinium DTPA, Contrast Media, Image Interpretation, Computer-Assisted, Observer Variation, Magnetic Resonance Imaging, Stroke Volume, Severity of Illness Index, Retrospective Studies, Ventricular Function, Left, Adult, Aged, Middle Aged
Divisions:	Faculty of Health and Life Sciences Faculty of Health and Life Sciences > Clinical Directorate
Depositing User:	Symplectic Admin
Date Deposited:	11 Jan 2022 15:53
Last Modified:	15 Mar 2024 20:16
DOI:	10.1186/s13104-015-1007-1
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3146265