Automated truncation method for myocardial T2* measurement in thalassemia

Abstract

Purpose: To propose an automated truncation method for myocardial T2* measurement and evaluate this method on a large population of patients with iron loading in the heart and scanned at multiple magnetic resonance imaging (MRI) centers.

Materials and methods: A total of 550 thalassemia patients were scanned at 20 international centers using a variety of MR scanners (Siemens, Philips, or GE). A single mid-ventricular short axis slice was imaged. All patient data were anonymized before the T2* were measured by expert observers using standard techniques. These same datasets were then retrospectively processed using the proposed automated truncation method by another independent observer and the resulting T2* measurements were compared with those of expert readings.

Results: The T2* measurements using the automated method showed good agreement with those measured by expert observers using standard techniques (P = 0.95) with a low coefficient of variation (1.6%).

Conclusion: This study demonstrates feasibility and good reproducibility of a new automated truncation method for myocardial T2* measurement. This approach simplifies the overall analysis and can be easily incorporated into T2* analysis software to facilitate further development of a fully automated myocardial tissue iron quantification.