Abstract #2009

Automatic Detection and Quantification of Myocardial Scar in Patients with Prior Myocardial Infarction at 3T without Contrast Agents

Xinheng Zhang^1,2, Hsin-Jung Yang¹, Guan Wang^1,3, Ivan Cokic¹, Qi Yang¹, Sotirios Tsaftaris⁴, and Rohan Dharmakumar¹

¹Biomedical Imaging Research Institute, Cedars-Sinai Medical Center, Los Angeles, CA, United States, ²Department of Bioengineering, University of California, Los Angeles, Los Angeles, CA, United States, ³Department of Radiology, The First Affiliated Hospital of China Medical University, Shenyang, China, ⁴Institute of Digital Communications, University of Edinburgh, Edinburgh, United Kingdom

Native T1 maps at 3T has the capacity to accurately characterize chronic myocardial infarction (MI) territories, however it requires accurate identification of remote myocardium which in some cases is limited by image contrast between infarcted and remote myocardial territories. To overcome this limitation, we evaluated multiple automatic segmentation algorithms. Native T1 maps acquired in chronic MI patients were segmented using Gaussian Mixture Model, Otsu’s and K-means methods. K-means approach showed the best performance when compared to LGE. We conclude that K-means approach can accurately delineate MI territories.

This abstract and the presentation materials are available to members only; a login is required.

Join Here