Meeting Banner
Abstract #2125

Transfer Learning Based Automated Myocardial T2 and Extracellular Volume Quantification

Yanjie Zhu1,2, Ahmed S. Fahmy2, Chong Duan2, and Reza Nezafat2

1Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, 2Department of Medicine (Cardiovascular Division), Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, United States

Manual analysis of myocardial tissue mapping is time consuming. Deep learning has a potential to facilitate the analysis but requires big training datasets. In this study, a deep fully convolutional neural network, trained using native T1 mapping dataset, is used for T2 and extracellular volume (ECV) quantification based on transfer learning. We prospectively acquired T2 (401 patients) and ECV maps (381 patients) to access the network performance. Compared with the manually analyzed reference values, the transfer learning-based automated analysis platform shows good performance for myocardial T2 and ECV mapping. The platform has potential to fully automate myocardial tissue mapping.

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

Join Here