Abstract #3091

Enhancing Mass Detection & Classification in Breast Tissue using Strain-Encoded (SENC) MRI

Ahmed Amr Harouni¹, Riham H. El Khouli², Jakir Hossain³, David A. Bluemke², Nael F. Osman⁴, Michael A. Jacobs⁵

¹Electrical & Computer Engineering, Johns Hopkins University, Baltimore , MD, United States; ²Radiology & Imaging Sciences, National Institute of Health, Bethesda, MD, United States; ³Electrical & Computer Engineering, Johns Hopkins University, Baltimore, MD, United States; ⁴Department of Radiology, Johns Hopkins University, Baltimore, MD, United States; ⁵Department of Radiology & Oncology, Johns Hopkins University School of Medicine, Baltimore, MD, United States

MRI has proven to have high sensitivity and moderate specificity in detecting breast cancer. Mechanical properties can increase specificity, by calculating the tissues stiffness. In this work, we demonstrate the use of strain-encoded MRI to measure strain that is inversely proportional to stiffness. We measure the compression and relaxation response of tissues. Phantom results show that using compression and relaxation complementary information with high CNR, we were able to detect and classify masses while ruling out image-artifacts. Moreover, Ex-vivo results show that SENC is able to detect masses that would be useful in clinical setting.