Abstract #0415

Multi-Resolution Registration and Segmentation for cardiac BOLD MRI

Ilkay Oksuz^1,2, Rohan Dharmakumar^3,4, and Sotirios A. Tsaftaris^2,5

¹Diagnostic Radiology, Yale University, New Haven, CT, United States, ²IMT Institute for Advanced Studies Lucca, Lucca, Italy, ³Biomedical Imaging Research Institute, Cedars Sinai Medical Center, Los Angeles, CA, United States, ⁴University of California, Los Angeles, CA, United States, ⁵The University of Edinburgh, Edinburgh, United Kingdom

Cardiac Phase-resolved Blood Oxygen-Level-Dependent (CP-BOLD) MRI is a new contrast and stress-free approach for detecting myocardial ischemia, that identifies the ischemic myocardium by examining changes in myocardial signal intensity patterns as a function of cardiac phase. But, these changes coupled with cardiac motion, challenge automated standard CINE MR myocardial segmentation and registration techniques resulting in a significant drop of segmentation and registration accuracy. We propose a dictionary learning based multi-resolution registration scheme for supervised learning and sparse representation of the myocardium. Our results show an improvement of 15% myocardial segmentation w.r.t. the state of the art.

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

Join Here