ISMRM22 - Extending Cardiovascular MR: Learning CMR Image Reconstruction

Joint Annual Meeting ISMRM-ESMRMB & ISMRT 31st Annual Meeting • 07-12 May 2022 • London, UK

Extending Cardiovascular MR: Learning CMR Image Reconstruction

Sunrise Course

ORGANIZERS: Sila Kurugol, Tobias Wech, Christopher François, Ruud van Heeswijk

Tuesday, 10 May 2022

ICC Capital Suite 8-9

08:00 - 09:00

Moderators: Jesse Hamilton

Skill Level: Basic to Intermediate

Session Number: S-Tu-02

Session Number: S-Tu-02

Overview
These four sunrise sessions will demonstrate several applications into which cardiovascular MR has recently expanded. This includes patient groups such adult congenital complications as well as technical advances such as artificial intelligence, low-field scanners, and advanced motion correction.

Target Audience
Clinicians, physicians, physicists, and engineers interested in the latest advances in cardiovascular MR.

Educational Objectives
As a result of attending this course, participants should be able to:
- Describe how CMR can be used in patients with congenital heart defects;
- Describe how artificial intelligence can be applied to CMR;
- Describe the advantages and disadvantages of low-field CMR vs. high-field CMR; and
- Describe the effects of motion and the correction thereof.

	08:00		Motion-Corrected Reconstruction in CMR View the Presentation Christopher Roy Image quality in cardiovascular magnetic resonance imaging (CMR) is highly dependent on the methods used to compensate for both cardiac and respiratory motion. Additionally, in non-cooperative patients, especially in pediatrics, bulk patient movement can significantly degrade image quality. As a result, motion correction strategies for CMR make up a complex and on-going field of research. This talk will provide an overview of both well established and emerging directions for technologies that can provide high quality images of the heart despite the presence of motion during the examination.
	08:30		Physics-Based Deep Learning for CMR Image Reconstruction View the Presentation Mehmet Akcakaya Deep learning (DL) techniques have emerged as a powerful reconstruction approach for high-quality accelerated MRI. Among these, physics-based DL reconstruction approaches, which incorporate the MRI encoding operator to solve a regularized least squares problem, have gained interest due to its improved generalization abilities. Our purpose is to look at physics-based DL methods in the context of CMR reconstruction.