Meeting Banner
Abstract #2136

Application of Deep Learning Reconstruction for Denoising of Compressed Sensing non-contrast coronary MRA images to achieve improved Diagnostic Confidence.

Yoko Kato1, Bharath Ambale-Venkatesh2, Yoshimori Kassai3, John Pitts4, Larry Kasuboski4, Jason Ortman1, Shelton Caruthers4, and Joao A.C. Lima1

1Cardiology, Johns Hopkins University, Baltimore, MD, United States, 2Radiology, Johns Hopkins University, Baltimore, MD, United States, 3Canon Medical Systems Corporation, Otawara, Japan, 4Canon Medical Research USA, Inc., Cleveland, OH, United States

Non-contrast Magnetic resonance coronary artery (MRCA) image acquisition has technical limitations of long acquisition time or reduced image resolution. We explore the use of a denoising approach with deep learning image reconstruction (dDLR) from k-space data. We investigate the effect of various levels of dDLR on Compressed Sensing non-contrast MRCA (CS-MRCA) images and optimize dDLR algorithms that achieve the best diagnostic confidence (DC) and a high signal-to-noise-ratio (SNR).

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

Join Here