Integrating SAMER Retrospective Motion Correction into 3D Deep Learning Image Reconstruction for High-Quality, Fast, and Robust Brain MRI by Daniel Polak (Siemens Healthineers, Erlangen, Germany), et al.
Patient motion is one of the most common sources of image degradation in clinical neuroimaging, manifesting as blurring, ringing, or in severe cases ghosting/folding artifacts. Radiologists are often forced to either interpret images despite these artifacts – risking missed or misinterpreted pathologies – or to request costly and time-consuming repeat scans.
In pediatric patients, motion is typically more severe, often necessitating anesthesia, which increases both procedural risks and costs.
Motion artifacts also reduce the reliability of automated quantitative clinical tools, such as those used for brain morphometry and the identification or segmentation of hemorrhages, edema, and tumors. Such tools are increasingly important in the screening, monitoring, and treatment of neurodegenerative diseases such as Alzheimer’s.
Scout Accelerated Motion Estimation and Reduction (SAMER) is a retrospective motion correction technique for brain imaging that enables fast motion estimation and artifact correction without the need for external tracking hardware. Clinically evaluated in both adult and pediatric patient populations, SAMER has significantly reduced the number of non-diagnostic motion cases.
More recently, SAMER has been integrated into a deep learning-based image reconstruction framework to support highly accelerated, motion-robust 3D brain imaging.
In this article, the authors review the combined DL-SAMER technique and demonstrate its effectiveness in vivo using MPRAGE and SPACE acquisitions at R = 6 acceleration.
Shout out to the co-authors:
Dominik Nickel, Daniel Nicolas Splitthoff, Bryan Clifford, Yan Tu Huang, Wei-Ching Lo, Shohei Fujita, Susie Y. Huang, John Conklin, Lawrence L. Wald, Stephen Cauley
DL-SAMER: A new deep learning + retrospective motion correction method for fast, motion-robust 3D brain #MRI — effective even in challenging pediatric cases.
Check it out 👉 marketing.webassets.siemens-healthineers.com/7a1c195a6a39...
#NeuroSky #RadSky #MagnetomWorld
@harvardmed.bsky.social