OSE Seminar by Dr. Steen Moller on Leveraging Thermal noise in MRI
Departmental News
Posted: March 28, 2023
Speaker: Dr. Steen Moller of the Center for Magnetic Resonance Research, University of Minnesota (“CMRR”).
Date: Thursday, March 30, 2023
Time: 12:15 PM - 1:15 PM
Location: CHTM, Room 101, and via Zoom.
Abstract:
Magnetic resonance imaging (MRI) has become one of the most powerful tools for investigating the human brain, and functional magnetic resonance imaging (fMRI) is widely used for probing the living brain. However, virtually all fMRI studies have relatively poor signal-to-noise ratio (SNR). Today's talk will be divided into 3 parts. In the first I will present the mathematical and physics grounding for how fast imaging is performed; then present the mathematical and theoretical grounding of NORDIC PCA for lower the impact of thermal noise and then demonstrate the benefits of integrating Deep learning reconstructions with conventional denoising techniques to push fMRI beyond conventional resolutions.
Biography:
Dr. Steen Moeller is a Research Associate Professor in Radiology at the Center for Magnetic Resonance Research, University of Minnesota (“CMRR”). After obtaining a Master’s Degree (Cand.Scient.) in Mathematics and Physics from the University of Aalborg, Denmark in 1997, Dr. Moeller conducted research on inverse problems for electrical impedance tomography and was awarded the Ph.D. degree in 2002 from the Department of Mathematical Sciences, University of Aalborg. He joined CMRR as a post-doctoral associate for Dr. Kamil Ugurbil in 2003 and focused on mathematical algorithms for accelerated image reconstructions at ultra high fields utilizing multi-channel arrays. His work was focused on the implications for functional MRI and hardware design criteria. In 2006, Dr. Moeller joined Dr Michael Garwood’s group for investigating and developed mathematical tools for the SWIFT (Sweep Imaging with Fourier Transformation) technique developed at CMRR. The technologies developed were licensed by Steady State Imaging, LLC and subsequently licensed by General Electric. Since 2010, Dr. Moeller has been involved in the technology development for the Human Connectome Project and has developed robust mathematical techniques for fast image reconstruction and new novel Deep learning and Denoising technology.
