Low-Dimensional Models in High-Dimensional MRI

Speaker: Michael Lustig, PhD

Date: March 28

Time: 1 pm

Location: 1240 Computer Sciences, 1210 W. Dayton St., Madison, WI


Abstract: Magnetic resonance imaging (MRI) is a powerful, ionizing-radiation-free medical imaging modality. The vast physical and physiological parameters, which MRI is sensitive to, makes it possible to visualize both structure and function in the body. However the prolonged time necessary to capture the information in this large parameter space remains a major limitation of this phenomenal modality.
In this talk I will cover the work that is done in our group to capture the rich, dynamical information in MRI. In particular I will focus on two techniques which provide comprehensive multi-contrast exams. The first leverages an unsupervised multi-scale low rank modeling, which enables low-dimensional representation for reconstructing dynamic 3D MRI data at unprecedented spatiotemporal resolution. This representation also enables compressed storage, which in combination with a stochastic optimization approach renders the reconstruction of 100’s of Gb of images feasible.
The 2nd technique, named T2 Shuffling leverages a low dimensional model to provide sharp, multi-contrast MR images in 3D, and in a single scan. T2 shuffling has been extensively used in the clinic. More recently it is used as a basis for a targeted short exam, that enables same-day appointments at 1/3 of the time and cost. This talk is based on the PhD work of Drs. Frank Ong, and Jonathan Tamir, and collaborations with Stanford Children’s hospital, UCSF and UW-Madison.


Recorded Lectures: Mp4 Lecture