Dynamic magnetic resonance imaging (MRI) has important clinical applications (e.g., cardiac disease diagnosis, neurological behavior studies). It captures an object in motion by acquiring data across time, then reconstructing a sequence of images from them. Goals in advancing dynamic MRI include fast imaging with high spatial and temporal resolution. With the development of fast data acquisition methods, we also desire efficient image reconstruction techniques.

This talk covers efficient dynamic MRI reconstruction using handcrafted models, to address the goal of fast and high resolution imaging. Specifically, we consider models that account for acquisition process, image properties, and artifact correction. We also discuss efficient algorithms to solve the large-scale inverse problems that arise during reconstruction. To illustrate our model-based reconstruction framework, we will use magnetic field inhomogeneity estimation and task functional MRI modeling as examples.

Claire's advisors are Anna Gilbert and Jeffrey Fessler.

Zoom link: https://umich.zoom.us/j/95306219220 Speaker(s): Claire Lin (UM)