Bernice Grafstein Lecture in Neuroscience: Addressing Barriers in High Field fMRI Acquisitions
The 2021 Bernice Grafstein Lecture in Neuroscience will be held virtually on Thursday, October 28, at 3 PM, Eastern Time. The Intergrate Program in Neuroscience is pleased to welcome Dr. Lawrence Wald as this year's keynote speaker.
The Bernice Grafstein Lecture in Neuroscience is an opportunity to discover and learn about cutting-edge neuroscience from a world-renowned invited speaker. The lecture is generously supported by Dr. Bernice Grafstein, herself a pioneer of neuroscience.
This lecture is open to the public.
Register Here
Key Note Speaker: Dr. Lawrence Wald, PhD
Dr. Wald is currently a Professor of Radiology at Harvard Medical School and Affiliated Faculty of the Harvard-MIT Division Health Sciences Technology. His work has explored the benefits and challenges of highly parallel MRI and its application to highly accelerated image encoding and parallel excitation and ultra-high field MRI (7 Tesla) methodology for brain imaging including improved methods for matrix shimming and gradient coil design. His lab also focuses on motion mitigation methods, portable MRI technology, and is developing a prototype functional Magnetic Particle Imaging scanner.
Talk Abstract:听
MRI acquisition methodology has concentrated on the considerable engineering challenges of rapidly applying the strong magnetic fields needed to polarize the spins, rapidly encode the spatial information and sensitively detect the MR signal. Happily, many pure engineering challenges have fallen. For example, the latest gradient coils under development handle unprecedented heat removal, current, voltage, force, torque, eddy current and other gradient-magnet interactions. These successes have forced us to confront biological ceilings imposed by the stimulation of peripheral nerves during gradient switching. Also, the MR gradient coil is now no-longer alone in its role of encoding the image information. The detector array has improved to the point where it produces the majority of the final Nyquist-sampled data space, as opposed to being encoded in the signal by the gradient coil.
In this talk, I describe our attempts to improve these two sources of image encoding (gradient and detector coil array), and then pivot our methodology toward this view of shared image encoding to try to make them work better together. By dove-tailing the strengths and weaknesses of the gradient coil and detector array, we can go one step further and create new sampling strategies. Successful recent offspring of this marriage include 鈥渂lipped-CAIPI鈥� Simultaneous MultiSlice (SMS) fMRI, 鈥渨ave-CAIPI鈥� anatomical imaging and EPTI for either fMRI or anatomical imaging, and finally, a new capability to extend the image reconstruction to non-stationary objects and thus mitigate motion confounds.
