New massively-parallel array coils for extending the reach of MRI and concurrent Transcranial Magnetic Stimulation
New massively-parallel array coils for extending the reach of MRI and concurrent Transcranial Magnetic Stimulation
Date 09 Jun 2022
Location online event

In this talk, I will describe a new generation of massively-parallel array coils aimed at improving human MR neuroimaging well as simultaneous imaging and neuromodulation. An overarching goal is to push the limits of MRI resolution and speed, particularly for functional and diffusion MR imaging, enabling non-invasive studies of brain structure and function at finer spatial scales.

Extending the reach of MRI & concurrent Transcranial Magnetic Stimulation

Specifically, I will discuss the synergistic combination of high-channel count radiofrequency (RF) receive arrays with “B0” shim arrays for accelerating image acquisition while reducing image artifacts (particularly at 7 Tesla and beyond). In addition to new array coils for neuroimaging, we are also developing multi-channel, MR-compatible transcranial magnetic simulation (TMS) coils that allow the electric field to be dynamically steered and reoriented over the cortical surface. This provides new degrees of freedom for probing the activity of multiple brain circuits either simultaneously or sequentially. Bringing together two types of array coils, we are integrating a 48-ch TMS array with a 28-ch RF receive array to provide unprecedented image quality for simultaneous TMS-fMRI experiments.

Exploiting these hardware capabilities further, we have begun exploring the use of the current-carrying loops in our TMS coils for MRI applications such as diffusion encoding, B0 shimming, and zoomed imaging. The unifying theme of these hardware subsystems is “local field control” using the many degrees of freedom provided by large arrays to dynamically shape both the B-field and E-field for imaging and neuromodulation, thus extending the utility of MRI and TMS for systems neuroscience.

Discussion points of the webinar:

Is it possible to use the same close-fitting coil array for both neuromodulation and for MRI spatial encoding/shimming?

What opportunities and challenges arise for ultra-high field brain MRI as the RF receive array channel count grows from 32 to 64 to 128 channels?

What can we do with a rich basis set of rapidly-switchable, spatially distinct B0 fields, beyond just conventional B0 shimming?

Jason Stockmann, PhD

Massachusetts General Hospital, Harvard Medical School

Jason Stockmann, PhD, is broadly interested in magnetic resonance imaging hardware and acquisition methods for improving data quality for both structural and functional imaging. He has worked on diverse MRI scanners ranging in field strength by two orders of magnitude, from low-field (70 mT) to ultra-high field (7 Tesla). He explores synergistic combinations of hardware, pulse sequences, and image reconstruction methods that address unmet needs in MRI research, especially for diffusion and functional brain imaging with echo planar imaging (EPI) acquisitions. A major thrust of this work has been to develop multi-coil (MC) shim arrays and associated amplifier hardware and optimization methods to improve magnetic field homogeneity inside the body, thus reducing image distortions and other artifacts. More recently, he and colleagues have applied MC arrays to perform dynamic local field control, creating tailored nonlinear field offsets for (i) improving lipid suppression in spectroscopy, (ii) selectively exciting and imaging target anatomy with increased efficiency, and (iii) providing supplementary spatial encoding.

Dr. Stockmann is also interested in low-field, portable MRI for point-of-care brain imaging. He has contributed to Dr. Lawrence Wald and Dr. Clarissa Cooley’s program to build a lightweight prototype brain scanner based on a Halbach array of permanent magnets (80mT main magnetic field). His primary role in this project has been to design pulse sequences and RF pulses that are robust to extreme field inhomogeneity. He has also helped build a generalized reconstruction framework that incorporates the full signal forward model including field nonlinearity. In parallel with this work, he has developed an interest in open-source hardware for MRI research and education. To this end, he contributed to a team effort by Dr. Wald’s group to build 20 tabletop MRI scanners (0.2 Tesla) for an undergraduate engineering lab course at MIT, at a cost of less than $10K per scanner.

He is strongly committed to open-source science and reproducible research across sites. All of his hardware designs and software are available online or by request.