Compact 7T and various updates

It has been nearly two months since I posted. In the meantime, several things have happened in my public research sphere.

The compact 7T magnet was delivered to UCSF. This was a major operation involving 6 days of cross-country trip of an 8-ton, high-tech, and sensitive machine riding on a giant truck. (I didn't travel this time.) The link here shows the last phase of the transportation, when the magnet was carried on a forklift across the streets of San Francisco to Byer's Hall. I learn that the building is a few miles away from the one I visited 21 years ago, with my PhD advisor John Clarke and a postdoc, Michael Moessle, to discuss low-field MRI of breast. More on that later.

I submitted 3 first-authored abstracts to ISMRM for its Scientific Meeting of 2027. Three more were submitted as co-authored abstracts. It is reasonable to think that AI must have been involved in many (if not most) abstracts submitted. I think ISMRM should officially use AI to reduce the painful burden of abstract formatting, by automatically converting a user-provided word or pdf file into HTML.

I reviewed two journal articles, one for MRI and the other for MRM.

An R21 research grant proposal was submitted to NIH in collaboration with Mayo Clinic; I am a co-PI with Mayo Clinic leading the effort.

Local rotating frame paper

When I started my GE career in 2008, I was given a project on parallel transmit pulse design. New to the field, I started with reading publications on the subject, and by late 2009, I had figured that many textbook treatments of spatial RF pulse design methods can be reformulated through what I called the local rotating frame transformation. And to me the new formulation looked very advantageous. For one thing, I was able to explain why so-called low-tip-angle method works well for not-so-low flip angles. In physics, that's sufficient to gain recognition and to be published. Well, people in the MRI field were not so impressed. No patent application was approved, and no journal publication followed despite multiple attempts by 2010. Then I had to move on.

Now the work, expanded and revised, is finally published (50 day free-access link) in an Elsevier journal, Magnetic Resonance Imaging, after a peer review with 3 reviewers. I feel like a composer whose symphony took decades to get out and be played (read). With three Appendices and 15 double-column pages, it is a substantial work for a single author paper. I just returned the proof. It is out there, waiting to ring a bell in (= resonate with) someone interested. 

PS. The plus: The journal does not charge authors to publish, has a single-paragraph abstract, and has good typesetting fonts for equations (except for vector arrows).

Talk at Yale conference

               

From what I read (link), the Yale School of Medicine founded a new research center in 2025, named Yale Biomedical Imaging Institute. The institute, apparently having a PET and MR focus, held the first conference on Biomedical Imaging Technology & Translational Science (BITTS) earlier this week. On Tuesday, I gave a presentation on 7T MRI at the GE lunch symposium, to the (mostly) Yale audience, along with one of my senior GE-HTIC colleague, a CT expert. Throughout the session, the audience was engaged and their questions were sincere. It was a good event, and I am very grateful to my colleagues who organized the team, invited me, and gladly provided transportation from upstate New York.

I cannot help but thinking, exactly 10 years ago, I was on the faculty of biomedical engineering, helping to establish a brand-new imaging institute in Korea, at Sungkyunkwan university. As I was travelling at that time, the researchers here had travelled to GE and Siemens headquarters to compare their 7T MRI offerings. As in that other place, YBII had young and experienced researchers presenting on their animal imaging, fMRI analysis, multimodal imaging, and MRI physics research. Do the worlds repeat themselves? ... in which people take turns to act different roles? 

I wish every success of the people and the institute of YBII. 

Two upcoming conferences

This week Yale university will hold its first ever Biomedical Imaging Technology and Translational Science (BITTS) conference in its west haven campus. The conference will feature different medical imaging modalities and be sponsored by major industry partners including GE HealthCare. Link: BITTS Conference | Yale Biomedical Imaging Institute.

In November, University of Iowa will host the 5th high-performance neuroimaging workshop, with focus on applications of MAGNUS 3T head-only MRI, which is expanding its research user base and will be commercially available next year. GE HealthCare and Magtron Inc. co-sponsor the meeting. Link: https://mri.medicine.uiowa.edu/5th-high-performance-neuroimaging-symposium.

Back to the basics on 2D line scan and DIANA

Prof. Jang-Yeon Park and his graduate student Jae-Youn Keum at SKKU published a two-author paper (link) in MRM on the theoretical analysis and experimental mitigation of image artifacts caused by so-called pseudo-steady-state (PSS) in gradient-echo based 2D line scan. The PSS has plagued Direct Imaging of Neuronal Activity since its inception through insidious artifacts that mimicked neuronal excitation signals. The new publication addresses the problem head-on, and can help control or eliminate such artifacts to make the line scan method more reliable. Kudos to the team of Prof. Park, whose incessant pursuit of the holy grail of MR neuroimaging continues to stimulate the field and refine methodology.

7th installation of MAGNUS in June

Somewhere in Texas, another research-dedicated 3T MAGNUS scanner was installed in June. This is the 7th in the world, and 5th in the USA. At least one more will come this year, continuing the trend of growing adoption of high-performance gradient systems for brain research. 

Helium-free MRI on the rise, and GE's compact 7T scanner

Helium-free MRI is on the rise, spurred by high helium cost as well as general awareness of resource conservation needs around the globe. GE had a very early technical lead on this front, but it is only very recent that it is devoting full efforts to commercialization of the technology for mainstream MRI scanners. In this week's ISMRM meeting, "Freelium" 1.5 T scanners were announced (again) as products-to-be of GE HealthCare. For research, a low-helium highlight was GE's Compact 7T scanner, presented by Tom Foo in a scientific oral session, as an Annual Meeting Program Committee-selected talk.

Combining high-field (7T), strong & fast gradients, and compact geometry is technically very difficult for even conventional liquid helium-filled magnets. Doing this on a helium-free system is almost insane because of the difficulty in magnet heating control. Making human 7T compact is therefore a very high-risk proposal.

The world's first such system was announced in the talk with impressive 7T brain images, achieved with minimal tailored hardware or processing for image quality. The technology might meaningfully change the 7T research landscape in the coming decade.