Protocol development for a Very Low Field (0.047 T) MRI system

Very low-field (VLF, <0.1 T) portable MRI scanners are emerging as a new
category of systems with much lower costs than conventional MRI scanners and
the ability to be used in sites and applications where conventional MRI is
impossible.

In September 2022 we received approval from the METC to start scanning human
volunteers on a commercial VLF MRI system from Hyperfine, specifically for
protocol optimization for a number of different contrasts. This study has
started and is providing very useful information on how best to set up such
protocols. However, the Hyperfine unit is an expensive commercial system, with
limitations on open software development as well as hardware adaptations, which
places it outside the realm of many low and middle income countries (LMICs),
for example. We have extensive funding from the National Institutes of Health,
the European Research Council and the NWO to develop a VLF system with
increased portability, lower cost and open-source hardware and software which
can be used in LMICs.

In this proposal we intend to study a number of different aspects of VLF MRI.
These studies will be performed in parallel to those on the Hyperfine unit,
allowing comparison of results and optimization of many aspects of the LUMC VLF
unit. In addition to a study to optimize image contrast in the brain using a
number of different MRI techniques, which will parallel the approved study on
the Hyperfine, we will evaluate a number of techniques for enhancing image
quality which cannot be performed on a commercial system, namely:

i) Active noise cancellation
ii) Correction for small drifts in the magnetic field during the scan
iii) Undersampled data collection and AI reconstructions

In addition to neurological measurements, the flexibility of our 0.047 T system
allows other parts of the body to be studied, unlike the Hyperfine system. We
will develop techniques to measure the lipid/muscle content in both the arm and
leg, which can ultimately be used in the estimation of malnutrition in
developing countries.

Dissemination of findings will be performed via articles in world leading
journals on MR techniques. Such developments and publication of results are
common to the MR community, and lead to a continuous development and
improvement of the capabilities of MR scanners.

Our overall aim is to measure several distinct MR parameters at 0.047 Tesla,
and to use these to derive optimal MRI-protocols in terms of contrast and SNR.
The images produced will then be compared with those from the commercial
Hyperfine system. Some of these new developments may subsequently be used in
clinical research protocols (which are not a part of this protocol), other
developments are more fundamental technical MR developments for which
applications will only benefit in the future.

Only projects aiming at the development, optimization and/or interpretation of
non-invasive MR techniques are included. MRI-protocol development will follow
the usual roadmap of MR physics research that consists of an iterative process
of identification of new requirements or artefacts in existing techniques,
MRI-protocol optimization, sequence development, pilot experiments, quality
review meetings, and finally back to identification of sources of artefacts.
For each of the studies (noise cancellation, data undersampling, SNR
comparisons for six different MR contrast mechanisms, and development of
lipid/muscle ratios we will scan 20 healthy volunteers (total number=240) to
publish results of these new protocols/sequences.

All of the issues concerning safety and risk at 0.047 T are much lower than the
corresponding considerations at clinical MRI field strengths (1.5 and 3T).
Projectile forces are proportional to the magnetic field multiplied by the
spatial gradient of the magnetic field, as so are ~30 and 60 times less than
1.5T and 3T respectively, power deposition in the subject is proportional to
the square of the magnetic field, so is ~900 and 3600 times less than at 1.5T
and 3T. The scans are almost silent, no special clothing is needed, and the
environment is much less claustrophobic than for a conventional MRI scan. Many
thousands of scans have been performed on both patients and healthy volunteers
on VLF scanners with field strengths of 0.08 and 0.064 T.