By Steve Bush 20th January 2022
University of Hong Kong researchers have used artificial intelligence to perform a brain scan with a conventional rare-earth magnet rather than a large superconducting magnet, and without a screened room.
The ‘ultra-low field magnet’ – according to MRI nomenclature – was designed using almost 90kg of samarium cobalt as its field source – used because it is less temperature sensitive than neodymium-iron-boron – combined with iron pole pieces to make a head-sized uniform 0.055 Tesla field within a structure wide enough to accommodate the head and shoulders of the subject, and tall enough to minimise claustrophobia.
With such a low field – 3T field from a superconducting magnet is more normal – electromagnetic interference dominates the signal to noise ratio.
One way to increase signal-to-noise to usable levels is to build a screened room around the scanner, which was against the aim of the project, which is to bring brain scanning to poorer countries.
Instead, the team positioned ten EMI sense coils in and around the machine and its control electronics, then interspersed EMI sensing into the scanning procedure.
Using this data, a convolutional neural network was trained to build a model of the EMI present where the subject’s head was, and then that result was subtracted from the actual scan data.
Results were usable, and within 5% of the scanner tested in a screened room, according to to the team.
Contrast and resolution were not as good as a conventional 3T MRI scanner, but scans were clear, and good enough for a clinician to identify tumours and signs of stroke in human subjects.
Power consumption was a wall-socket-friendly 1.2kW.
Two further advantages are that it is quiet in operation, and does little to heat or pull surgical clips within the brain cavity.
The scanner and results are described in the clearly-written Nature Communications paper ‘A low-cost and shielding-free ultra-low-field brain MRI scanner‘.
Future work, according to the paper, could involve building a neodymium-iron-boron magnet version, which would offer a higher field and cost less, with its temperature coefficient compensated for by measuring field strength during operation and feeding this figure into the maths.
Using a Halbach magnet array could further decrease the magnet size, weight and cost further, said the researchers, but there is a limit to how much the magnetic box can be shrunk before people will be un-nerved within it – scans take five to seven minutes.
The paper credits earlier and concurrent work, including Hyperfine’s commercial 0.064T MRI brain scanner.
Design data has been made available on GitHub
Tagged with: Magnet MRI research
Your email address will not be published. Required fields are marked *
Get our news, blogs and comments straight to your inbox! Sign up for the Electronics Weekly newsletters: Mannerisms, Gadget Master and the Daily and Weekly roundups.
Read our special supplement celebrating 60 years of Electronics Weekly and looking ahead to the future of the industry.
Read the Electronics Weekly @ 60 supplement »
Read the first ever Electronics Weekly online: 7th September 1960. We've scanned the very first edition so you can enjoy it.
Read the very first edition »
Electronics Weekly teams up with RS Grass Roots to highlight the brightest young electronic engineers in the UK today.
Read our special supplement celebrating 60 years of Electronics Weekly and looking ahead to the future of the industry.
Read the Electronics Weekly @ 60 supplement »
Read the first ever Electronics Weekly online: 7th September 1960. We've scanned the very first edition so you can enjoy it.
Read the very first edition »
Tune into this Xilinx interview: Responding to platform-based embedded design
Tune into this podcast to hear from Chetan Khona (Director Industrial, Vision, Healthcare & Sciences at Xilinx) about how Xilinx and the semiconductor industry is responding to customer demands.
By using this website you are consenting to the use of cookies. Electronics Weekly is owned by Metropolis International Group Limited, a member of the Metropolis Group; you can view our privacy and cookies policy here.