NEW YORK (Reuters Health) – A new MRI scanner that can be built at low cost but produces high-quality neuroimaging could improve access to the technology in low- and middle-income countries, researchers suggest.
Because the scanner needs no shield and depends on ultra-low magnetic field (0.055 Tesla) to produce images, it is more compact and far less expensive than current versions. Its developers believe it can be constructed for less than $20,000 as compared with millions of dollars for a traditional scanner, the researchers note in Nature Communications.
“The high cost of procuring, siting/installing, maintaining and operating the current clinical scanners constitutes a major roadblock in MRI accessibility in healthcare,” Dr. Ed X. Wu, a professor of biomedical engineering at the University of Hong Kong, and colleagues write. “Low-cost, low-power, compact, open, and shielding-free (ultra-low magnetic field) MRI for brain imaging as demonstrated here aims to complement rather than compete with existing high performance clinical MRI in healthcare.”
Ultra-low magnetic field (ULF) “MRI holds several inherent attractions when compared to high-field MRI,” the researchers note. “They include open magnet configuration for patient comfort, low acoustic noise levels during scanning, low sensitivity to metallic implants, less image susceptibility (to) artifacts at air/tissue interfaces, and extremely low (radiofrequency) specific absorption rate (SAR).”
Currently MRI accessibility around the world is low, Dr. Wu and his colleagues write, and is concentrated in high-income countries. To develop the new prototype, the team used deep-learning software to improve the images created by the low field and get around the need for a shield – which in traditional machines blocks electromagnetic waves that make the signal noisy.
The researchers tested their new machine, which along with being much smaller and lighter than the traditional devices also can be plugged into a standard AC socket, on 25 patients to see if the machine could aid in the diagnosis of neurological diseases, including stroke, tumors, and brain trauma.
While the images in the study from the ultra-low-field device are not quite as sharp as those from a 3T machine, they are clear enough to show pathology, the team found.
“This scanner has the potential of being implemented in low-cost medical settings where MRI technology is needed but yet readily available,” said Dr. Tamer Ibrahim, a professor of bioengineering and psychiatry at the University of Pittsburgh and director of the 7 Tesla Bioengineering Research Program and the Radiofrequency Research Facility at the University of Pittsburgh.
“The cost of MRI systems has been relatively dropping over the years,” Dr. Ibrahim, who was not involved in the research, told Reuters Health by email. “That being said, it is still difficult for such technology to reach remote places. This technology can alleviate this issue. This technology will not replace standard MRI point of care, which is typically done at 1.5 T, 3T, and more recently 7T. The images presented are really excellent quality for a 0.055T MRI field strength. Its implementation can significantly impact the healthcare of remote communities.”
Dr. Haris Sair, director of neuroradiology at Johns Hopkins School of Medicine in Baltimore, Maryland, also sees real possibilities in the new prototype.
“A low-cost portable MRI scanner has the potential to redefine patient care in areas where imaging resources may be limited, for example in rural areas or developing countries,” said Dr. Sair, who also was not involved in the research. “Furthermore, even in resource-rich areas, it may help patients who are precluded from obtaining traditional MRI due to risks of transportation or the presence of MR-contraindicated implants and medical devices.”
Still, he’d like to see more research. “You’d have to do a decent-sized study with a good number of patients in a real-world setting,” Dr. Sair told Reuters Health by phone.
One concern is that the artificial intelligence processing might lead to artifacts in the images. But, if more research illuminates any limitations of the technology and “if people have knowledge of what specific things they need to look out for in terms of potentially misinterpreting an image, it could be a tremendously useful tool,” he said.
The authors did not respond to requests for comment.
SOURCE: https://go.nature.com/3m4wDxx Nature Communications, online December 14, 2021.
Source: Read Full Article