, suggesting that it could have potential for use in screening high-risk patients.
Although previous studies had shown that the approach could hold up to high-field MRI, the new study was a blind comparison in which raters did not have access to the high-field images.
In addition to portability, the device has potential advantages over high-field MRI, including low cost and no need for high-field physical shielding. It could be used for point-of-care testing, especially in remote or low-resource areas. It does not produce ionizing radiation, and has been used in intensive care units and pediatric facilities.
Advantages and limitations
The device isn’t ready for general use in MS. It performed well in periventricular lesions but less well in other areas. Ongoing research could improve its performance, including multiplanar imaging and image analysis.
“I think it still needs some work, but to me if it’s less expensive it will be particularly better for third-world countries and that sort of place, or possibly for use in the field in the United States or in North America. If something is detected, you can then bring the person in for a better scan, but I don’t know how sensitive it is – how much pathology you might miss. But in countries where there are no MRIs, it’s certainly better than nothing,” said Anne Cross, MD, who comoderated the session at the Americas Committee for Treatment and Research in Multiple Sclerosis, where the study was presented.
She also noted that the device is potentially safer than high-field MRI. “I don’t think it would be something insurance companies or patients would want to pay $1,000 for when they could get a better scan somewhere, but it’ll get better,” said Dr. Cross, who is a professor of neurology and chair of neuroimmunology at Washington University in St. Louis.
How reliable are low-field images?
In previous work, in which evaluators compared the two scans side by side, the researchers showed in 36 patients that the device performed well, compared with a 64mT scanner. “When we look at tandem evaluations, we can identify dissemination in space in 80%. When a patient has at least one lesion that is larger than 4 millimeters in its largest diameter, we are able to detect it in the ultralow field MRI with 100% sensitivity. The open question here is, what is the diagnostic utility of these scanners when we don’t have any information about the high-field images?” said Serhat Okar, MD, during his presentation of the study. Dr. Okar is a neurologist and postdoctoral researcher at the National Institutes of Health.
To answer that question, the researchers asked two raters to examine scans from the low-field MRI, but only an independent party evaluator had access to both scans.
The study included 55 MS patients who were seen for either clinical or research purposes. The average age was 41 years, and 43 patients were female. Two neuroradiologists served as scan raters. Rater 1 had 17 years of experience, and rater 2 had 9 years of experience. They each conducted assessments for periventricular, juxtacortical, infratentorial, deep white matter, and deep gray matter lesions, as well as dissemination in space. They marked the scan and filled out an online form with number of observed lesions and whether they observed dissemination in space, with responses checked against a high-field image by an independent neuroradiologist for true positive and false positive findings.
There was significant discordance between raters for observation of dissemination in space, with rater 1 reporting 81% positivity and reader 2, 49%. False positive analyses revealed a difference in their approaches: Rater 1 was more conservative in marking lesions, which led to fewer true positive and fewer false positive findings. Both raters had good performance in the periventricular lesions with similar, low rates of false positives.
Other areas were a different story. Both raters found a greater number of true positive and false positive areas in the juxtacortical, deep white matter, and deep gray matter areas.
The study was funded by Hyperfine. Dr. Okar and Dr. Cross have no relevant financial disclosures.