In one of the first deployments of a point-of-care MRI scanner, Dr. Kevin Sheth of Yale School of Medicine discussed the application of low-field POC MRI, outlining the constraints of conventional MRI and advantages of MRI at the bedside.
Neuroimaging is used extensively for diagnosis, triage, and management of brain injury, Sheth said. Noncontrast CT and MRI have been gold standards for diagnosing stroke, and MRI is increasingly being recognized for detecting both ischemia and intracranial hemorrhage.
That said, the limitations of conventional MRI are numerous and include the need for dedicated imaging suites and trained MRI technologists, as well as the special magnet cooling requirements necessitated by superconducting scanners. There is also the need to remove ferrous materials from the room and to transport patients who may be very ill out of their hospital rooms to be imaged, Sheth said.
In contrast, point-of-care MRI with low-field MRI scanners avoids many of these challenges. Such systems can be plugged into standard power outlets, have no cooling requirements, are significantly lower in cost, and do not require clinical equipment such as oxygen tanks, ventilators, intravenous poles, and dialysis machines to be moved out of the room, he explained.
In his ISMRM talk, Sheth discussed how he and colleagues at Yale's neuroscience intensive care unit (ICU) evaluated a 0.064-tesla POC MRI scanner under development. They enrolled 246 patients between July 2018 and June 2020; a total of 72 patients did not undergo POC MRI for reasons such as claustrophobia, patient size constraint, or scanner malfunction. In total, they obtained 211 exams on 174 patients. The central nervous system (CNS) pathologies that were scanned included ischemic stroke, hemorrhagic stroke, subarachnoid hemorrhage, traumatic brain injury, and brain tumor.
The scanner's portability allowed for serial MRI scans to be performed, Sheth said.
"One of the things that low-field, point-of-care imaging allowed us to do, which is very rare for conventional MRIs at our institution, is the ability to perform serial exams," he noted. "We were able to do this in 16 patients."
In addition, trained MR technicians were not needed -- bedside exams could be performed by research staff -- and scanning parameters were controlled via a tablet.
A total of 20 patients had COVID-19 with an inability to ascertain their mental status. Hemorrhagic stroke was the second most common pathology, with 12 patients presenting with it, followed by ischemic stroke (nine patients). Other pathologies included traumatic brain injury, brain tumor, and subarachnoid hemorrhage.
"All point-of-care MRI findings were in agreement with conventional radiology reports," Sheth said.
Furthermore, use of the POC MRI scanner enabled radiologists to identify pathologies in patients who could not undergo conventional MRI: In one patient who was too unstable to be transported to conventional imaging, a right cerebellum infarction was detected, and in another patient who was paralyzed and sedated, evidence of a large, hemispheric infarction was detected.
On the other hand, one patient had a diffuse subarachnoid hemorrhage that was not observed with the point-of-care MRI scanner. The scanner's inability to perform susceptibility-weighted imaging precluded the ability to examine microhemorrhages that one might see in conditions such as cerebral amyloid angiopathy.
Still, the overall data demonstrated that point-of-care MRI was undeniably valuable in the ICU, according to Sheth.
"These findings, in short, demonstrated the first successful deployment of a portable MRI to the bedsides of patients with critical illness, sidestepping the risks of transporting patients with critical illnesses by allowing these exams to occur in the immediate, bedside environment," he said. "We believe that the pandemic and nonpandemic scanning experience demonstrates the promise for bedside neurological assessment in other resource-limited settings."
Sheth plans to explore broadening the use of point-of-care MRI beyond the neuro ICU, working with clinicians at Massachusetts General Hospital to generate multicenter data. In addition, the researchers plan to incorporate artificial intelligence for better image analysis.
Copyright © 2021 AuntMinnie.com