Lang et al. have proposed dividing the usage of pMRI imaging in ED into three groups based on their field strength and portability (39). The first group are “easy-to-site suite” scanners using a standard superconducting solenoid magnet architecture; Examples are High-field (3T) (40, 41) and Mid-field (between 0.5 T and 1.0 T) head-only MRI scanners (42–44). The second group would be truly portable scanners that would operate at a low field (50 mT to 200 mT), need unconventional electromagnetic interference (EMI) mitigation, and must operate with substantially reduced electrical power and without water cooling or cryogenics. The 64 mT Hyperfine Swoop scanner (Guilford, CT, USA), the first FDA-cleared clinical pMRI scanner (12, 45), and the 80 mT “Halbach-bulb” rare-earth magnet configuration (38, 46, 47) are some of the advances in this class. The ultra-low-field shield-free pMRI using a Samarium-cobalt magnet enables the cancellation of EMI and elimination of the traditional radiofrequency shielding cage. This scanner promisingly met the clinical needs to diagnose brain tumors and stroke (48). The third group are “hand-held” level MRI devices, likely with greatly reduced imaging capabilities, but inexpensive and small enough to be considered MR detectors or monitoring devices more than diagnostic imaging devices. An example of this is a 7 kg device “MR Cap” (49). This type of instrument could warn of early changes and impending events by continuously imaging the brain in ED or ICU, particularly in patients that are difficult for clinical evaluation and examination (e.g. sedated patients).

Stroke is the second leading cause of mortality and morbidity worldwide and affects 15 million people each year. Neuroimaging plays a significant role in both the diagnosis and treatment of stroke within the golden time (first 3–4.5 h after the event). The availability of open-bore systems has enabled better temporal access than fixed CT or conventional MRI and resulted in timely therapeutic intervention and reduced door-to-needle time (9). DWI remains the gold standard sequence imaging for acute stroke and is available on low-field and very low-field/ultra-low-field MR scanners (29). In 2020, Cahn et al. suggested the clinical utility of POC MRI for evaluation of stroke patients. Using the 64 mT scanner (50), they examined acute stroke patients and were able to produce 3D clinical quality images by performing DWI sequences and suggested it as a safe and viable tool in complex clinical care environments (17). Zubair et al. (13) and Hovis et al. (15) in 2021 reported using 64 mT pMRI for evaluation of patients with suspected stroke. The implementation of pMRI resulted in faster work-up, decreased hospital stays, and yielded reliable results compared to non-contrast CT scan and conventional MRI. In 2022, Yuen et al. utilized 64 mT pMRI as first-line diagnostic imaging for the detection of ischemic stroke. POC MRI was able to capture lesions as small as 4 mm. Infarcts were detected as hyper-intense regions on T2-W, FLAIR, and DWI sequences with a sensitivity of 98, 100, and 86%, respectively. Stroke volume measurements were consistent between low-field pMRI and conventional high-field MRI studies. Significant correlations were seen between low-field pMRI stroke volumes and stroke severity, and consequently functional outcome at discharge. They validated the clinical use of low-field pMRI as a novel imaging solution for patients with ischemic stroke, especially in resource-limited environments (18).

In 2020, Shah et al. analyzed the ICH detection capacity of a bedside POC MRI scanner. Prospectively, using a 64 mT scanner, a pathologic lesion was identified on every exam with 100% sensitivity by T2-W and FLAIR sequences. They suggested that low-field, POC MRI may be used to detect hemorrhagic strokes at the bedside (19). In 2021 Mazurek et al. evaluated the use of low-field MRI (64 mT) for assessing ICH in comparison with conventional 1.5/3 T MRI or non-contrast CT scan. PMRI could detect ICH with 80.4% sensitivity and 96.6% specificity. They found that hematoma volume measurements derived from pMRI exams correlated with impaired cognitive status and worse functional capacity at discharge. They proposed the Low-field pMRI as a useful technology in resource-limited settings (16). With the same methodology, Sheth et al. assessed mass effects and the resulting MLS in brain-injured patients using pMRI. There was significant agreement between pMRI and conventional MRI or CT. Low-field pMRI identified MLS with a sensitivity of 93% and specificity of 96% and could predict poor clinical outcomes at discharge. They regarded bedside pMRI as valuable for detecting mass effects (14).

In patients admitted to the ICU, transportation to an MRI suite may place the patient at numerous risks, including compromise of venous or arterial access, endotracheal tube displacement, hypoxia, hypotension, increased intracranial pressure, and cardiac arrest. In patients with contagious infections such as COVID-19 other concerns such as exposure to a larger patient population and hospital staff and the need for decontamination of radiology suites are logistical barriers (12, 22). On the other hand, substantial delays between ordering and obtaining an MRI may further increase patients' hospital length of stay and subsequent total hospitalization costs. Importantly, limited access to MRI hinders timely and targeted therapeutic decision-making in this patient population (4, 11, 21).

Low-field pMRI (64 mT) produces less than 5 Gauss beyond the 5-foot (1.52 meter) safety zone of the magnet's center, which is significantly lower than conventional MRI. It can be feasibly employed in the presence of all ICU equipment during scanning without adverse events (26). Additionally, POC MRI may improve the functionality of existing fixed MRI machines by removing complicated and time-sensitive critically ill patients from the scheduled waiting list (11, 20, 21). In 2020–21, Turpin et al. (22) and Mazurek et al. (23) reported their experience of imaging critically ill COVID-19 patients with a self-shielding, 64 mT pMRI. The use of pMRI was safe, and feasible and led to changes in clinical management based on imaging results. Also, the diagnostic MRI quality was graded to be adequate (85%) with FLAIR sequences demonstrating the highest quality, and DWI scoring the lowest (22). Similarly, Sheth et al. assessed the utility of pMRI in complex clinical settings such as neuroscience labs and COVID-19 ICUs and reached similar results (12). However, Kuoy et al. obtained a diagnostic rate of 72% and mentioned that motion artifact, habitus, and lower field-strength may necessitate higher field-strength MRI subsequently. They also found that the turnaround time was reduced for patients in the ICU (5.3 vs. 11.7 h for fixed scanner) but not for patients in the ED (3.4 vs. 3.7 h for fixed scanner), suggesting that ease of transport and proximity to the fixed scanner may affect turnaround times (21). Regarding the time-dependent nature of critically ill patients and particularly those resuscitated from cardiac arrest, Beekman et al. demonstrated that pMRI could successfully identify patients with hypoxic-ischemic brain injury (HIBI). Accurate outcome prediction is required to prevent withdrawal of life-sustaining therapy in comatose survivors with good functional outcomes; this is possible through serial MR imaging and quantitative analyses using FLAIR intensity (11). The same applies to patients on extracorporeal membrane oxygenation (ECMO) as both logistic and diagnostic advantages of POC MRI led to early detection and timely intervention in acute brain injuries and could improve their outcomes (20).