Key points are not available for this paper at this time.
BACKGROUND: Despite improving outcomes, extracorporeal membrane oxygenation (ECMO) remains associated with high rates of acute brain injury, impacting survival and outcomes. Clinical imaging modalities, including computed tomography and standard-field magnetic resonance imaging (≥1.5 Tesla), have either limited early detection of ischemic-type injuries or are incompatible with ECMO, causing diagnostic delays. A Food and Drug Administration-cleared ultra-low-field portable magnetic resonance imaging (pMRI; 0.064 Tesla) has shown promise in acute brain injury detection and use with MRI-unsafe equipment, but pediatric ECMO data are limited. METHODS: This institutional review board-approved, single-center study was conducted at a quaternary children's hospital. Neonatal and pediatric patients on ECMO from September 2023 to August 2024 were screened. Enrolled subjects underwent pMRI using Swoop (Hyperfine, Guilford, CT). Feasibility (resource use, concurrent equipment) and safety metrics (ECMO flow/vitals, cannula displacement) were collected along with patient/ECMO variables and all head imaging. Two blinded pediatric neuroradiologists independently reviewed images. RESULTS: Of 41 screened patients, 20 were enrolled, 16 had pMRI attempted, and 11 completed full imaging (T1, T2, fluid-attenuated inversion recovery, and diffusion-weighted imaging). Median staff required for pMRI was 6 (interquartile range, 5-7) versus 7 (interquartile range, 7-7) for head computed tomography. Median pMRI positioning and imaging time was 66 (interquartile range, 61-70) minutes versus 75 (interquartile range, 70-79) minutes for head computed tomography with intrahospital transport. One subject had a ≥20% but clinically nonsignificant decrease in blood pressure; no safety events were attributable to pMRI. A total of 70% to 75% of image sets were of diagnostic quality. CONCLUSIONS: pMRI is feasible in pediatric ECMO with no clinically relevant complications, lower resource use, and generation of diagnostic quality images in our cohort. REGISTRATION INFORMATION: URL: https://www.clinicaltrials.gov; Unique Identifier: NCT06074406.
Wallisch et al. (Wed,) studied this question.