553 - Adequacy of an in-NICU 1T MRI Compared to 3T MRI for Clinical Management

Poster Number: 553
Publication Number: 553.1623

Background: In the neonatal period, brain magnetic resonance imaging (MRI) is a valuable clinical tool that offers the ability to visualize brain injury and assess brain growth and development. At the Brigham and Women’s Hospital Neonatal Intensive Care Unit (NICU), we began using the Embrace scanner, a 1-Tesla (1T), in-unit MRI clinically in March of 2020. This scanner allows for easier and safer access to brain MRIs for high-risk infants.

Objective: This retrospective study aims to assess the added diagnostic utility and clinical value of a 3T MRI after obtaining a 1T MRI.

Design/Methods: We identified a cohort of infants with an initial MRI in the 1T scanner that had repeat imaging within 14 days in a 3-Tesla (3T) scanner. All infants were admitted to the level III NICU at Brigham and Women’s Hospital. The MRIs were done for clinical purposes (per BWH clinical practice guidelines) or as part of research protocols. Follow-up 3T MRIs were performed to obtain higher resolution imaging, magnetic resonance angiogram (MRA), magnetic resonance venography (MRV), or as part of a research protocol. Both clinical and research MRIs were reported in the same format and were interpreted by pediatric neuroradiologists at Boston Children’s Hospital.

Results: 15 clinical and 19 research MRIs were performed on 1T with 13 scans in full-term infants and 21 in preterm infants (half prior to term equivalent). 26/34 (76%) of the 1T MRIs were interpreted as abnormal. 28 clinical and 6 research MRIs were performed on 3T with 13 scans in full-term infants, 4 in preterm infants prior to TEA, and 17 in preterm infants at TEA. 27/34 (79%) of the 3T scans were found to have an abnormality. In 31 cases (91%), the 3T MRI showed similar or expected evolution of known findings found at 1T (examples in Figure 1). Variation existed in three cases where either an abnormality was defined as artifact, or a more extensive recognition of injury occurred (Table 1, Figure 2). There appeared to be no clinical impact of the change between the 1T and 3T imaging results.

Conclusion(s): Images from 1T MRI are sufficient for characterizing a wide range of neonatal brain injuries and abnormalities. During the study period, there were limitations associated with the 1T scanner, such as the absence of susceptibility-weighted imaging (SWI) and MRA/MRV. Moreover, in cases involving known or suspected malformations, 3T MRI offers increased signal-to-noise ratio (SNR) allowing increased spatial resolution. However, within our cohort, these limitations did not impact clinical management.