Two accelerated magnetic resonance imaging (MRI) protocols for pediatric brain imaging, GOBrain and Deep Resolve Swift Brain, developed by Siemens Healthineers (Erlangen, Germany), were evaluated in a series of clinically relevant pediatric cases at 3 Tesla. Pediatric patients are particularly prone to motion, may be uncooperative, and often require sedation, especially in emergency settings. Consequently, there is a persistent clinical demand for fast brain MRI protocols that provide diagnostically sufficient image quality while minimizing examination time. Contemporary turbo spin-echo (TSE)-based clinical protocols commonly integrate parallel imaging (PI) and simultaneous multi-slice (SMS) techniques to achieve substantial reductions in scan time. Recent advances in three-dimensional volumetric encoding, compressed sensing, and deep learning (DL)-based reconstruction have further mitigated geometry-factor-related noise amplification, enabling higher acceleration factors (GOBrain). In parallel, echo-planar imaging (EPI) has emerged as a promising approach for ultrafast multi-contrast imaging. To overcome the limitations of single-shot EPI, a multi-shot EPI-based brain MRI protocol combined with the DL-based reconstruction method Deep Resolve Swift Brain has been developed. This approach leverages the efficiency of EPI while improving image quality. Using these accelerated protocols, a comprehensive diagnostic multi-contrast brain MRI examination, particularly suited to triage and emergency imaging, can be completed in minutes. This case overview, including therapy-related leukencephalopathy in acute lymphoblastic leukemia (ALL), a brain abscess, traumatic diffuse axonal injury (DAI), a posterior circulation infarction due to vertebral artery dissection, leuokostasis syndrome, and a posterior fossa tumor with obstructive hydrocephalus, demonstrates the potential clinical feasibility of both protocols in pediatric neuroimaging. Both protocols position them as supplementary options alongside established imaging protocols, while dedicated high-resolution protocols might remain necessary for subtle pathological findings, such as focal cortical dysplasia, and for neuronavigation until larger comparative studies are available.
Kendel et al. (Thu,) studied this question.