Background: There is no consensus on the optimum timing of MRI in neonates with hypoxic-ischemic encephalopathy treated with hypothermia. Reliable early imaging assessment might help managing treatment.
Objective: To assess non-random differences between early and late MRI that might influence intensive-care decisions.
Materials and methods: This single-center retrospective study included all asphyxiated term neonates eligible for hypothermia treatment November 2009-July 2012. MRI scans were systematically performed at day 4 (early MRI) and day 11 of life as part of routine protocol. Two experienced pediatric radiologists reviewed both scans according to three assessment methods: a pattern classification, a scoring system and a simplified classification. Agreement between early and late imaging findings was assessed using Cohen's kappa coefficients.
Results: Thirty-three neonates were included. Interobserver agreement was excellent. Early MRI detected all severe injuries. Agreement between early and late MRI was excellent for the simplified classification (κ = 0.82), good for the pattern classification (κ = 0.64), and good to excellent for 3 scores out of 4 in the scoring system (κ = 0.70-0.89).
Conclusion: Early MRI may provide valuable information about brain injury to help parents and neonatologists in intensive-care decisions at the end of hypothermia treatment.
Keywords: Brain injury; Diffusion-weighted magnetic resonance imaging; Hypoxia-ischemia brain; Induced hypothermia; Neonates.