The objective of our study was to perform MRI in 2 groups of infants, A) term AGA with neonatal asphyxia (cord blood ph<7.1, 5 min Apgar ≤4, need for CPR at birth and hypoxic encephalopathy) and B) very low birth weight infants (VLBW) with grade III-IV intracranial hemorrhage (ICH) and posthemorrhagic ventriculomegaly requiring shunt insertion, and correlate MRI with neurologic outcome. MRI was performed at 6 mo corrected age with sagittal T1 weighted images and axial T1 and T2 weighted images (± contrast). Neurologic exams were categorized as normal, neuromotor impairment (with no developmental delay) and cerebral palsy (CP) MRI readings and neurologic exams were performed by independent examiners.

Results: 24 infants; 10 (gp.A) and 14 (gp.B) were followed to a mean corrected age of 24 mo. Of 10 term infants (A), 6 had normal neurologic exams (all 6 had normal MRI); and 1 had neuromotor impairment (MRI was normal). The remaining 3 had severe CP; MRI was normal in 1, thinning of corpus collosum with normal myelin pattern in 1 and increased signals from basal ganglia and delayed myelination in the third. The sensitivity of the MRI in predicting outcome was 50%, specificity was 100%. Of the 14 VLBW infants with III-IV ICH (B), 4 infants had normal neurologic exam; the MRI was normal in 1, thin corpus with ventricular dilation seen in another, tenting of corpus and gyri by shunt seen in the third and cerebral necrosis seen in the remaining infant. 6 infants had neuromotor impairment and 4 had CP. Abnormalities in MRI in these 10 included cerebral necrosis in 4, dilated lateral ventricles in 3, trapped 4th ventricle in 2, dysgenesis/agenesis of corpus in 8, and delayed myelination in 3. The sensitivity of the MRI for predicting neurologic outcome was 100% while the specificity was 25%.

Conclusion: MRI has high specificity for term infants with asphyxia and sensitivity for VLBW infants with III-IV ICH.