Objectives: To illustrate the early magnetic resonance imaging (MRI) findings in neonates with hypoxic ischemic encephalopathy (HIE) and to define the fallacies of cranial ultrasound (US) as a definitive diagnostic modality for intracranial pathology associated with HIE.
Patients & Methods: The study included 150 neonates with HIE classified into 3 groups on the basis of Sarnat & Sarnat staging system of HIE. Cranial US was performed using real-time sector scanner with 7.5 MHz transducer. MRI imaging was performed after neonatal sedation using a 1.5 Tesla–MR imaging system with head coil n=15 and neonate is in supine position.
Results: Cranial US detected intracranial pathologies in 62 neonates with periventricular echogenicity was the most frequent finding. Cranial MRI showed abnormal findings in 123 neonates. Bright basal ganglia (BG) were the most frequent finding and reported as the solitary lesion in 35 neonates, appeared voluminous in 9 neonates and associated with hemorrhage in 15 neonates. Bright basal ganglia was associated with thalamic affection in 33 neonates, 2 of them had only thalamic affection while the other 31 had other pathologies, 6 had internal capsule affection, 2 had internal capsule and pontine affection, 18 had internal capsule and pontine affection associated with hemorrhage, 4 had cerebellar affection and a neonate had internal capsule affection and white matter cystic changes. In 19 neonates bright basal ganglia (BG) were associated with variable findings including bright peri-rolandic cortex, dilated ventricles, slit-like ventricles, white matter abnormalities and agenesis of corpus callosum. 12 neonates showed varied pathologies mainly affecting cortex and/or white matter. Compared to cranial MRI, cranial US showed sensitivity rate of 44.5%, specificity rate of 71%, positive and negative predictive values of 85.4% and 25%, respectively and 50% accuracy rate as a diagnostic procedure for intracranial pathology in HIE neonates. Some cranial US examinations despite being true positive were misleading concerning the missed cortical pathology and other brain affections as detected by cranial MRI. Furthermore, four newborns with normal cranial US were found to have cortical affections. Cranial MRI defined intracranial bleeding during examination of 33 neonates, cranial US correctly defined the presence of hemorrhage in 20, but failed to define the site of hemorrhage in two of them.
Conclusion: Early radiological evaluation of newborns with HIE could define the presence and extent of intracranial pathology prior to its progression thus allowing earlier interference and cranial MRI is superior to cranial US as an evaluation modality for the more detailed examination with special concern of cortical affections and intracranial bleeding. Diffusion-weighted images allowed recognition of white matter affections.
|
