Background/aims Brain magnetic resonance imaging (MRI) is the reference technique for the assessment of qualitative injuries and cerebral growth. However cerebral ultrasonography (CUS) remains the standard follow-up imaging for most neonatologists. The aim of the study is to compare term corrected age MRI to a protocol combining term and neonatal CUS (tCUS/nCUS).
Methods Between January 2013 and March 2014, preterm (PT) ≤30 gestational weeks (GW) received the same day term corrected age MRI and tCUS in addition to classic nCUS. Supratentorial (intraventricular haemorrhage (IVH), nodular and cystic periventricular leukomalacia (PVL)) and cerebellar lesions were assessed separately. Corpus callosum length (CCL), a myelination marker, was measured on MRI and tCUS. Sensitivity (Se) and specificity (Sp) were calculated for qualitative lesions, metrics were compared by linear regression.
Results The study included prospectively 102 PT. Mean GW and birth weight were 26.9 (±2.2) and 1127 g (±313). Table described the results. Linear regression coefficient for CCL was 0.7.
Conclusion Additional tCUS raises diagnosis rate of mild IVH and cystic PVL and is reliable for CCL measures. However MRI remains superior for nodular white matter and cerebellar lesions.