Background and aims Preterm infants are at risk of brain injury. Cranial ultrasound is frequently used in neonatal care to detect and monitor brain injury. Anatomical structures and abnormalities can be distinguished by differences in echogenicity.

Our primary objective was to reliably measure sonographic grey values in basal ganglia. Secondary objectives included the influence of gestational age at birth on echogenicity and aspects of deep grey matter change at 30 weeks corrected GA.

Methods We prospectively collected CUS-data of 229 preterm infants (<29 weeks gestation). Parasagittal images through the gangliothalamic ovoid were assessed on mean grey value in putamen and globus pallidus. Intra- and interobserver for placement of ROI were analysed.

Results The method used produced a reliable globus pallidus to putamen ratio (GPP ratio). Mean GPP ratio was 0.786 (±0.085). Extreme preterm infants have significantly lower GPP at birth than did preterm infants above 28 weeks (0.755 ± 0.081 vs 0.808 ± 0.091; P-value <0.01). At 30 weeks corrected GA this was still the case (0.723 ± 0.051 vs. 0.818 ± 0.063; P-value <0.01).

Conclusion The putamen of extremely preterm infants is more hyperechoic then putamen of preterm infants of 29 weeks of gestation. Objective measurement of grey values can help to study brain injury.