Feasibility and reproducibility of shear wave elastography in pediatric cranial ultrasound

Head ultrasound (US) is commonly used to evaluate the neonatal brain but may be limited by its lack of sensitivity and specificity. Ultrasound shear wave elastography (SWE) might provide additional information to conventional gray-scale imaging.

To assess whether SWE of brain parenchyma can be (1) successfully performed at a large academic medical center where US technologists perform the majority of examinations and (2) used to detect intracranial pathology.

Pediatric patients undergoing head ultrasound underwent simultaneous SWE examination. We included normal examinations (n=70) and those with intracranial pathology (n=8) for analysis. We analyzed inter-reader variability and examination success rates and assessed the stiffness of white matter and deep gray nuclei in normal and pathological states across multiple gestational age groups.

Average gestational age of the term, pre-term and extreme pre-term groups were 38.4±1.2 weeks, 29.0±3.7 weeks and 28.3±3.1 weeks, respectively. Overall examination success rate was 79.5%. We observed a decrease in the SWE examination time from the first month (5.9±3.7 min) to the second month (4.1±1.7 min; P=0.01). Forty-one repeat examinations were performed on 14 children by different technologists, with an intraclass correlation coefficient (ICC) of 0.91. Mean stiffness in the periventricular white matter was lower than in the deep gray nuclei in all gestational age groups: term group (1.3 m/s vs. 1.5 m/s, P<0.001), pre-term (1.3 m/s vs. 1.4 m/s P=0.12), and extremely preterm group (1.2 m/s vs. 1.4 m/s, P=0.001). Mean stiffness for the deep gray nuclei differed between the term (1.5±0.3 m/s) and pre-term (1.4±0.2 m/s) groups (P<0.01). No significant differences in white matter stiffness were seen in relation to gestational age. Infants with large intraparenchymal hemorrhage had increased white matter stiffness (1.3±0.1 m/s) and deep gray nuclei stiffness (1.6±0.2 m/s) compared to full-term infants with normal head ultrasounds. These differences approached statistical significance with P=0.09 and P=0.06, respectively.

We demonstrated that SWE performed by pediatric sonography technologists is reproducible. We found differences in stiffness between deep gray nuclei and periventricular white matter across multiple age groups.