Cranial ultrasound is a reliable first step imaging in children with suspected craniosynostosis

L Pogliani1, G V Zuccotti2, M Furlanetto3

  • 1Department of Paediatrics, University of Milan, Luigi Sacco Hospital, Via GB Grassi 74, Milan, Italy. laura_pogliani@libero.it.

Insights

Cranial ultrasound scan (CUS) is a highly accurate and sensitive tool for diagnosing craniosynostosis in infants under one year. This method avoids radiation exposure, making it a reliable first-step imaging evaluation for suspected cases.

Area of Science:

  • Pediatric Radiology
  • Medical Imaging
  • Neonatal Care

Background:

  • Craniosynostosis (CS) diagnosis in infants often relies on skull radiography (SR) and computed tomography (CT), exposing them to ionizing radiation.
  • Ultrasound studies suggest a potential role for this modality in CS diagnosis.
  • There is a need for safer, reliable imaging techniques for early CS detection.

Purpose of the Study:

  • To assess the diagnostic accuracy of cranial ultrasound scan (CUS).
  • To determine if CUS is a reliable first-step imaging evaluation for craniosynostosis in newborns.
  • To reduce ionizing radiation exposure in infants with abnormal head shapes.

Main Methods:

  • A cohort of 196 infants with suspected craniosynostosis underwent CUS.
  • Infants with confirmed CUS findings were referred for volumetric CT scan.
  • Infants with negative CUS or low clinical suspicion underwent clinical follow-up.

Main Results:

  • CUS demonstrated high specificity and sensitivity in diagnosing craniosynostosis.
  • Two false positives were noted in the initial phase, highlighting a learning curve.
  • CT confirmed CUS findings in diagnosed cases; negative CUS correlated with normal head shape evolution in follow-up.

Conclusions:

  • CUS is a highly specific and sensitive imaging technique for craniosynostosis screening in infants under one year.
  • Expert use of CUS can serve as a reliable first-step evaluation, avoiding radiation.
  • Centralization of cases and centralization are recommended due to the operator-dependent nature and learning curve of CUS.
Abstract