Madhurima Chetan, Akudziwe Mawere, Ludovica Griffanti, Gaurav Bhalerao, Thomas Nichols, Mark Jenkinson, Kari Ashton, Jayaratnam Jayamohan, Shailendra Magdum, Gregory Thomas, Rosanna C Ching, Steven Wall, David Johnson, Tim Lawrence, Natalie Voets, Fintan Sheerin
Patients with craniosynostosis are at risk of intracranial space constriction and raised intracranial pressure. CT imaging to determine cranial growth rates contributes to surgical decision-making. Exterior head circumference charts are still commonly used due to a sparsity of normative data reporting intracranial volume (ICV) growth patterns based on CT. In this retrospective study, we developed and validated a semiautomated approach to measure ICV from CT imaging. Using 1579 CTs from healthy children, we generated normative growth curves for ICV for male and female children aged 0 to 18 years. This constitutes the largest single-site pediatric CT-derived ICV cohort to date. ICV was then measured in 518 CTs in children with unoperated craniosynostosis and, using the nomograms, a corresponding ICV centile was calculated. Automatically calculated ICVs showed excellent agreement with ICVs segmented manually in a subset of patients (n=105, ICC: 0.981, P<0.001). Children with craniosynostosis under the age of 1 had higher ICV values than controls (males: W=18775, P<0.001; females: W=7681, P<0.001). There was a negative correlation between ICV centile and age in the craniosynostosis cohort, compared to the normal cohort (P=0.001). These normative growth curves for ICV enable rapid and accurate comparison of an individual patient's ICV relative to patterns of typical childhood head growth, as well as enabling the tracking of an individual patient's ICV and centile trends over time. In the future, this may help to address fundamental questions about head growth in different subtypes of craniosynostosis.