Advisor: Brani Vidakovic, PhD (ISyE/BME, Georgia Tech)
Co-Advisor: Douglas Robertson, MD, PhD (BME, Georgia Tech/Emory)
Gary May, PhD (ECE, Georgia Tech)
Barbara Boyan, PhD (VCU)
Joseph Williams, MD (CHOA)
Title: Craniosynostosis developmental quantification and severity assessment
Quantifying disease progression is one the most difficult challenges facing physicians as they look to provide the most effective treatment customized for individual patients especially in the case of surgical intervention. In more severe cases of craniosynostosis, when the cranial sutures between the bone plates prematurely fuse causing a constriction of skull development, surgery is often pursued to correct the deformity, but the optimal timing and technique of this surgery is not well characterized as well as the cause of postoperative complications.
The goal of this project is to develop computational algorithms that provide a comprehensive analysis of individual pediatric skulls in preparation for surgery using their CT scans and provide insights into what makes specific cases more severe than others. These same analytics are then used more broadly to evaluate normal and abnormal cranial development in the different forms of craniosynostosis. Then utilizing these cranial features and clinical information from a large patient population, predictive models will be developed for the identification of risk factors and severity markers related to postoperative intracranial pressure and resynostosis of the fused suture or other sutures. These outcomes are especially important because patients with these complications frequently seek successive follow-up surgeries for additional correction.
This comprehensive analysis of the skull is proposed through characterizing the intracranial volume asymmetries and measuring the cranial sutures from preoperative CT scans for synostosis differentiation and severity. Then the bone plates, separated by their corresponding sutures, are characterized using maximum intensity projections with wavelet statistical image improvement techniques to help identify signs of high underlying intracranial pressures along with three-dimensional surface fitting and vector analysis to quantify bone thickness around the skull. These techniques combine to provide an extensive assessment of the skull for further understanding of craniosynostosis and an identification of the cases that pose the greatest risk.