Data-driven quantification and prediction of pre-surgical local head volume distributions and post-surgical development in patients with craniosynostosis

Project summary Children with craniosynostosis typically undergo surgical treatment to remove the brain growth constraints and correct for the malformations produced a volume overgrowth parallel to the fused sutures that compensates for the local growth restrictions. However, there is a large variability of surgical techniques and outcomes among institutions, and it is common for patients with suboptimal treatments to require additional invasive surgeries for three main reasons: (1) local volume anomalies and their progression have not been characterized to estimate how much local volume patients need during treatment; (2) our knowledge about how treatment modifies cranial growth is limited and post-surgical growth predictions are not possible; and (3) there are no objective and personalized methods to evaluate long-term outcomes, so treatment selection remains subjective. Although our team and others have created methods to quantify cranial and head shape anomalies using imaging data, existing methods cannot characterize the abnormal patterns of local volume development in craniosynostosis, are age-agnostic, and do not account for sex, which is an essential modulator of development. Moreover, our recently created data-driven normative cranial bone development model cannot be used to predict growth of a surgically modified cranium, and large post-surgical datasets have not been traditionally available to characterize volume development after treatment of patients with craniosynostosis. This limited knowledge about the abnormal pre- and post-surgical local volume development has hindered the translation of existing methods to plan, quantify, compare and predict surgical outcomes. Hence, treatment selection remains subjective and the identification of relapsing patients still relies on subjective interpretation of variable clinical symptoms with low predictive value. We demonstrated that CT and 3D photogrammetry provide the same quantification of head anatomy, and Children’s Hospital Colorado incorporated pre- and post-surgical 3D photogrammetry acquisition in the standard clinical protocol of patients with craniosynostosis. In this project, we will leverage our retrospective dataset to characterize local head volume development in patients with craniosynostosis before surgery, study how treatment modifies growth patterns and identify relapsing patients using objective data. The main goals of this project are: (1) to quantitatively characterize the local head volume distributions and their temporal progression in patients with craniosynostosis; and (2) to predict and evaluate local head volume growth after treatment of craniosynostosis. This project will create the necessary objective and quantitative knowledge to achieve personalized optimal treatments, and software tools to objectively evaluate patients using non-invasive 3D photogrammetry before and after surgical treatment.