Towards a radiation free numerical modelling framework to predict spring assisted correction of scaphocephaly

Sagittal Craniosynostosis (SC) is a congenital craniofacial malformation, involving premature sagittal suture ossification; spring-assisted cranioplasty (SAC) – insertion of metallic distractors for skull reshaping – is an established method for treating SC. Surgical outcomes are predictable using numerical modelling, however published methods rely on computed tomography (CT) scans availability, which are not routinely performed. We investigated a simplified method, based on radiation-free 3D stereophotogrammetry scans. Eight SAC patients (age 5.1 ± 0.4 months) with preoperative CT and 3D stereophotogrammetry scans were included. Information on osteotomies, spring model and post-operative spring opening were recorded. For each patient, two preoperative models (PREOP) were created: i) CT model and ii) S model, created by processing patient specific 3D surface scans using population averaged skin and skull thickness and suture locations. Each model was imported into ANSYS Mechanical (Analysis System Inc., Canonsburg, PA) to simulate spring expansion. Spring expansion and cranial index (CI - skull width over length) at times equivalent to immediate postop (POSTOP) and follow up (FU) were extracted and compared with in-vivo measurements. Overall expansion patterns were very similar for the 2 models at both POSTOP and FU. Both models had comparable outcomes when predicting spring expansion. Spring induced CI increase was similar, with a difference of 1.2%±0.8% for POSTOP and 1.6%±0.6% for FU. This work shows that a simplified model created from the head surface shape yields acceptable results in terms of spring expansion prediction. Further modelling refinements will allow the use of this predictive tool during preoperative planning.Spring-assisted cranioplasty (SAC) –insertion of metallic distractors helping skull reshaping – is a method for treating sagittal craniosynostosis, caused by premature sagittal suture closure. We present a method for predicting SAC outcomes, relying on radiation-free 3D stereophotogrammetry scans. Eight patients with preoperative CT and 3D stereophotogrammetry scans were recruited; results of spring expansion simulation were compared between models created using CT versus 3D scan data. Expansion patterns and extent of reshaping were very similar. This work proves that SAC preoperative planning can be carried out using non-ionising imaging. Further modelling refinements will allow clinical adoption of this predictive tool.