Validation of an in-silico modelling platform for outcome prediction in spring assisted posterior vault expansion

Abstract

BackgroundSpring-Assisted Posterior Vault Expansion has been adopted at Great Ormond Street Hospital for Children, London, UK to treat raised intracranial pressure in patients affected by syndromic craniosynostosis, a congenital calvarial anomaly which causes premature fusion of skull sutures. This procedure aims at normalising head shape and augmenting intracranial volume by means of metallic springs which expand the back portion of the skull. The aim of this study is to create and validate a 3D numerical model able to predict the outcome of spring cranioplasty in patients affected by syndromic craniosynostosis, suitable for clinical adoption for preoperative surgical planning. MethodsRetrospective spring expansion measurements retrieved from x-ray images of 50 patients were used to tune the skull viscoelastic properties for syndromic cases. Pre-operative computed tomography (CT) data relative to 14 patients were processed to extract patient-specific skull shape, replicate surgical cuts and simulate spring insertion. For each patient, the predicted finite element post-operative skull shape model was compared with the respective post-operative 3D CT data. FindingsThe comparison of the sagittal and transverse cross-sections of the simulated end-of-expansion calvaria and the post-operative skull shapes extracted from CT images showed a good shape matching for the whole population. The finite element model compared well in terms of post-operative intracranial volume prediction (R2 = 0.92, p < 0.0001). InterpretationThese preliminary results show that Finite Element Modelling has great potential for outcome prediction of spring assisted posterior vault expansion. Further optimisation will make it suitable for clinical deployment.