草榴社区 Cloud
Cloud native EDA tools & pre-optimized hardware platforms
Request a Free Trial →
Simpleware Case Study: Computational Modeling of Patient-Specific Craniosynostosis Correction
Overview
Lambdoid craniosynostosis (LC) is a rare non-syndromic craniosynostosis characterized by fusion at birth of the lambdoid sutures at the back of the head. While surgical options such as spring-assisted cranioplasty are available, the aesthetic results are often suboptimal. To tackle this problem, a parametric Finite Element (FE) model of the LC skulls was developed to optimize future spring surgeries.
Skull geometries from three different LC patients who had undergone spring correction were reconstructed from pre-operative CT data in Simpleware software. Simulation was carried out to compare the skull growth between the pre-operative CT imaging and the surgery. By reproducing this procedure in a virtual setting, this approach may help to improve LC surgeries in future.
Highlights
- FE models created in Simpleware software from CT images used to better understand craniosynostosis interventions
- Surface deviation in Simpleware CAD compares skull growth
- Simulation used to compare pre-op CT imaging and surgical intervention, and after surgical intervention
- Goal of improving future surgical practice
Thanks to
S. Bozkurt[1,2], A. Borghi[2,3], L.S. Van de Lande[2,3], N.U. Owase Jeelani[2,3], D.J. Dunaway[2,3], S. Schievano[2,3]
- , University College London, UK and
- , University College London, UK
- , Great Ormond Street Hospital for Children, UK
Computational modelling of patient specific spring assisted lambdoid craniosynostosis correction. Sci Rep 10, 18693 (2020).
Data Acquisition and Reconstruction
Three LC patients undergoing spring-assisted surgery at the Great Ormond Street Hospital Craniofacial Unit for abnormal skull shape were CT scanned before and after their procedures. The CT images were imported to Simpleware ScanIP software for reconstruction of the skull at both stages. Segmentation and processing were carried out to identify the bone of the calvarium to the maxilla and suture structures.
Figure 1. Patient-specific pre- and post-operative skull models reconstructed from the CT images using Simpleware software.
Meshing for Simulation
Finite Element (FE) models of the skull were generated in Simpleware FE using structural 3D tetrahedral elements, with materials modeled with appropriate material properties for different anatomical regions. Skull growth between the pre-operative imaging and surgery was simulated in MSC Marc by approximating growth using thermal expansion coefficients, before performing the osteotomies on the skull models. The intracranial volume (ICV) was used as the parameter to represent skull size at different stages, including from the pre-operative CT reconstructions in Simpleware ScanIP by selecting the internal surface of the cranial vault.
The osteotomies on the skulls at the time of surgery were replicated based on the estimated intra-operative ICV by following the traces remaining visible from the procedure in Simpleware ScanIP. The skull geometries with osteotomies were then re-meshed, and spring implantation simulated using spring/dashpot link elements in MSC Marc to show the difference between surgery and the post-operative CT scans.
Figure 2. The FE models simulating spring assisted cranial expansion with osteotomies.
Surface Deviation Analysis
Simpleware CAD was used to undertake surface deviation analysis between the expanded FE skull models and the post-operative CT skull reconstructions, after obtaining volume registration using landmarks on the anterior nasal spine and frontozygomatic sutures not affected by the surgery. The simulations were then performed iteratively using thermal FE models by adjusting parameters to account for average negative and positive surface deviations for the entire skull in each model.
Figure 3. Surface deviation between the FE models and post-operative skull models reconstructed from CT images.
Results
The reconstructed skull models show the shape of the intracranial cavities at the time of pre-operative CT, with the testing allowing for comparison between the three patients. Analysis between the FE models and the post-operative skull models from CT showed a relatively low of rate of surface deviation on the frontal and temporal bones, increasing on the posterior skull surfaces that were expanded by the springs.
Figure 4: Displacement maps for the skull models for the spring assisted cranioplasty and post-operative skull growth
Conclusions
This study was able to approach the complexity of skull growth patterns in craniosynostosis based on mechanical effects, comparing pre-operative CT imaging, surgical interventions, and post-surgical interventions. Simulation results were validated using post-operative reconstructions from CT images. The obtained simulation results demonstrate that the final shape of the skull after surgery depends on the performed osteotomies, with relatively longer cuts mostly resulting in in hinging and expansion in the cranium, while a minimal cut allows the gap between the edges of the osteotomy to enlarge. As a result, better understanding the size and locations of the osteotomies based on patient-specific findings is crucial.
Future research could, then, develop more complex skull models to take into account factors such as age and bone formation. The project shows the value of using parametric FE models to simulate surgical outcomes in LC corrected by spring-assisted cranioplasty. By reproducing these procedures, parameters can be tuned for surgical planning, with larger population studies possibly being able to determine a specific set of values for patients to use the model as a planning tool. Based on this model, future investigations of spring types and locations using parametric FE models could help optimize the function and aesthetic outcomes in LC surgical corrections.
Learn More
Read the full open access paper by Bozkurt, S., Borghi, A., van de Lande, L.S., Owase Jeelani, N.U., Dunaway, D.J., Schievano, S., 2020. . Scientific Reports, 10, 18693.
Any Questions?
Do you have any questions about this case study or how to use Simpleware software for your own workflows?