Related Experiment Video
Updated: Aug 9, 2025
10:42
A Postoperative Evaluation Guideline for Computer-Assisted Reconstruction of the Mandible
Published on: January 28, 2020
6.6K
Reconstruction of the mandible from partial inputs for virtual surgery planning
Ryan L Gillingham1, Tinashe E M Mutsvangwa2, Johan van der Merwe1
1Department of Mechanical & Mechatronic Engineering, University of Stellenbosch, Stellenbosch, 7600, South Africa.
Medical Engineering & Physics
|February 15, 2023
Summary
Statistical Shape Models (SSMs) offer superior accuracy for virtual surgery planning (VSP) in mandibular reconstruction compared to Sparse Prediction Models (SPMs). SSMs are recommended when patient anatomy is not typical.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Computer-Aided Surgery
Background:
- Virtual surgery planning (VSP) is crucial for reconstructing mandibular defects.
- Accurate prediction of anatomical form is essential for successful VSP.
- Statistical Shape Models (SSMs) and Sparse Prediction Models (SPMs) are emerging techniques for VSP.
Purpose of the Study:
- To compare the accuracy of SSMs and SPMs against standard mirroring techniques for VSP of mandibular defects.
- To evaluate the models' ability to reproduce clinically relevant metrics.
Main Methods:
- CT scans from 100 individuals were used, split into 80:20 training/testing sets.
- SSMs were developed using principal component analysis on segmented meshes.
- SPMs were constructed using regressions between cephalometric measurements.
Main Results:
- SSMs demonstrated lower measurement and surface-to-surface errors compared to SPMs and mirroring.
- Measurement errors for SSMs ranged from 0.45˚ to 3.67˚ and 0.66 mm to 2.54 mm.
- Surface-to-surface errors for SSMs were 1.06 mm to 1.33 mm.
Conclusions:
- SSMs are recommended for VSP of mandibular defects, especially when normal patient anatomy is absent.
- SSMs provide more accurate and reliable results for virtual surgical planning.
- Further research may explore hybrid approaches combining SSMs and SPMs.
