Related Experiment Video
Updated: May 20, 2025
07:11
Treatment of Facial Deformities using 3D Planning and Printing of Patient-Specific Implants
Published on: May 23, 2020
7.2K
Facial surgery preview based on the orthognathic treatment prediction
Huijun Han1, Congyi Zhang2, Lifeng Zhu3
1Discipline of Orthodontics, Faculty of Dentistry, the University of Hong Kong, Hong Kong Special Administrative Region of China; Department of Computer Science and Engineering, Texas A&M University, TX, USA.
Computer Methods and Programs in Biomedicine
|April 26, 2025
Summary
This study introduces an automated pipeline for 3D facial previews after orthognathic surgery, improving visualization accuracy without extra medical images. The method enhances patient consultations by providing realistic postsurgical appearance predictions.
Area of Science:
- Medical imaging and computer vision
- 3D reconstruction and visualization
- Surgical planning and simulation
Background:
- Orthognathic surgery consultations require accurate facial appearance change visualization.
- Current methods are often inefficient and inaccurate due to data limitations and treatment complexity.
- A need exists for automated, precise 3D previews of postsurgical facial appearances.
Purpose of the Study:
- To develop a fully automated pipeline for accurate and efficient 3D postsurgical facial appearance previews.
- To generate previews without requiring additional medical images.
- To incorporate novel aesthetic criteria for improved prediction accuracy.
Main Methods:
- Developed an automated pipeline incorporating aesthetic criteria (mouth-convexity, asymmetry).
- Implemented a robust data augmentation scheme to address data limitations.
- Evaluated performance using Chamfer and Hausdorff distances, and conducted a user study with medical professionals and engineers.
Main Results:
- Achieved high prediction accuracy (Hausdorff Distance: 9.00 mm, Chamfer Distance: 2.50 mm), outperforming state-of-the-art.
- Demonstrated competitive results even without synthesized data (Hausdorff Distance: 9.43 mm, Chamfer Distance: 2.94 mm).
- User study showed no significant difference in distinguishing generated faces from real outcomes (p > 0.05), indicating robustness.
Conclusions:
- The proposed method offers a practical and effective solution for orthognathic surgery consultations.
- Improves efficiency and accuracy of 3D postsurgical facial appearance previews for doctors and patients.
- Has potential for practical application in pre-surgical visualization and decision-making.
