Optimization Of 3d Reconstruction Surface Rendering Algorithm For Osferion Bone Void Filling
3D reconstruction visualizes the 3D models from 2D medical image slices, which is proven helpful to doctors and surgeons in diagnosing and surgical planning on OSferion bone defects, which is well known for its fast absorption rate. Among the 3D reconstruction algorithms, surface rendering algorithm...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | http://eprints.usm.my/60339/1/DANIEL%20CHIN%20JIE%20YUAN%20-%20TESIS24.pdf http://eprints.usm.my/60339/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | 3D reconstruction visualizes the 3D models from 2D medical image slices, which is proven helpful to doctors and surgeons in diagnosing and surgical planning on OSferion bone defects, which is well known for its fast absorption rate. Among the 3D reconstruction algorithms, surface rendering algorithms are more suitable for effectively visualizing the bones’ structure and shape. However, surface rendering algorithms have two main problems, the massive number of triangular patches generated during the reconstruction process and the slow reconstruction speed, especially in reconstructing huge medical image datasets. Also, with the attempt to generate rendering-device-agnostic models by reducing the 3D models’ file size, the surface of the models is easily deformed due to the reduced number of triangular patches. Thus, the objectives are to enhance the Marching Cubes or the Marching Tetrahedra algorithm for large CT/MRI datasets so that the reconstructed 3D models are rendering-device-agnostic and optimized and to improve the quality of the 3D models after reducing the number of vertices and faces so that the surface of the 3D models can be improved. The impact of this research includes 3D models that are rendering-device-agnostic so that doctors and surgeons have access to the 3D models at anytime, anywhere. The proposed improvement method, which is Marching Cubes with 3D data smoothing and surface smoothing box kernel size of 11, mesh decimation reduction factor of 0.1, successfully increased the reconstruction accuracy by 6.26%, decreased the number of vertices and faces by 89.82%, and decreased the reconstruction and rendering time by 52.45% and 90.74% seconds respectively |
|---|