Agentic data science with the Data Agent Kit
cloud.google.com·7h ago
AI Research
arXiv CS.AI·Aug 11·5 min read
Coarse-to-Fine Registration of Jawbone CT and Intraoral Scan Data Using GeDi and ICP with Pseudo-IOS Ground Truth
TL;DR
A novel registration method for jawbone CT and intraoral scan data was developed, achieving submillimeter accuracy using a combination of GeDi and ICP algorithms. This approach significantly outperformed traditional methods, particularly in the presence of metal artifacts.
✦ Why It Matters
Implement the GeDi+ICP method to enhance registration accuracy in dental imaging workflows today.
Key Takeaways
- Utilize the GeDi+ICP method for improved registration accuracy.
- Generate pseudo-IOS data for reliable evaluation of registration techniques.
- Be aware of metal artifacts but note their limited impact on GeDi+ICP.
- Conduct further tests across diverse jaw and artifact conditions for validation.
How It Works
The proposed method begins with GeDi, a local descriptor that provides a robust initial alignment of the jawbone CT and IOS data. This is followed by the ICP algorithm, which iteratively refines the alignment by minimizing the distance between corresponding points in the two datasets.
The pseudo-IOS framework allows for a controlled evaluation by creating a synthetic intraoral scan from the CT data, ensuring that the true transformation is known.