# Construction of unbiased dental template and parametric dental model for precision digital dentistry

**Authors:** Lei Ma, Jingyang Zhang, Ke Deng, Peng Xue, Zhiming Cui, Yu Fang, Minhui Tang, Yue Zhao, Min Zhu, Zhongxiang Ding, Dinggang Shen  
**arXiv:** [2304.03556](https://arxiv.org/abs/2304.03556)  
**Date:** 2023-04-07  
**Summary page:** https://www.summarizepaper.com/en/paper/2304.03556v1/

## Key points

- Dental templates and parametric dental models are essential tools in digital dentistry
- Constructing an unbiased dental template and accurate parametric dental models is challenging due to complex anatomical structures and low volume ratio of teeth
- Researchers developed an innovative approach to construct an unbiased dental template by accurately constructing a dental atlas from CBCT images with guidance from teeth segmentation
- The study utilized CBCT images of 159 subjects from their digital archive, with resolutions ranging from 0.1mm to 0.4mm
- Prior to construction of the dental atlas, pre-processing was performed on the CBCT images by standardizing their resolutions and normalizing voxel intensities
- An approach involving registration between enhanced segmentation images was proposed to enhance CBCT images and teeth segmentation for accurate atlas construction
- Parametric dental models were built using Principal Component Analysis (PCA) to determine shape subspaces of each tooth based on point-to-point correspondences between the models
- Contributions of this study include developing an unbiased dental template, proposing an approach for enhancing CBCT images and teeth segmentation, and constructing parametric dental models for various applications in digital dentistry.
- The developed dental template and parametric dental models are available on GitHub for public use.

## Summary

In the field of digital dentistry, dental templates and parametric dental models are essential tools for various applications. However, constructing an unbiased dental template and accurate parametric dental models remains a challenging task due to the complex anatomical and morphological dental structures, as well as the low volume ratio of teeth. To address these challenges, a team of researchers developed an innovative approach to construct an unbiased dental template by accurately constructing a dental atlas from cone-beam computed tomography (CBCT) images with guidance from teeth segmentation. The study utilized CBCT images of 159 subjects from their digital archive. The resolution of the CBCT images varied from 0.1mm to 0.4mm with X and Y dimensions both set to 400, while the Z dimension ranged from 256 to 328. Prior to the construction of the dental atlas, pre-processing was performed on the CBCT images by standardizing their resolutions to 0.4mm x 0.4mm x 0.4mm and normalizing voxel intensities of the images to range between zero and one. To enhance CBCT images and their teeth segmentation images through image cropping, image masking, and segmentation intensity reassigning, they proposed an approach that involved performing registration between enhanced segmentation images to guide registration between masked CBCT images further reducing influence on adjacent anatomical structures on atlas construction while improving accuracy in teeth region registration. Leveraging the constructed unbiased dental template, they built parametric dental models for dentition and each individual tooth by estimating point-to-point correspondences between the dental models using Principal Component Analysis (PCA) to determine shape subspaces of parametric dental models. The contributions of this study include developing an unbiased dental template by accurately constructing a dental atlas from CBCT images using enhanced segmentation techniques; proposing an approach that enhances CBCT images and teeth segmentation for accurate atlas construction; constructing parametric dental models by analyzing shape variations of dentition and every tooth using the unbiased dental template. The developed dental template and parametric dental models are available on GitHub for public use in various applications in digital dentistry.