EUVIMEDEUVIMED entdecken
Open Access

Geo-Net: Geometry-Guided Pretraining for Tooth Point Cloud Segmentation

Y Liu iD, X Liu, C Yang, Y Yang iD, H Chen iD, Y Yuan

Journal of Dental Research · 2024

DOI: 10.1177/00220345241292566 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Dental Research, 2024. Verfasst von Y Liu und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong SAR, PR China. Hauptthemen laut MEDLINE: Algorithms, Imaging, Three-Dimensional. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 24 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Research Grants Council, University Grants Committee.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurately delineating individual teeth in 3-dimensional tooth point clouds is an important orthodontic application. Learning-based segmentation methods rely on labeled datasets, which are typically limited in scale due to the labor-intensive process of annotating each tooth. In this article, we propose a self-supervised pretraining framework, named Geo-Net, to boost segmentation performance by leveraging large-scale unlabeled data. The framework is based on the scalable masked autoencoders, and 2 geometry-guided designs, curvature-aware patching algorithm (CPA) and scale-aware reconstruction (SCR), are proposed to enhance the masked pretraining for tooth point cloud segmentation. In particular, CPA is designed to assemble informative patches as the reconstruction unit, guided by the estimated pointwise curvatures. Aimed at equipping the pretrained encoder with scale-aware modeling capacity, we also propose SCR to perform multiple reconstructions across shallow and deep layers. In vitro experiments reveal that after pretraining with large-scale unlabeled data, the proposed Geo-Net can outperform the supervised counterparts in mean Intersection of Union (mIoU) with the same amount of annotated labeled data. The code and data are available at https://github.com/yifliu3/Geo-Net .

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

AlgorithmsImaging, Three-DimensionalHumansToothImage Processing, Computer-AssistedNeural Networks, Computer

Stichwörter der Autor:innen

artificial intelligencecomputer vision/convolutional neural networksdeep learning/machine learningdental anatomydental informatics/bioinformaticselectronic dental records

Autor:innen und Institutionen

  • Department of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong SAR, PR China.
  • Faculty of Dentistry, The University of Hong Kong, Hong Kong SAR, PR China.

ORCID-Profile: Y Liu, Y Yang, H Chen

Förderung

  • Research Grants Council, University Grants Committee

Zitieren

APA

Liu, Y., Liu, X., Yang, C., Yang, Y., Chen, H., & Yuan, Y. (2024). Geo-Net: Geometry-Guided Pretraining for Tooth Point Cloud Segmentation. Journal of Dental Research. https://doi.org/10.1177/00220345241292566

Vancouver

Liu Y, Liu X, Yang C, Yang Y, Chen H, Yuan Y. Geo-Net: Geometry-Guided Pretraining for Tooth Point Cloud Segmentation. Journal of Dental Research. 2024. doi:10.1177/00220345241292566

BibTeX
@article{Liu2024,
  author = {Liu, Y and Liu, X and Yang, C and Yang, Y and Chen, H and Yuan, Y},
  title = {Geo-Net: Geometry-Guided Pretraining for Tooth Point Cloud Segmentation},
  journal = {Journal of Dental Research},
  year = {2024},
  doi = {10.1177/00220345241292566}
}
Kostenlos starten →