MeshSNet: Deep Multi-scale Mesh Feature Learning for End-to-End Tooth Labeling on 3D Dental Surfaces

Wang, Li; Wu, Tai-Hsien; Liu, Mingxia; Durán, Francisca; Ko, Ching Chang; Shen, Dinggang

doi:10.1007/978-3-030-32226-7_93

Cited by 30 publications

(35 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, their models were not comprehensively evaluated in terms of clinical utility and applicability. The model proposed by Lian et al (2019) was trained to handle only scans with regular morphology and without third molars, hence producing problematic segmentation for complicated cases, such as scans with missing teeth or third molars.…”

Section: Discussionmentioning

confidence: 99%

“…The network in Zanjani et al Zanjani, Moin, Verheij, et al 2019) considers geometric information such as face normals, but advanced information such as shapes or jaws is not included. Sun et al (2020) and Lian et al (2019) designed networks to learn from meshes or vertices, but their models' robustness and generalization ability are yet unsatisfied or not verified for clinical applications. In our study, we sample rich geometric information from meshes during preprocessing.…”

Section: Discussionmentioning

confidence: 99%

“…We compare our method with extensive baselines, including Xu et al (2018), the backbone networks in recent works (Lian et al 2019;Zanjani, Moin, Verheij, et al 2019)-specifically, PointNet (Qi et al 2017) and PointCNN (Li et al 2018)-and Dynamic Graph CNN (Wang et al 2019). The settings of the hyperparameters are listed in the Appendix.…”

Section: Methodsmentioning

confidence: 99%

“…Though with agreeable interpretability, these methods cannot generalize to various teeth morphologies and usually require human involvement for postcorrection. Current advances employed deep learning techniques to develop more generic solutions (Zhao et al 2006;Zou et al 2015;Li and Wang 2016;Xu et al 2018;Lian et al 2019;Zanjani, Moin, Verheij, et al 2019;Lian et al 2020;Sun et al 2020). These methods showed better performance than geometry-based approaches.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Toward Clinically Applicable 3-Dimensional Tooth Segmentation via Deep Learning

et al. 2021

View full text Add to dashboard Cite

Digital dentistry plays a pivotal role in dental health care. A critical step in many digital dental systems is to accurately delineate individual teeth and the gingiva in the 3-dimension intraoral scanned mesh data. However, previous state-of-the-art methods are either time-consuming or error prone, hence hindering their clinical applicability. This article presents an accurate, efficient, and fully automated deep learning model trained on a data set of 4,000 intraoral scanned data annotated by experienced human experts. On a holdout data set of 200 scans, our model achieves a per-face accuracy, average-area accuracy, and area under the receiver operating characteristic curve of 96.94%, 98.26%, and 0.9991, respectively, significantly outperforming the state-of-the-art baselines. In addition, our model takes only about 24 s to generate segmentation outputs, as opposed to >5 min by the baseline and 15 min by human experts. A clinical performance test of 500 patients with malocclusion and/or abnormal teeth shows that 96.9% of the segmentations are satisfactory for clinical applications, 2.9% automatically trigger alarms for human improvement, and only 0.2% of them need rework. Our research demonstrates the potential for deep learning to improve the efficacy and efficiency of dental treatment and digital dentistry.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Toward Clinically Applicable 3-Dimensional Tooth Segmentation via Deep Learning

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Considering the natural correlations between the two tasks (e.g., each tooth's landmarks depend primarily on its local geometry), a two-stage framework leveraging mesh deep learning (called TS-MDL) is proposed in this paper for joint In Stage 1, we propose an end-to-end deep neural network, called iMeshSegNet, to perform tooth segmentation on 3D intraoral scans. iMeshSegNet improves the implementation of the multi-scale graph-constrained learning module in its forerunner MeshSegNet [6], [7], a neural network for automatic tooth segmentation. In Stage 2, we extract cells that belong to individual teeth based on the segmentation results generated by Stage 1.…”

Section: Introductionmentioning

confidence: 99%

Two-Stage Mesh Deep Learning for Automated Tooth Segmentation and Landmark Localization on 3D Intraoral Scans

Wu,

Lian,

Lee

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Accurately segmenting teeth and identifying the corresponding anatomical landmarks on dental mesh models are essential in computer-aided orthodontic treatment. Manually performing these two tasks is timeconsuming, tedious, and, more importantly, highly dependent on orthodontists' experiences due to the abnormality and large-scale variance of patients' teeth. Some machine learning-based methods have been designed and applied in the orthodontic field to automatically segment dental meshes (e.g., intraoral scans). In contrast, the number of studies on tooth landmark localization is still limited. This paper proposes a two-stage framework based on mesh deep learning (called TS-MDL) for joint tooth labeling and landmark identification on raw intraoral scans. Our TS-MDL first adopts an end-to-end iMeshSegNet method (i.e., a variant of the existing MeshSegNet with both improved accuracy and efficiency) to label each tooth on the downsampled scan. Guided by the segmentation outputs, our TS-MDL further selects each tooth's region of interest (ROI) on the original mesh to construct a light-weight variant of the pioneering PointNet (i.e., PointNet-Reg) for regressing the corresponding landmark heatmaps. Our TS-MDL was evaluated on a real-clinical dataset, showing promising segmentation and localization performance. Specifically, iMeshSegNet in the first stage of TS-MDL reached an averaged Dice similarity coefficient (DSC) at 0.953 ± 0.076, significantly outperforming the original MeshSegNet. In the

show abstract