Skip to main content
. 2022 Sep 28;10(10):1892. doi: 10.3390/healthcare10101892

Table 7.

GNN selected studies.

Authors Name and Year Methods Results Authors Suggestions/
Conclusions
Zhang et al., (2021) [102] PointNet, DGCNN, PointNet++, PointCNN, MeshSegNet Accuracy = 95.25, mIoU = 88.99 TSGCNet cannot robustly handle special cases with 12 teeth.
Zheng et al., (2022) [103] Modified Dynamic Graph CNN (DGCNN) mIoU = 97.49, accuracy = 98.94 The proposed teeth segmentation is robust to rotten, missing, crowded, and ectopic-tooth cases.