TUCNet: A channel and spatial attention-based graph convolutional network for teeth upsampling and completion

• Published in: Computers in biology and medicine Vol. 166; p. 107519
• Main Authors: Liu, Mengting, Li, Xiaojie, Liu, Jie, Liu, Weifan, Yu, Zekuan
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.11.2023; Elsevier Limited
• Subjects: Algorithms; Artificial neural networks; CAD; Channel and spatial attention; Computer aided design; Datasets; Deep learning; Dental materials; Dental prosthetics; Dentistry; Feature extraction; Humans; Imaging, Three-Dimensional - methods; Internal Medicine; Medical diagnosis; Modules; Neural Networks, Computer; Oral cavity; Orthodontics; Other; Point cloud completion; Symmetry; Teaching methods; Teeth; Teeth point cloud completion; Three dimensional models; Tooth - diagnostic imaging; Upsample transformer; Upsample transformer; Channel and spatial attention; Teeth point cloud completion; Point cloud completion
• DOI: 10.1016/j.compbiomed.2023.107519
• ISSN: 0010-4825; 1879-0534

With the increasing popularity of the use of 3D scanning equipment in capturing oral cavity in dental health applications, the quality of 3D dental models has become vital in oral prosthodontics and orthodontics. However, the point cloud data obtained can often be sparse and thus missing information. To address this issue, we construct a high-resolution teeth point cloud completion method named TUCNet to fill up the sparse and incomplete oral point cloud collected and output a dense and complete teeth point cloud. First, we propose a Channel and Spatial Attentive EdgeConv (CSAE) module to fuse local and global contexts in the point feature extraction. Second, we propose a CSAE-based point cloud upsample (CPCU) module to gradually increase the number of points in the point clouds. TUCNet employs a tree-based approach to generate complete point clouds, where child points are derived through a splitting process from parent points following each CPCU. The CPCU learns the up-sampling pattern of each parent point by combining the attention mechanism and the point deconvolution operation. Skip connections are introduced between CPCUs to summarize the split mode of the previous layer of CPCUs, which is used to generate the split mode of the current CPCUs. We conduct numerous experiments on the teeth point cloud completion dataset and the PCN dataset. The experimental results show that our TUCNet not only achieves the state-of-the-art performance on the teeth dataset, but also achieves excellent performance on the PCN dataset. •A novel convolutional network TUCNet for the completion of teeth point clouds.•A Channel and Spatial Attentive EdgeConv (CSAE) block for local feature extraction.•A CSAE-based point cloud upsample (CPCU) module for points upsampling.•State-of-the-art performance on the FDDI teeth dataset.