Graph Neural Networks (GNNs) have led to state-of-the-art performance on a variety of machine learning tasks such as recommendation, node classification and link prediction. Graph neural network models generate node embeddings by merging nodes features with the aggregated neighboring nodes information. Most existing GNN models exploit a single type of aggregator (e.g., mean-pooling) to aggregate neighboring nodes information, and then add or concatenate the output of aggregator to the current representation vector of the center node. However, using only a single type of aggregator is difficult to capture the different aspects of neighboring information and the simple addition or concatenation update methods limit the expressive capability of GNNs. Not only that, existing supervised or semi-supervised GNN models are trained based on the loss function of the node label, which leads to the neglect of graph structure information. In this paper, we propose a novel graph neural network architecture, Graph Attention & Interaction Network (GAIN), for inductive learning on graphs. Unlike the previous GNN models that only utilize a single type of aggregation method, we use multiple types of aggregators to gather neighboring information in different aspects and integrate the outputs of these aggregators through the aggregator-level attention mechanism. Furthermore, we design a graph regularized loss to better capture the topological relationship of the nodes in the graph. Additionally, we first present the concept of graph feature interaction and propose a vector-wise explicit feature interaction mechanism to update the node embeddings. We conduct comprehensive experiments on two node-classification benchmarks and a real-world financial news dataset. The experiments demonstrate our GAIN model outperforms current state-of-the-art performances on all the tasks.
Objectives
This study aimed to classify the alveolar ridge of the anterior maxillary edentulous and investigate the incidence of perforation and associated risk factors via virtual implant placement using cone beam computed tomography (CBCT).
Methods
The morphology of 85 patients who have lost a single tooth in the maxillary esthetic zone was assessed by CBCT. The width and height of the residual crest alveolar bone were measured. Root form implants (3.3 mm*10 mm and 3.3 mm*12 mm) were placed virtually in the edentulous area, and the risk factors associated with perforation were analyzed.
Results
Class Ib was the most common type of ridge (n = 26; 30.6%). Concavity bone thickness (CT) was significantly different (P < 0.05) among the four types of alveolar bone. The long axis angle of alveolar bone (LAAB) and the differences between the LAAB and implant placement angle (IPA) in the nonperforation areas were significantly lower compared with those in the perforation areas. After implantation, the overall occurrence of labial bone perforation was 56.5%, and perforation mostly occurred in Class II ridges.
Conclusions
Class II ridges and lateral incisors were at relatively high risk for labial perforation after implant placement. We thus recommend that clinicians determine the ridge classification, tooth type, CT, LAAB and IPA.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.