An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease
In the diagnosis of Alzheimer’s Disease (AD), the brain network analysis method is often used. The traditional network can only reflect the pairwise association between two brain regions, but ignore the higher-order relationship between them. Therefore, a brain network construction method based on h...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/12/11/2632 |
| _version_ | 1797468572856877056 |
|---|---|
| author | Xinlei Wang Junchang Xin Zhongyang Wang Chuangang Li Zhiqiong Wang |
| author_facet | Xinlei Wang Junchang Xin Zhongyang Wang Chuangang Li Zhiqiong Wang |
| author_sort | Xinlei Wang |
| collection | DOAJ |
| description | In the diagnosis of Alzheimer’s Disease (AD), the brain network analysis method is often used. The traditional network can only reflect the pairwise association between two brain regions, but ignore the higher-order relationship between them. Therefore, a brain network construction method based on hypergraph, called hyperbrain network, is adopted. The brain network constructed by the conventional static hyperbrain network cannot reflect the dynamic changes in brain activity. Based on this, the construction of a dynamic hyperbrain network is proposed. In addition, graph convolutional networks also play a huge role in AD diagnosis. Therefore, an evolving hypergraph convolutional network for the dynamic hyperbrain network is proposed, and the attention mechanism is added to further enhance the ability of representation learning, and then it is used for the aided diagnosis of AD. The experimental results show that the proposed method can effectively improve the accuracy of AD diagnosis up to 99.09%, which is a 0.3 percent improvement over the best existing methods. |
| first_indexed | 2024-03-09T19:09:22Z |
| format | Article |
| id | doaj.art-688f889deea842e3a2f378af9fd43f8e |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-09T19:09:22Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-688f889deea842e3a2f378af9fd43f8e2023-11-24T04:18:24ZengMDPI AGDiagnostics2075-44182022-10-011211263210.3390/diagnostics12112632An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s DiseaseXinlei Wang0Junchang Xin1Zhongyang Wang2Chuangang Li3Zhiqiong Wang4School of Computer Science and Engineering, Northeastern University, Shenyang 110169, ChinaSchool of Computer Science and Engineering, Northeastern University, Shenyang 110169, ChinaSchool of Computer Science and Engineering, Northeastern University, Shenyang 110169, ChinaCollege of Medicine and Biological Information Engineering, Northeastern University, Shenyang 110169, ChinaCollege of Medicine and Biological Information Engineering, Northeastern University, Shenyang 110169, ChinaIn the diagnosis of Alzheimer’s Disease (AD), the brain network analysis method is often used. The traditional network can only reflect the pairwise association between two brain regions, but ignore the higher-order relationship between them. Therefore, a brain network construction method based on hypergraph, called hyperbrain network, is adopted. The brain network constructed by the conventional static hyperbrain network cannot reflect the dynamic changes in brain activity. Based on this, the construction of a dynamic hyperbrain network is proposed. In addition, graph convolutional networks also play a huge role in AD diagnosis. Therefore, an evolving hypergraph convolutional network for the dynamic hyperbrain network is proposed, and the attention mechanism is added to further enhance the ability of representation learning, and then it is used for the aided diagnosis of AD. The experimental results show that the proposed method can effectively improve the accuracy of AD diagnosis up to 99.09%, which is a 0.3 percent improvement over the best existing methods.https://www.mdpi.com/2075-4418/12/11/2632Alzheimer’s diseasehyperbrain networkevolving hypergraph convolutional networkattention mechanism |
| spellingShingle | Xinlei Wang Junchang Xin Zhongyang Wang Chuangang Li Zhiqiong Wang An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease Diagnostics Alzheimer’s disease hyperbrain network evolving hypergraph convolutional network attention mechanism |
| title | An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease |
| title_full | An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease |
| title_fullStr | An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease |
| title_full_unstemmed | An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease |
| title_short | An Evolving Hypergraph Convolutional Network for the Diagnosis of Alzheimer’s Disease |
| title_sort | evolving hypergraph convolutional network for the diagnosis of alzheimer s disease |
| topic | Alzheimer’s disease hyperbrain network evolving hypergraph convolutional network attention mechanism |
| url | https://www.mdpi.com/2075-4418/12/11/2632 |
| work_keys_str_mv | AT xinleiwang anevolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT junchangxin anevolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT zhongyangwang anevolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT chuangangli anevolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT zhiqiongwang anevolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT xinleiwang evolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT junchangxin evolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT zhongyangwang evolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT chuangangli evolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease AT zhiqiongwang evolvinghypergraphconvolutionalnetworkforthediagnosisofalzheimersdisease |