EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods
Alzheimer’s disease (AD) is the leading form of dementia worldwide. AD disrupts neuronal pathways and thus is commonly viewed as a network disorder. Many studies demonstrate the power of functional connectivity (FC) graph-based biomarkers for automated diagnosis of AD using electroencepha...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9878348/ |
| _version_ | 1797805113920716800 |
|---|---|
| author | Dominik Klepl Fei He Min Wu Daniel J. Blackburn Ptolemaios Sarrigiannis |
| author_facet | Dominik Klepl Fei He Min Wu Daniel J. Blackburn Ptolemaios Sarrigiannis |
| author_sort | Dominik Klepl |
| collection | DOAJ |
| description | Alzheimer’s disease (AD) is the leading form of dementia worldwide. AD disrupts neuronal pathways and thus is commonly viewed as a network disorder. Many studies demonstrate the power of functional connectivity (FC) graph-based biomarkers for automated diagnosis of AD using electroencephalography (EEG). However, various FC measures are commonly utilised, as each aims to quantify a unique aspect of brain coupling. Graph neural networks (GNN) provide a powerful framework for learning on graphs. While a growing number of studies use GNN to classify EEG brain graphs, it is unclear which method should be utilised to estimate the brain graph. We use eight FC measures to estimate FC brain graphs from sensor-level EEG signals. GNN models are trained in order to compare the performance of the selected FC measures. Additionally, three baseline models based on literature are trained for comparison. We show that GNN models perform significantly better than the other baseline models. Moreover, using FC measures to estimate brain graphs improves the performance of GNN compared to models trained using a fixed graph based on the spatial distance between the EEG sensors. However, no FC measure performs consistently better than the other measures. The best GNN reaches 0.984 area under sensitivity-specificity curve (AUC) and 92% accuracy, whereas the best baseline model, a convolutional neural network, has 0.924 AUC and 84.7% accuracy. |
| first_indexed | 2024-03-13T05:47:14Z |
| format | Article |
| id | doaj.art-c78813f5390e42f08ef2c1bd9352730d |
| institution | Directory Open Access Journal |
| issn | 1558-0210 |
| language | English |
| last_indexed | 2024-03-13T05:47:14Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| spelling | doaj.art-c78813f5390e42f08ef2c1bd9352730d2023-06-13T20:09:10ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1558-02102022-01-01302651266010.1109/TNSRE.2022.32049139878348EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity MethodsDominik Klepl0https://orcid.org/0000-0001-7584-9074Fei He1https://orcid.org/0000-0001-9176-6674Min Wu2https://orcid.org/0000-0003-0977-3600Daniel J. Blackburn3https://orcid.org/0000-0001-8886-1283Ptolemaios Sarrigiannis4Centre for Computational Science and Mathematical Modelling, Coventry University, Coventry, U.K.Centre for Computational Science and Mathematical Modelling, Coventry University, Coventry, U.K.Institute for Infocomm Research, A*STAR, Connexis North Tower, SingaporeDepartment of Neuroscience, The University of Sheffield, Sheffield, U.K.Department of Neurophysiology, Royal Devon and Exeter NHS Foundation Trust, Exeter, U.K.Alzheimer’s disease (AD) is the leading form of dementia worldwide. AD disrupts neuronal pathways and thus is commonly viewed as a network disorder. Many studies demonstrate the power of functional connectivity (FC) graph-based biomarkers for automated diagnosis of AD using electroencephalography (EEG). However, various FC measures are commonly utilised, as each aims to quantify a unique aspect of brain coupling. Graph neural networks (GNN) provide a powerful framework for learning on graphs. While a growing number of studies use GNN to classify EEG brain graphs, it is unclear which method should be utilised to estimate the brain graph. We use eight FC measures to estimate FC brain graphs from sensor-level EEG signals. GNN models are trained in order to compare the performance of the selected FC measures. Additionally, three baseline models based on literature are trained for comparison. We show that GNN models perform significantly better than the other baseline models. Moreover, using FC measures to estimate brain graphs improves the performance of GNN compared to models trained using a fixed graph based on the spatial distance between the EEG sensors. However, no FC measure performs consistently better than the other measures. The best GNN reaches 0.984 area under sensitivity-specificity curve (AUC) and 92% accuracy, whereas the best baseline model, a convolutional neural network, has 0.924 AUC and 84.7% accuracy.https://ieeexplore.ieee.org/document/9878348/Alzheimer’s diseasegraph neural networkclassificationEEGfunctional connectivity |
| spellingShingle | Dominik Klepl Fei He Min Wu Daniel J. Blackburn Ptolemaios Sarrigiannis EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods IEEE Transactions on Neural Systems and Rehabilitation Engineering Alzheimer’s disease graph neural network classification EEG functional connectivity |
| title | EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods |
| title_full | EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods |
| title_fullStr | EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods |
| title_full_unstemmed | EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods |
| title_short | EEG-Based Graph Neural Network Classification of Alzheimer’s Disease: An Empirical Evaluation of Functional Connectivity Methods |
| title_sort | eeg based graph neural network classification of alzheimer x2019 s disease an empirical evaluation of functional connectivity methods |
| topic | Alzheimer’s disease graph neural network classification EEG functional connectivity |
| url | https://ieeexplore.ieee.org/document/9878348/ |
| work_keys_str_mv | AT dominikklepl eegbasedgraphneuralnetworkclassificationofalzheimerx2019sdiseaseanempiricalevaluationoffunctionalconnectivitymethods AT feihe eegbasedgraphneuralnetworkclassificationofalzheimerx2019sdiseaseanempiricalevaluationoffunctionalconnectivitymethods AT minwu eegbasedgraphneuralnetworkclassificationofalzheimerx2019sdiseaseanempiricalevaluationoffunctionalconnectivitymethods AT danieljblackburn eegbasedgraphneuralnetworkclassificationofalzheimerx2019sdiseaseanempiricalevaluationoffunctionalconnectivitymethods AT ptolemaiossarrigiannis eegbasedgraphneuralnetworkclassificationofalzheimerx2019sdiseaseanempiricalevaluationoffunctionalconnectivitymethods |