From Electroencephalogram Data to Brain Networks: Graph-Learning-Based Brain Disease Diagnosis

Brain networks are built according to the structures or neural activities of different brain regions, which can be modeled as complex networks. Many studies exploit brains from the perspective of graph learning to diagnose the nerve diseases of brains. However, many of these algorithms are unable to automatically construct brain function topology based on electroencephalogram (EEG) and fail to capture the global features of multichannel EEG signals for whole-graph embedding. To address these challenging issues, we propose an attention-based whole-graph learning model for the diagnosis of brain diseases, namely, MAINS, which can adaptively construct brain functional topology from EEG signals and effectively embed multiple node features and the global structural features of brain networks into the whole-graph representations. We validated the model by conducting classification (diagnosis) experiments on real EEG datasets. Comprehensive experimental results demonstrate the superiority of the proposed approach over state-of-the-art methods.

An electroencephalogram (EEG) is a recording of electrical activity in the brain used to diagnose neurological disorders, which are medically defined as those involving the brain, spinal cord, nerves, and muscles. However, research statistics1 establish that 50% of EEG records in epilepsy patients are misdiagnosed as healthy by experts, leading to delays in medical care and significant risks to patients. There remains a need to develop automatic analysis algorithms and efficient tools for EEG signals to achieve rapid analysis and auxiliary diagnosis of neurological diseases. Besides EEG datasets, numerous research groups share datasets of brains in other formats, including magnetic resonance imaging (MRI), diffusion tensor imaging (DTI), computed tomography of brains, etc. The sharing of these images and network datasets makes it possible for network scientists to mine complex brain networks and promote the rapid development of network neuroscience.