Alzheimer's Disease Diagnosis Based on the EEG Analysis

Alzheimer's Disease (AD) is a neurodegenerative disease that usually hurts the central nervous system. By detecting and intervening in the early stage, AD patients could be prevented from memory loss and function decrease. Electroencephalography (EEG) is the recording of the brain's electric action, which reflects the activity of nervous in the brain. Recent studies showed that AD patients' brain action will change in the early stage, and EEG might become a potential marker for AD's early diagnosis. To record and analyze the abnormal transformation in the brain with EEG, researchers and engineers has developed approaches to extract and analyze EEG feature. For the feature extraction, power spectrum analysis, event-related potentials (ERPs) and connectivity analysis could be as the marker for AD diagnosis. And in the feature analysis, Deep learning algorithms such as convolutional neural network (CNN), transfer learning (TL) and generative adversarial network (GAN) could help the automatic diagnosis of AD in the clinic. Based on above review, this article compares advantages and disadvantages of the AD diagnosis based on EEG analysis. In the discussion, different approaches for EEG analysis used in the clinic for AD patients will be introduced and compared.