Loading...

Designing EEG-based Deep Neural Network for Analysis of Functional and Effective Brain Connectivity

Shoushtari, Shirin | 2021

293 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54251 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Mohammadzadeh, Hoda; Amini, Arash
  7. Abstract:
  8. Brain states analysis during consciousness is emerging research in brain-computer interface(BCI). Emotion recognition can be applied to learn brain states and stages of neural activities. Therefore, emotion recognition is crucial to the analysis of brain functioning. Electrical signals such as electroencephalogram (EEG), electrocardiogram (ECG) and functional magnetic resonance imaging(fMRI) are frequently used in emotion recognition researches. Convenience in recording, non-invasive nature and high temporal resolution are the factors that have made EEG popular in brain researches. EEG can be used to identify brain region activity solely or the connectivity of various regions in time in the emotion recognition issues. Connectivity between regions can be explained in three categories. The first category of brain connectivity is structural connectivity which describes the connectome between neighboring neurons. Functional and effective connectivity indicates a temporal and causal dependency between different regions of the brain. Brain connectivity is used in various applications including Autism and Alzheimer's detection and emotion recognition. The goal of this study is emotion recognition using functional and effective connectivity in time and deep neural networks. we have utilized EEG signals with visual stimuli to explore the importance of brain connectivity in emotion recognition. Since EEG are affected by artifacts and noise, we have used appropriate preprocessing before feature extraction. We have also compared the role of features from excluded brain regions and connectivity among brain regions. Using neural networks in feature extraction and classification can enhance accuracy significantly. Convolutional neural networks(CNN) can be utilized to model electrodes position in the feature extraction process and recurrent neural networks(RNN) can decode temporal changes of brain connectivity. Since changes of brain connectivity are associated with different emotions, graph analysis of brain connectivity is used to model connectivity patterns in time. Graph analysis of brain connectivity is a novel method that can enhance the accuracy of the classification and reduce calculations
  9. Keywords:
  10. Emotion Recognition ; Deep Neural Networks ; Functional Brain Connectivity Network ; Brain-Computer Interface (BCI) ; Functional Magnetic Resonance Imaging (FMRI) ; Electroencephalography

 Digital Object List

 Bookmark