Loading...
Search
Search in this resource
sort by
Send to Friend
 Digital Object List
 Bookmark
 Review List
  1. Simple search
  2. Search result
  3. ارزیابی بار ذهنی و تأثیر آن بر ارتباطات مغزی با پردازش توأم سیگنال‌ های EEG و fNIRS
ارزیابی بار ذهنی و تأثیر آن بر ارتباطات مغزی با پردازش توأم سیگنال‌ های EEG و fNIRS
145 viewed

ارزیابی بار ذهنی و تأثیر آن بر ارتباطات مغزی با پردازش توأم سیگنال‌ های EEG و fNIRS

محدثه قره محمدلو Qare Mohammadlou, Mohaddeseh

Mental Workload Assessment and its Effect on Brain Connectivities using Simulataneous EEG and fNIRS

Qare Mohammadlou, Mohaddeseh | 2024

145 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57377 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Shamsollahi, Mohammad Bagher
  7. Abstract:
  8. Mental Work Load (MWL) is a fundamental concept for examining human performance and the amount of psychological pressure. It represents the level of brain activity during task execution, an individual's ability to process information during work, and the mental and cognitive resources required to perform duties. The accurate detection of mental workload levels, as demonstrated in this research, is of paramount importance. It plays a crucial role in enhancing task performance in various fields, including the design and assessment of systems related to transportation vehicles, human-computer and brain-computer interfaces, and in educational settings. By evaluating mental workload, we can ensure that these systems are optimized for human performance, thereby improving safety, efficiency, and learning outcomes. In this research, mental workload assessment was conducted simultaneously using EEG and fNIRS signals (including oxygenated and deoxygenated hemoglobin concentration signals). Given the high temporal accuracy of the EEG signal but its susceptibility to noise, and the fNIRS signal's resistance to noise but low temporal accuracy, these two signals complement each other. The classification of low and high levels of mental workload was done using functional connectivity features (MSC, PLV, PLI, PCC, and SPCC), effective connectivity features (Granger causality, PDC, and DTF) from both EEG and fNIRS signals, and statistical features from the differential fNIRS signal. The results of workload classification in this study show that using the MSC functional connectivity feature for the EEG signal resulted in an accuracy of 70.6%. The accuracy of the OXY and DEOXY signals was 65.5% and 65.9%, respectively. Using both EEG and fNIRS signals together improved the accuracy to 73.3%, with an enhancement of about 3%. The MSC feature provided more favorable classification accuracy than other brain connectivity features. In combined classification using the PLI feature, accuracy improved by 8 to 12% compared to just EEG or fNIRS. Additionally, examining brain connectivity metrics showed increased values of PCC, PLV, and MSC in higher mental workload levels. In another section, which appears to have not been discussed in prior studies, the brain connectivity between the two modalities of EEG and fNIRS was calculated, examining the flow of information between the two signals. The results indicate that the information output from EEG to fNIRS is greater than the information input from fNIRS to EEG, suggesting that EEG has a greater influence on fNIRS
  9. Keywords:
  10. Electroencephalography ; Classification ; Machine Learning ; Brain Signal ; Mental Work Load (MWL) ; Effective Connectivity ; Brain Connectivity ; Functional Connectivity ; Functional Near-Infrared Spectroscopy

 Digital Object List

 Bookmark

No TOC