Related Experiment Video
Updated: May 14, 2025
13:40
Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking
Published on: December 16, 2010
16.6K
EEG-Based Engagement Monitoring in Cognitive Games.
Yusuf Ahmed1,2, Martin Ferguson-Pell3, Kim Adams3
1Department of Occupational Therapy, Faculty of Rehabilitation Medicine, University of Alberta, 8205 114 St NW, Edmonton, AB T6G 2G4, Canada.
Sensors (Basel, Switzerland)
|April 12, 2025
Summary
This study shows that electroencephalography (EEG) can accurately measure user engagement in cognitive rehabilitation games for both young and older adults. This technology offers a promising alternative to traditional questionnaires for assessing engagement levels.
Area of Science:
- Neuroscience
- Human-Computer Interaction
- Gerontology
Background:
- Cognitive decline and dementia prevention are critical global health concerns.
- Cognitive rehabilitation games show promise but depend on user engagement for efficacy.
- Assessing engagement via questionnaires is difficult for individuals with cognitive impairments.
Purpose of the Study:
- To investigate the link between video game difficulty, electroencephalography (EEG) engagement indices, and self-reported flow states.
- To develop a machine learning algorithm for classifying user engagement into high and low states.
- To explore the feasibility of using EEG for engagement analysis in adults and older adults.
Main Methods:
- Twenty-seven participants, including older adults, played a video game at varying difficulty levels while EEG data was collected using EPOCX.
- Participants completed the flow state scale for occupational tasks after each difficulty level.
- Machine learning algorithms were developed to classify user engagement based on EEG signals.
Main Results:
- Self-reported engagement scores were highest at the optimal game difficulty level (p = 0.027).
- A combination of three EEG indices yielded the highest classification performance, with F1 scores of 89% (within-subject) and 81% (cross-subject).
- Engagement classification achieved F1 scores of 90% for young adults and 85% for older adults.
Conclusions:
- EEG data can be effectively utilized for engagement analysis in cognitive rehabilitation games across different age groups.
- Machine learning models based on EEG show high accuracy in distinguishing between high and low user engagement.
- This approach offers a more objective and accessible method for assessing engagement in populations with cognitive decline.
