- 21 views
- 32 downloads
A Hybrid Brain-Computer Interface using P300 and Steady State Visual Evoked Potentials
-
- Author / Creator
- Goncalves Mussi, Matheus
-
BACKGROUND: Children experiencing neurological impairment can experience limitations in their functional abilities. For people with severe physical disabilities, brain-computer interfaces (BCI) are a potential solution to access computers when other assistive technologies prove to be inaccessible. Although BCIs can help individuals accomplish a number of activities, some traditional BCI methods yield insufficient performance to be used in online applications. Hybrid BCI (hBCI) systems aim to improve the system’s performance by combining brain signal paradigms, or brain signals with other inputs.
OBJECTIVES: The purpose of this study was to develop and test an EEG-based hBCI system using P300 and steady-state visual evoked potentials (SSVEP) simultaneously, and compare the performance of the developed hBCI against the pure P300 and SSVEP BCI in offline and online scenarios.
METHODS: This study validated the system and potential measures with adults without disabilities. It includes two parts. The system was developed in part 1 with eight neurotypical adults who tested the system at different stages of design. Using the user-centered design, the system was modified based on the volunteers’ opinions and the final system was used in part 2. Six different neurotypical adults, divided into two groups, tested the system in part 2. The participants performed six sessions over three weeks, two with each paradigm (P300, SSVEP, hybrid). The second group used a system modified slightly to improve performance. The performed task was programmed so that three targets flickered at different frequencies to generate the SSVEP response and frames appeared semi-randomly to generate the P300 response. The system was evaluated in accordance with Kübler’s usability measures of effectiveness, efficiency, and satisfaction. For the effectiveness, three types of accuracies were calculated during the sessions: offline, continuous and selection. For the efficiency, the response time for online sets was measured and the information transfer rate was calculated. For the satisfaction, the NASA TLX questionnaire was used to evaluate the workload of each paradigm.
RESULTS: For group 1, the average selection accuracy for the pure P300 was 83.33%±11.86, for the pure SSVEP it was 49.52% ±17.79, and for the hybrid it was 49.44% ±13.04. For group 2, the average selection accuracy for the pure P300 was 96.39% ±4.29, for the pure SSVEP it was 49.72% ±11.54, and for the hybrid it was 49.44% ±17.31. For group 1, the ITR for the pure P300 was 64.53 bits/min, for the pure SSVEP it was 6.73 bits/min, and for the hybrid it was 6.68 bits/min. For group 2, the ITR for the pure P300 was 111.27 bits/min, for the pure SSVEP it was 6.90 bits/min, and for the hybrid it was 6.68 bits/min. The workload was calculated for each system (0 lowest and 10 highest). The average final workload was 3.27±1.59 for the P300, 5.02±1.22 for the SSVEP and 5.36±1.49 for the hybrid. Post- analysis showed that the lower accuracy on the hybrid was a consequence of the lower accuracy of the SSVEP.
CONCLUSION: The hybrid combination of the P300 and SSVEP did not result in the expected improvement in this study. The attempt to use a short sampling window size of 0.5s might account for the SSVEP’s poor performance, which consequently negatively affected the hybrid performance. Although changes made from group 1 to group 2 improved the time response for all paradigms, the changes were unable to sufficiently improve the SSVEP accuracy. Recommendations to increase the accuracy of the system are suggested for future studies. -
- Subjects / Keywords
-
- Graduation date
- Spring 2022
-
- Type of Item
- Thesis
-
- Degree
- Master of Science
-
- License
- This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.