Using Implicit Interaction to Enhance the Applicability of EEG-based Brain-Computer Interface for Untrained Users


Student thesis: Doctoral Thesis

View graph of relations


Related Research Unit(s)


Awarding Institution
  • MoBen BENAYOUN (Supervisor)
Award date3 May 2022


Current brain-computer interfaces (BCIs) have been typically utilized in assisting and improving life for disabled people in terms of physical and mental conditions. Even though remarkable progress has been made in recent decades, BCIs still face some critical obstacles, including massive training load, low effectiveness, low efficiency, and significant variation in performance across users. These drawbacks limit the development of BCIs, especially the applicability of EEG-based BCIs for untrained users. A possible solution is an implicit interaction, making the EEG-based BCIs adapt to the untrained users, ensuring that the relevant implicit information is retrieved, analyzed, and used appropriately. This research is devoted to studying the implicit interaction related to the applicability of EEG-based BCI. It aims to investigate how implicit interaction can enhance the applicability of EEG-based BCI in certain undefined purpose activities for untrained users.

This research reviews different literature to help understand and define the concepts in this research, such as implicit interaction, applicability, brain-computer interfaces, and untrained users (Chapter 2). It provides a methodology that can be applied to this research by discussing the user-centered approach, its adaptation to BCI, and its practice and application in this thesis (Chapter 3).

This research investigates the implicit interaction in enhancing the applicability of EEG-based BCIs for untrained users. It follows an adapted circle of user-centered approach: It starts to plan the user-centered process using literature research to provide a relevant theoretical perspective and inform the whole research process. Then, it specifies the user requirements through considering the current experience of EEG-based BCIs, and it derives the usage implications that implicit interaction systems could usefully support (Chapter 4). Next, to understand and specify the context of use, it explores the key contextual factors of EEG-based BCIs for untrained users, which could be used to enhance the applicability (Chapter 5). After that, it describes the design of an implicit interaction system using the user-centered process to improve its applicability. An information exchange framework with the user assessing model (UAM) and the system generating model (SGM) is proposed, and a multi-level approach to implement the model is produced as the design solution. The multi-level approach consists of four main activities: user-centric calibration, real-time automatic consensus, hybrid system, and multi-channel biofeedback (Chapter 6). Finally, this thesis investigates the impact of implicit interaction on applicability with a mixed lab and field experiment methodology: a lab-based study that compares different prototypes with potential users and a field-based study that examines the overall role of implicit interaction with the actual usage during an exhibition, The Brain Factory project (Chapter 7). Implicit interaction is shown to result in enhanced applicability of EEG-based BCI for untrained users.

This research proposes an implicit interaction framework to provide a foundation and instruct the implicit BCI system development to enhance the applicability in specific applications with undefined purposes. And it designs a hybrid system for untrained users with four advanced BCI approaches to enhance the BCI applications in an art project where the objective is hard to define. In conclusion, specific contributions in different aspects and potential avenues for further research are highlighted (Chapter 8).

    Research areas

  • Implicit interaction, applicability, EEG, BCI, Brain-computer interfaces