REdefining Brain-Computer Interfaces to Enable their users to achieve controL mastery – REBEL

To redefine BCI user training, we will first work on understanding and modeling BCI skill acquisition from a neurophysiological point of view. In other words, we first aim at identifying what are the EEG features defining good EEG patterns (that are successfully recognized by the BCI), and how they evolve with training. Then, we will propose new EEG signal processing tools to quantify such training-related EEG features in real-time. This will enable us to identify objectives to reach with BCI training and a way to quantify and guide the user’s progress during training. Afterwards, we will combine these new EEG features and BCI training models with recommendations and principles from human learning and education psychology to propose new and relevant feedback and training tasks to radically improve BCI training. In particular, we will propose adaptive and adapted training tasks and provide the users with an explanatory feedback (indicating what is good or bad about the EEG patterns performed) based on our new training-related EEG features. Finally we will extensively evaluate and validate our new BCI, first on healthy users, then on a few motor-impaired ones. Overall, we target a new BCI design leading to a fast acquisition of reliable BCI control skills.

So far, concerning the improvement of BCI user training, we performed a number of surveys to identify the challenges, research directions and tools that are the most promising to improve it. We also modeled the factors influencing control performance for BCI based on mental imagery. Our model notably identifies three main categories of factors: spatial abilities, attentional abilities and technology acceptance. It identifies the relationships between these factors, and how they could be manipulated. We also proposed a framework to design adaptive BCI that take into account the user profile, in order to improve user training, thanks to the model factors mentioned above. Finally, we defined new metrics of performances to study and quantify BCI control skills, and how they evolve.

Regarding improving BCI user training, we explored training tasks and feedback that take into account some factors in our BCI control model. We notably designed, tested and validated a spatial ability training approach. Our model also revealed that anxious and less autonomous users tend to have lower BCI performances. To address this, we developed a learning companion that provides social and emotional feedback, under the form of oral support and suggestion together with emotional facial expressions. Integrating this learning companion to a standard BCI training led to improved user experience. Finally, our model revealed that the less the user feels in control, the lower their performance. We thus proposed an adaptive feedback, which modulates the perceive difficulty in order to improve the feeling of control.

The next steps consist in continuing the analyses we started about spatial abilities training, and in assessing the impact of biased and tactile feedback on BCI performances (the experiments are completed, the analysis in progress). We will also explore another factor from our model: attention. We will notably look for markers of attention in EEG signals, in order to provide the users with attention feedback or attentional training. We are also about to start experiments with our training approaches with Stroke patients with Bordeaux hospital. Finally, we also plan to explore new content for motor imagery BCI, based on more accurate and relevant localization of the motor cortex, in order to improve training.

Journals:

C. Jeunet, F. Lotte, J.-M. Batail, P. Philip, J.-A. Micoulaud-Franchi, “How to improve clinical neurofeedback using a human-factor centered standpoint? A short review of the insights provided by the literature on BCI”, Neuroscience, 2017

F. Yger, M. Bérar, F. Lotte, « Riemannian approaches in Brain-Computer Interfaces: a review », IEEE Transactions on Neural System and Rehabilitation Engineering, 2017

M. Arns et al, « Neurofeedback: one of today’s techniques in psychiatry? », L’Encéphale, 2017

Chavarriaga, M. Fried-Oken, S. Kleih, F. Lotte, R. Scherer, « Heading for new shores! Overcoming pitfalls in BCI design », Brain-Computer Interfaces, pp. 1-14, 2016

Conferences:

C. Jeunet, B. N’Kaoua & F. Lotte, “Towards a cognitive model of MI-BCI user training”, Int. BCI Conf., 2017

F. Lotte, C. Jeunet, “Online classification accuracy is a poor metric to study mental-imagery based BCI user learning: An experimental demonstration and new metrics”, Int. BCI Conf., 2017

L. Pillette, C. Jeunet, B. Mansencal, R. N’Kambou, B. N’Kaoua, F. Lotte, “PEANUT : Personalised Emotional Agent for Neurotechnology User-Training”, Int. BCI Conf., 2017

J. Mladenovic, J. Frey, M. Bonnet-Save, J. Mattout, F. Lotte, « The Impact of Flow in an EEG-based Brain Computer Interface », Int. BCI Conf., 2017

S. Teillet, F. Lotte, B. N’Kaoua, C. Jeunet, « Towards a Spatial Ability Training to Improve Motor Imagery based Brain-Computer Interfaces (MI-BCIs) Performance: a Pilot Study« , IEEE SMC, 2016

Book chapters:

Lotte, CS Nam, A Nijholt, “Evolution of Brain-Computer Interfaces”, BCI Handbook, Taylor & Francis, 2018

J Mladenovic, J Mattout, and F Lotte, “A Generic Framework for Adaptive EEG-Based BCI Training and Operation”, BCI Handbook, Taylor & Francis, 2018

C Jeunet, S Debener, F Lotte, J Mattout, R Scherer, and C Zich, “Mind the Traps: Design Guidelines for Rigorous BCI Experiments”, BCI Handbook, Taylor & Francis, 2018

Brain-Computer Interfaces (BCI) are communication systems that enable their users to send commands to computers through brain activity only, this activity being measured and processed by the BCI (usually using ElectroEncephaloGraphy – EEG). Making computer control possible without any physical activity, BCI have promised to revolutionize many application areas, notably assistive technologies for paralyzed users (e.g., for wheelchair control) and human-computer interaction (HCI). Despite this promising potential, BCI are still barely used outside laboratories, due to a poor reliability. For instance, current BCI based on only 2 imagined movements correctly recognize less than 80% of the users’ mental commands, on average, while between 10 to 30% of BCI users (depending on the BCI type) cannot control a BCI at all.
Designing a reliable BCI requires to consider it as a co-adaptive system, with its users learning to produce distinct brain activity patterns that the machine learns to recognize using signal processing. Indeed, BCI control is a skill that the user has to learn. Most research efforts so far have been dedicated to signal processing or human-computer interaction techniques, i.e., on the computer side. Unfortunately BCI user training is as essential but 1) is only scarcely studied and 2) standard approaches are only based on heuristics, without satisfying human learning principles. Thus, currently poor BCI reliability is probably due to a large extent to highly suboptimal user training.
Therefore, to obtain a much higher reliability for BCI we need a major rethinking of their fundamentals in algorithmics (signal processing, machine learning) and user training (feedback and training tasks). In particular, we propose to create a new generation of BCI that apply human learning principles to ensure the users can learn high quality control skills which will go much beyond those obtained with currently available systems, hence making Brain-Computer Interfaces reliable and trustable.
To do so, we will first work on understanding and modeling BCI skill acquisition from a neurophysiological point of view. In other words, we first aim at identifying what are the EEG features defining good EEG patterns (that are successfully recognized by the BCI), and how they evolve with training. Then, we will propose new EEG signal processing tools to quantify such training-related EEG features in real-time. This will enable us to identify objectives to reach with BCI training and a way to quantify and guide the user’s progress during training. Afterwards, we will combine these new EEG features and BCI training models with recommendations and principles from human learning and education psychology to propose new and relevant feedback and training tasks to radically improve BCI training. In particular, we will propose adaptive and adapted training tasks and provide the users with an explanatory feedback (indicating what is good or bad about the EEG patterns performed) based on our new training-related EEG features. Finally we will extensively evaluate and validate our new BCI, first on healthy users, then on a few motor-impaired ones. Overall, we target a new BCI design leading to a fast acquisition of reliable BCI control skills. Such a reliable BCI could actually positively change HCI as BCI have promised but failed to do so far.