Camille Jeunet

EEG-based workload estimation across affective contexts.

Workload estimation from electroencephalographic signals (EEG) offers a highly sensitive tool to adapt the human–computer interaction to the user state. To create systems that reliably work in the complexity of the real world, a robustness against contextual changes (e.g., mood), has to be achieved. To study the resilience of state-of-the-art EEG-based workload classification against stress we devise a novel experimental protocol, in which we manipulated the affective context (stressful/non-stressful) while the participant solved a task with two workload levels. We recorded self-ratings, behavior, and physiology from 24 participants to validate the protocol. We test the capability of different, subject-specific workload classifiers using either frequency-domain, time-domain, or both feature varieties to generalize across contexts. We show that the classifiers are able to transfer between affective contexts, though performance suffers independent of the used feature domain. However, cross-context training is a simple and powerful remedy allowing the extraction of features in all studied feature varieties that are more resilient to task-unrelated variations in signal characteristics. Especially for frequency-domain features, across-context training is leading to a performance comparable to within-context training and testing. We discuss the significance of the result for neurophysiology-based workload detection in particular and for the construction of reliable passive brain–computer interfaces in general.

Predicting Mental Imagery-Based BCI Performance from Personality, Cognitive Profile and Neurophysiological Patterns

Mental-Imagery based Brain-Computer Interfaces (MI-BCIs) allow their users to send commands to a computer using their brain-activity alone (typically measured by ElectroEncephaloGraphy—EEG), which is processed while they perform specific mental tasks. While very promising, MI-BCIs remain barely used outside laboratories because of the difficulty encountered by users to control them. Indeed, although some users obtain good control performances after training, a substantial proportion remains unable to reliably control an MI-BCI. This huge variability in user-performance led the community to look for predictors of MI-BCI control ability. However, these predictors were only explored for motor-imagery based BCIs, and mostly for a single training session per subject. In this study, 18 participants were instructed to learn to control an EEG-based MI-BCI by performing 3 MI-tasks, 2 of which were non-motor tasks, across 6 training sessions, on 6 different days. Relationships between the participants’ BCI control performances and their personality, cognitive profile and neurophysiological markers were explored. While no relevant relationships with neurophysiological markers were found, strong correlations between MI-BCI performances and mental-rotation scores (reflecting spatial abilities) were revealed. Also, a predictive model of MI-BCI performance based on psychometric questionnaire scores was proposed. A leave-one-subject-out cross validation process revealed the stability and reliability of this model: it enabled to predict participants’ performance with a mean error of less than 3 points. This study determined how users’ profiles impact their MI-BCI control ability and thus clears the way for designing novel MI-BCI training protocols, adapted to the profile of each user.

Towards improved BCI based on human learning principles

Although EEG-based BCI are very promising for numerous applications, they mostly remain prototypes not used outside laboratories, due to their low reliability. Poor BCI performances are partly due to imperfect EEG signal processing algorithms but also to the user, who may not be able to produce reliable EEG patterns. This paper presents some of our current work that aims at addressing the latter, i.e., at guiding users to learn BCI control mastery. First, this paper identifies some theoretical (based on human learning psychology models) and practical limitations of current standard BCI training approaches and thus the need for alternative ones. To try to address these limitations, we conducted a study to explore what kind of users can use a BCI and why, and will present the main results. We also present new feedback types we designed to help users to learn BCI control skills more efficiently.

Continuous Tactile Feedback for Motor-Imagery Based Brain-Computer Interaction in a Multitasking Context

Motor-Imagery based Brain Computer Interfaces (MI-BCIs) allow users to interact with computers by imagining limb movements. MI-BCIs are very promising for a wide range of applications as they offer a new and non-time locked modality of control. However, most MI-BCIs involve visual feedback to inform the user about the system’s decisions, which makes them difficult to use when integrated with visual interactive tasks. This paper presents our design and evaluation of a tactile feedback glove for MI-BCIs, which provides a continuously updated tactile feedback. We first determined the best parameters for this tactile feedback and then tested it in a multitasking environment: at the same time users were performing the MI tasks, they were asked to count distracters. Our results suggest that, as compared to an equivalent visual feedback, the use of tactile feedback leads to a higher recognition accuracy of the MI-BCI tasks and fewer errors in counting distracters.

Towards Explanatory Feedback for User Training in Brain-Computer Interfaces

Despite their potential for many applications, Brain -- Computer Interfaces (BCI) are still rarely used due to their low reliability and long training. These limitations are partly due to inappropriate training protocols, which includes the feedback provided to the user. While feedback should theoretically be explanatory, motivating and meaningful, current BCI feedback is usually boring, corrective only and difficult to understand. In this study, different features of the electroencephalogram signals were explored to be used as a richer, explanatory BCI feedback. First, based on offline mental imagery BCI data, muscular relaxation was notably found to be negatively correlated to BCI performance. Second, this study reports on an online BCI evaluation using muscular relaxation as additional feedback. While this additional feedback did not lead to significant change in BCI performance, this study showed that multiple feedbacks can be used without deteriorating performance and provided interesting insights for explanatory BCI feedback design.