Impact of the baseline temporal selection on the ERD/ERS analysis for Motor Imagery-based BCI.

Motor Imagery-based Brain-Computer Interfaces (MI-BCIs) are neurotechnologies that exploit the modulation of sensorimotor rhythms over the motor cortices, respectively known as Event-Related Desynchronization (ERD) and Synchronization (ERS). The interpretation of ERD/ERS is directly related to the selection of the baseline used to estimate them, and might result in a misleading ERD/ERS visualization. In fact, in BCI paradigms, if two trials are separated by a few seconds, taking a baseline close to the end of the previous trial could result in an over-estimation of the ERD, while taking a baseline too close to the upcoming trial could result in an under-estimation of the ERD. This phenomenon may cause a functional misinterpretation of the ERD/ERS phenomena in MI-BCI studies. This may also impair BCI performances for MI vs Rest classification, since such baselines are often used as resting states. In this paper, we propose to investigate the effect of several baseline time window selections on ERD/ERS modulations and BCI performances. Our results show that considering the selected temporal baseline effect is essential to analyze the modulations of ERD/ERS during MI-BCI use.

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

[This corrects the article DOI: 10.1371/journal.pone.0143962.].

Class-distinctiveness-based frequency band selection on the Riemannian manifold for oscillatory activity-based BCIs: preliminary results.

Considering user-specific settings is known to enhance Brain-Computer Interface (BCI) performances. In particular, the optimal frequency band for oscillatory activity classification is highly user-dependent and many frequency band selection methods have been developed in the past two decades. However, it is not well studied whether those conventional methods can be efficiently applied to the Riemannian BCIs, a recent family of BCI systems that utilize the non-Euclidean nature of the data unlike conventional BCI pipelines. In this paper, we proposed a novel frequency band selection method working on the Riemannian manifold. The frequency band is selected considering the class distinctiveness as quantified based on the inter-class distance and the intra-class variance on the manifold. An advantage of this method is that the frequency bandwidth can be adjusted for each individual without intensive optimization steps. In a comparative experiment using a public dataset of motor imagery-based BCI, our method showed a substantial improvement in average accuracy over both a fixed broad frequency band and a popular conventional frequency band selection method. In particular, our method substantially improved performances for subjects with initially low accuracies. This preliminary result suggests the importance of developing new user-specific setting algorithms considering the manifold properties, rather than directly applying methods developed prior to the rise of the Riemannian BCIs.

ERD modulations during motor imageries relate to users' traits and BCI performances.

Improving user performances is one of the major issues for Motor Imagery (MI) - based BCI control. MI-BCIs exploit the modulation of sensorimotor rhythms (SMR) over the motor and sensorimotor cortices to discriminate several mental states and enable user interaction. Such modulations are known as Event-Related Desynchronization (ERD) and Synchronization (ERS), coming from the mu (7-13 Hz) and beta (15-30 Hz) frequency bands. This kind of BCI opens up promising fields, particularly to control assistive technologies, for sport training or even for post-stroke motor rehabilitation. However, MI - BCIs remain barely used outside laboratories, notably due to their lack of robustness and usability (15 to 30% of users seem unable to gain control of an MI-BCI). One way to increase user performance would be to better understand the relationships between user traits and ERD/ERS modulations underlying BCI performance. Therefore, in this article we analyzed how cerebral motor patterns underlying MI tasks (i.e., ERDs and ERSs) are modulated depending (i) on nature of the task (i.e., right-hand MI and left-hand MI), (ii) the session during which the task was performed (i.e., calibration or user training) and (iii) on the characteristics of the user (e.g., age, gender, manual activity, personality traits) on a large MI-BCI data base of N=75 participants. One of the originality of this study is to combine the investigation of human factors related to the user's traits and the neurophysiological ERD modulations during the MI task. Our study revealed for the first time an association between ERD and self-control from the 16PF5 questionnaire.

Long-Term BCI Training of a Tetraplegic User: Adaptive Riemannian Classifiers and User Training.

While often presented as promising assistive technologies for motor-impaired users, electroencephalography (EEG)-based Brain-Computer Interfaces (BCIs) remain barely used outside laboratories due to low reliability in real-life conditions. There is thus a need to design long-term reliable BCIs that can be used outside-of-the-lab by end-users, e.g., severely motor-impaired ones. Therefore, we propose and evaluate the design of a multi-class Mental Task (MT)-based BCI for longitudinal training (20 sessions over 3 months) of a tetraplegic user for the CYBATHLON BCI series 2019. In this BCI championship, tetraplegic pilots are mentally driving a virtual car in a racing video game. We aimed at combining a progressive user MT-BCI training with a newly designed machine learning pipeline based on adaptive Riemannian classifiers shown to be promising for real-life applications. We followed a two step training process: the first 11 sessions served to train the user to control a 2-class MT-BCI by performing either two cognitive tasks (REST and MENTAL SUBTRACTION) or two motor-imagery tasks (LEFT-HAND and RIGHT-HAND). The second training step (9 remaining sessions) applied an adaptive, session-independent Riemannian classifier that combined all 4 MT classes used before. Moreover, as our Riemannian classifier was incrementally updated in an unsupervised way it would capture both within and between-session non-stationarity. Experimental evidences confirm the effectiveness of this approach. Namely, the classification accuracy improved by about 30% at the end of the training compared to initial sessions. We also studied the neural correlates of this performance improvement. Using a newly proposed BCI user learning metric, we could show our user learned to improve his BCI control by producing EEG signals matching increasingly more the BCI classifier training data distribution, rather than by improving his EEG class discrimination. However, the resulting improvement was effective only on synchronous (cue-based) BCI and it did not translate into improved CYBATHLON BCI game performances. For the sake of overcoming this in the future, we unveil possible reasons for these limited gaming performances and identify a number of promising future research directions. Importantly, we also report on the evolution of the user's neurophysiological patterns and user experience throughout the BCI training and competition.

Why we should systematically assess, control and report somatosensory impairments in BCI-based motor rehabilitation after stroke studies.

The neuronal loss resulting from stroke forces 80% of the patients to undergo motor rehabilitation, for which Brain-Computer Interfaces (BCIs) and NeuroFeedback (NF) can be used. During the rehabilitation, when patients attempt or imagine performing a movement, BCIs/NF provide them with a synchronized sensory (e.g., tactile) feedback based on their sensorimotor-related brain activity that aims at fostering brain plasticity and motor recovery. The co-activation of ascending (i.e., somatosensory) and descending (i.e., motor) networks indeed enables significant functional motor improvement, together with significant sensorimotor-related neurophysiological changes. Somatosensory abilities are essential for patients to perceive the feedback provided by the BCI system. Thus, somatosensory impairments may significantly alter the efficiency of BCI-based motor rehabilitation. In order to precisely understand and assess the impact of somatosensory impairments, we first review the literature on post-stroke BCI-based motor rehabilitation (14 randomized clinical trials). We show that despite the central role that somatosensory abilities play on BCI-based motor rehabilitation post-stroke, the latter are rarely reported and used as inclusion/exclusion criteria in the literature on the matter. We then argue that somatosensory abilities have repeatedly been shown to influence the motor rehabilitation outcome, in general. This stresses the importance of also considering them and reporting them in the literature in BCI-based rehabilitation after stroke, especially since half of post-stroke patients suffer from somatosensory impairments. We argue that somatosensory abilities should systematically be assessed, controlled and reported if we want to precisely assess the influence they have on BCI efficiency. Not doing so could result in the misinterpretation of reported results, while doing so could improve (1) our understanding of the mechanisms underlying motor recovery (2) our ability to adapt the therapy to the patients' impairments and (3) our comprehension of the between-subject and between-study variability of therapeutic outcomes mentioned in the literature.

Consensus on the reporting and experimental design of clinical and cognitive-behavioural neurofeedback studies (CRED-nf checklist).

Neurofeedback has begun to attract the attention and scrutiny of the scientific and medical mainstream. Here, neurofeedback researchers present a consensus-derived checklist that aims to improve the reporting and experimental design standards in the field.

Defining and quantifying users' mental imagery-based BCI skills: a first step.

OBJECTIVE: While promising for many applications, electroencephalography (EEG)-based brain-computer interfaces (BCIs) are still scarcely used outside laboratories, due to a poor reliability. It is thus necessary to study and fix this reliability issue. Doing so requires the use of appropriate reliability metrics to quantify both the classification algorithm and the BCI user's performances. So far, classification accuracy (CA) is the typical metric used for both aspects. However, we argue in this paper that CA is a poor metric to study BCI users' skills. Here, we propose a definition and new metrics to quantify such BCI skills for mental imagery (MI) BCIs, independently of any classification algorithm. APPROACH: We first show in this paper that CA is notably unspecific, discrete, training data and classifier dependent, and as such may not always reflect successful self-modulation of EEG patterns by the user. We then propose a definition of MI-BCI skills that reflects how well the user can self-modulate EEG patterns, and thus how well he could control an MI-BCI. Finally, we propose new performance metrics, classDis, restDist and classStab that specifically measure how distinct and stable the EEG patterns produced by the user are, independently of any classifier. MAIN RESULTS: By re-analyzing EEG data sets with such new metrics, we indeed confirmed that CA may hide some increase in MI-BCI skills or hide the user inability to self-modulate a given EEG pattern. On the other hand, our new metrics could reveal such skill improvements as well as identify when a mental task performed by a user was no different than rest EEG. SIGNIFICANCE: Our results showed that when studying MI-BCI users' skills, CA should be used with care, and complemented with metrics such as the new ones proposed. Our results also stressed the need to redefine BCI user training by considering the different BCI subskills and their measures. To promote the complementary use of our new metrics, we provide the Matlab code to compute them for free and open-source.

Using Recent BCI Literature to Deepen our Understanding of Clinical Neurofeedback: A Short Review.

In their recent paper, Alkoby et al. (2017) provide the readership with an extensive and very insightful review of the factors influencing NeuroFeedback (NF) performance. These factors are drawn from both the NF literature and the Brain-Computer Interface (BCI) literature. Our short review aims to complement Alkoby et al.'s review by reporting recent additions to the BCI literature. The object of this paper is to highlight this literature and discuss its potential relevance and usefulness to better understand the processes underlying NF and further improve the design of clinical trials assessing NF efficacy. Indeed, we are convinced that while NF and BCI are fundamentally different in many ways, both the BCI and NF communities could reach compelling achievements by building upon one another. By reviewing the recent BCI literature, we identified three types of factors that influence BCI performance: task-specific, cognitive/motivational and technology-acceptance-related factors. Since BCIs and NF share a common goal (i.e., learning to modulate specific neurophysiological patterns), similar cognitive and neurophysiological processes are likely to be involved during the training process. Thus, the literature on BCI training may help (1) to deepen our understanding of neurofeedback training processes and (2) to understand the variables that influence the clinical efficacy of NF. This may help to properly assess and/or control the influence of these variables during randomized controlled trials.

Why standard brain-computer interface (BCI) training protocols should be changed: an experimental study.

OBJECTIVE: While promising, electroencephaloraphy based brain-computer interfaces (BCIs) are barely used due to their lack of reliability: 15% to 30% of users are unable to control a BCI. Standard training protocols may be partly responsible as they do not satisfy recommendations from psychology. Our main objective was to determine in practice to what extent standard training protocols impact users' motor imagery based BCI (MI-BCI) control performance. APPROACH: We performed two experiments. The first consisted in evaluating the efficiency of a standard BCI training protocol for the acquisition of non-BCI related skills in a BCI-free context, which enabled us to rule out the possible impact of BCIs on the training outcome. Thus, participants (N = 54) were asked to perform simple motor tasks. The second experiment was aimed at measuring the correlations between motor tasks and MI-BCI performance. The ten best and ten worst performers of the first study were recruited for an MI-BCI experiment during which they had to learn to perform two MI tasks. We also assessed users' spatial ability and pre-training µ rhythm amplitude, as both have been related to MI-BCI performance in the literature. MAIN RESULTS: Around 17% of the participants were unable to learn to perform the motor tasks, which is close to the BCI illiteracy rate. This suggests that standard training protocols are suboptimal for skill teaching. No correlation was found between motor tasks and MI-BCI performance. However, spatial ability played an important role in MI-BCI performance. In addition, once the spatial ability covariable had been controlled for, using an ANCOVA, it appeared that participants who faced difficulty during the first experiment improved during the second while the others did not. SIGNIFICANCE: These studies suggest that (1) standard MI-BCI training protocols are suboptimal for skill teaching, (2) spatial ability is confirmed as impacting on MI-BCI performance, and (3) when faced with difficult pre-training, subjects seemed to explore more strategies and therefore learn better.

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.

Flaws in current human training protocols for spontaneous Brain-Computer Interfaces: lessons learned from instructional design.

While recent research on Brain-Computer Interfaces (BCI) has highlighted their potential for many applications, they remain barely used outside laboratories. The main reason is their lack of robustness. Indeed, with current BCI, mental state recognition is usually slow and often incorrect. Spontaneous BCI (i.e., mental imagery-based BCI) often rely on mutual learning efforts by the user and the machine, with BCI users learning to produce stable ElectroEncephaloGraphy (EEG) patterns (spontaneous BCI control being widely acknowledged as a skill) while the computer learns to automatically recognize these EEG patterns, using signal processing. Most research so far was focused on signal processing, mostly neglecting the human in the loop. However, how well the user masters the BCI skill is also a key element explaining BCI robustness. Indeed, if the user is not able to produce stable and distinct EEG patterns, then no signal processing algorithm would be able to recognize them. Unfortunately, despite the importance of BCI training protocols, they have been scarcely studied so far, and used mostly unchanged for years. In this paper, we advocate that current human training approaches for spontaneous BCI are most likely inappropriate. We notably study instructional design literature in order to identify the key requirements and guidelines for a successful training procedure that promotes a good and efficient skill learning. This literature study highlights that current spontaneous BCI user training procedures satisfy very few of these requirements and hence are likely to be suboptimal. We therefore identify the flaws in BCI training protocols according to instructional design principles, at several levels: in the instructions provided to the user, in the tasks he/she has to perform, and in the feedback provided. For each level, we propose new research directions that are theoretically expected to address some of these flaws and to help users learn the BCI skill more efficiently.