Nathalie Gayraud, PhD

Predicting a subject's ability to use a Brain Computer Interface (BCI) is one of the major issues in the BCI domain. Relevant applications of forecasting BCI performance include the ability to adapt the BCI to the needs and expectations of the user, assessing the efficiency of BCI use in stroke rehabilitation, and finally, homogenizing a research population. A limited number of recent studies have proposed the use of subjective questionnaires, such as the Motor Imagery Questionnaire… Show more

Predicting a subject's ability to use a Brain Computer Interface (BCI) is one of the major issues in the BCI domain. Relevant applications of forecasting BCI performance include the ability to adapt the BCI to the needs and expectations of the user, assessing the efficiency of BCI use in stroke rehabilitation, and finally, homogenizing a research population. A limited number of recent studies have proposed the use of subjective questionnaires, such as the Motor Imagery Questionnaire Revised-Second Edition (MIQ-RS). However, further research is necessary to confirm the effectiveness of this type of subjective questionnaire as a BCI performance estimation tool. In this study we aim to answer the following questions: can the MIQ-RS be used to estimate the performance of an MI-based BCI? If not, can we identify different markers that could be used as performance estimators? To answer these questions, we recorded EEG signals from 35 healthy volunteers during BCI use. The subjects had previously completed the MIQ-RS questionnaire. We conducted an offline analysis to assess the correlation between the questionnaire scores related to Kinesthetic and Motor imagery tasks and the performances of four classification methods. Our results showed no significant correlation between BCI performance and the MIQ-RS scores. However, we reveal that BCI performance is correlated to habits and frequency of practicing manual activities. Show less

Matching structural parcels across different subjects is an open problem in neuroscience. Even when produced by the same technique, parcellations tend to differ in the number, shape, and spatial localization of parcels across subjects. In this work, we propose a parcel matching method based on Optimal Transport. We test its performance by matching parcels of the Desikan atlas, parcels based on a functional criteria and structural parcels. We compare our technique against three other ways… Show more

Matching structural parcels across different subjects is an open problem in neuroscience. Even when produced by the same technique, parcellations tend to differ in the number, shape, and spatial localization of parcels across subjects. In this work, we propose a parcel matching method based on Optimal Transport. We test its performance by matching parcels of the Desikan atlas, parcels based on a functional criteria and structural parcels. We compare our technique against three other ways to match parcels which are based on the Euclidean distance, the cosine similarity, and the Kullback-Leibler divergence. Our results show that our method achieves the highest number of correct matches. Show less

Brain Computer Interfaces (BCIs) based on visual evoked potentials (VEP) allow for spelling from a keyboard of flashing characters. Among VEP BCIs, code-modulated visual evoked potentials (c-VEPs) are designed for high-speed communication. In c-VEPs, all characters flash simultaneously. In particular, each character flashes according to a predefined 63-bit binary sequence (m-sequence), circular-shifted by a different time lag. For a given character, the m-sequence evokes a VEP in the… Show more

Brain Computer Interfaces (BCIs) based on visual evoked potentials (VEP) allow for spelling from a keyboard of flashing characters. Among VEP BCIs, code-modulated visual evoked potentials (c-VEPs) are designed for high-speed communication. In c-VEPs, all characters flash simultaneously. In particular, each character flashes according to a predefined 63-bit binary sequence (m-sequence), circular-shifted by a different time lag. For a given character, the m-sequence evokes a VEP in the electroencephalogram (EEG) of the subject, which can be used as a template. This template is obtained during a calibration phase at the beginning of each session. Then, the system outputs the desired character after a predefined number of repetitions by estimating its time lag with respect to the template. Our work avoids the calibration phase, by extracting from the VEP relative lags between successive characters, and predicting the full word using a dictionary. Show less

Brain Computer Interfaces suffer from considerable cross-session and cross-subject variability,
which makes it hard for classification methods to generalize. We introduce a transfer learning method based on regularized discrete optimal transport with class labels in the interest of enhancing the generalization capacity of state-of-the-art classification methods. We demonstrate the potential of this approach by applying it to offline cross-subject transfer learning for the P300-Speller… Show more

Brain Computer Interfaces suffer from considerable cross-session and cross-subject variability,
which makes it hard for classification methods to generalize. We introduce a transfer learning method based on regularized discrete optimal transport with class labels in the interest of enhancing the generalization capacity of state-of-the-art classification methods. We demonstrate the potential of this approach by applying it to offline cross-subject transfer learning for the P300-Speller paradigm. We also simulate an online experiment to assess the feasibility of our method. Results show that our method is comparable to -and sometimes even outperforms- session-dependent classification. Show less

This work provides a theoretical analysis framework
for features that belong to the high dimensional Riemannian
manifold of symmetric positive definite matrices. In non-invasive
EEG-based Brain Computer Interfaces, such as the P300 speller,
these are sample covariance matrices of the epoched EEG signal
that are classified into two classes. An analysis of the class
shape on the manifold is performed, and the separability level
of the two classes is evaluated. The main… Show more

This work provides a theoretical analysis framework
for features that belong to the high dimensional Riemannian
manifold of symmetric positive definite matrices. In non-invasive
EEG-based Brain Computer Interfaces, such as the P300 speller,
these are sample covariance matrices of the epoched EEG signal
that are classified into two classes. An analysis of the class
shape on the manifold is performed, and the separability level
of the two classes is evaluated. The main contribution is the
Separability Marker (SM)-confidence method, a method that
appends a confidence marker to the prediction of a binary
classifier whose decision function is based on the comparison
of Riemannian distances. Show less

Motor imagery (MI) based BCI systems record and analyze the brain activity to determine users'
intentions while imagining moving some parts of their body. In order to build systems that are able to detect several commands, multiclass schemes need to be applied. Hierarchical methods allow solving multiclass problems by using a tree of binary classifiers, whose root discriminates between two groups, each one containing a half of the classes. Each succeeding node includes again only one half of… Show more

Motor imagery (MI) based BCI systems record and analyze the brain activity to determine users'
intentions while imagining moving some parts of their body. In order to build systems that are able to detect several commands, multiclass schemes need to be applied. Hierarchical methods allow solving multiclass problems by using a tree of binary classifiers, whose root discriminates between two groups, each one containing a half of the classes. Each succeeding node includes again only one half of the classes from the selected group, and the process is recursively repeated until each node contains a single class, from which the final decision can be inferred. In this study we compare a series of multiclass approaches to assert the benefits of hierarchical classification. The compared methods are based on two effective techniques for MI-discrimination, namely, Common Spatial Patterns (CSP) and Riemannian geometry, for which the hierarchical and non-hierarchical approaches have been considered. We include the CSP by Joint Diagonalization method (CSPbyJAD), which corresponds with a non-hierarchical approach; and its hierarchical counterpart, namely, Binary CSP. In addition, the non-hierarchical Minimum Distance to Riemannian Mean method (MDRM) is also evaluated, together with its analogous hierarchical approach; a contribution of the present work called Hierarchical MDRM algorithm (HMDRM). All these methods have been applied on dataset 2a of the BCI competition IV to facilitate their comparison. Show less

R-R interval signals that come from different subjects are regularly aligned and averaged according to the horological starting time of the recordings. We argue that the horological time is a faulty alignment criterion and provide evidence in the form of a new alignment method. Our main motivation is that the human heart rate (HR) rhythm follows a circadian cycle, whose pattern can vary among different classes of people. We propose two novel alignment algorithms that consider the HR circadian… Show more

R-R interval signals that come from different subjects are regularly aligned and averaged according to the horological starting time of the recordings. We argue that the horological time is a faulty alignment criterion and provide evidence in the form of a new alignment method. Our main motivation is that the human heart rate (HR) rhythm follows a circadian cycle, whose pattern can vary among different classes of people. We propose two novel alignment algorithms that consider the HR circadian rhythm, the Puzzle Piece Alignment Algorithm (PPA) and the Event Based Alignment Algorithm (EBA). First, we convert the R-R interval signal into a series of time windows and compute the mean HR per window. Then our algorithms search for matching circadian patterns to align the signals. We conduct experiments using R-R interval signals extracted from two databases in the Physionet Data Bank. Both algorithms are able to align the signals with respect to the circadian rhythmicity of HR. Furthermore, our findings confirm the presence of more than one pattern in the circadian HR rhythm. We suggest an automatic classification of signals according to the three most prominent patterns. Show less

Diffusion in social networks has been studied extensively in the past few years. Most previous work assumes that the underlying network is a static object that remains unchanged as the diffusion process progresses. However, there are several real-life networks that change dynamically over time. In this paper, we study diffusion on such evolving networks and extend the popular Independent Cascade and Linear Threshold models to account for network evolution. In particular, we introduce two… Show more

Diffusion in social networks has been studied extensively in the past few years. Most previous work assumes that the underlying network is a static object that remains unchanged as the diffusion process progresses. However, there are several real-life networks that change dynamically over time. In this paper, we study diffusion on such evolving networks and extend the popular Independent Cascade and Linear Threshold models to account for network evolution. In particular, we introduce two natural variations, a persistent and a transient one, to capture diffusions of different types. We consider the problem of influence maximization where the goal is to select a few influential nodes to initiate a diffusion with maximum spread. We show that, surprisingly, when considering evolving networks the diffusion function is no longer submodular for the transient models, and not even monotone for the transient Independent Cascade model. We also show that, depending on the model, delaying the activation of the initiators may improve diffusion. Our experiments, using three real datasets, demonstrate the effect of network evolution on the diffusion process, and highlight the importance of timing in the selection process. Show less