Investigating the effects of a sensorimotor rhythm-based BCI training on the cortical activity elicited by mental imagery.

OBJECTIVE: It is well known that to acquire sensorimotor (SMR)-based brain-computer interface (BCI) control requires a training period before users can achieve their best possible performances. Nevertheless, the effect of this training procedure on the cortical activity related to the mental imagery ability still requires investigation to be fully elucidated. The aim of this study was to gain insights into the effects of SMR-based BCI training on the cortical spectral activity associated with the performance of different mental imagery tasks. APPROACH: Linear cortical estimation and statistical brain mapping techniques were applied on high-density EEG data acquired from 18 healthy participants performing three different mental imagery tasks. Subjects were divided in two groups, one of BCI trained subjects, according to their previous exposure (at least six months before this study) to motor imagery-based BCI training, and one of subjects who were naive to any BCI paradigms. MAIN RESULTS: Cortical activation maps obtained for trained and naive subjects indicated different spectral and spatial activity patterns in response to the mental imagery tasks. Long-term effects of the previous SMR-based BCI training were observed on the motor cortical spectral activity specific to the BCI trained motor imagery task (simple hand movements) and partially generalized to more complex motor imagery task (playing tennis). Differently, mental imagery with spatial attention and memory content could elicit recognizable cortical spectral activity even in subjects completely naive to (BCI) training. SIGNIFICANCE: The present findings contribute to our understanding of BCI technology usage and might be of relevance in those clinical conditions when training to master a BCI application is challenging or even not possible.

Influence of P300 latency jitter on event related potential-based brain-computer interface performance.

OBJECTIVE: Several ERP-based brain-computer interfaces (BCIs) that can be controlled even without eye movements (covert attention) have been recently proposed. However, when compared to similar systems based on overt attention, they displayed significantly lower accuracy. In the current interpretation, this is ascribed to the absence of the contribution of short-latency visual evoked potentials (VEPs) in the tasks performed in the covert attention modality. This study aims to investigate if this decrement (i) is fully explained by the lack of VEP contribution to the classification accuracy; (ii) correlates with lower temporal stability of the single-trial P300 potentials elicited in the covert attention modality. APPROACH: We evaluated the latency jitter of P300 evoked potentials in three BCI interfaces exploiting either overt or covert attention modalities in 20 healthy subjects. The effect of attention modality on the P300 jitter, and the relative contribution of VEPs and P300 jitter to the classification accuracy have been analyzed. MAIN RESULTS: The P300 jitter is higher when the BCI is controlled in covert attention. Classification accuracy negatively correlates with jitter. Even disregarding short-latency VEPs, overt-attention BCI yields better accuracy than covert. When the latency jitter is compensated offline, the difference between accuracies is not significant. SIGNIFICANCE: The lower temporal stability of the P300 evoked potential generated during the tasks performed in covert attention modality should be regarded as the main contributing explanation of lower accuracy of covert-attention ERP-based BCIs.

Self-calibration algorithm in an asynchronous P300-based brain-computer interface.

OBJECTIVE: Reliability is a desirable characteristic of brain-computer interface (BCI) systems when they are intended to be used under non-experimental operating conditions. In addition, their overall usability is influenced by the complex and frequent procedures that are required for configuration and calibration. Earlier studies examined the issue of asynchronous control in P300-based BCIs, introducing dynamic stopping and automatic control suspension features. This report proposes and evaluates an algorithm for the automatic recalibration of the classifier's parameters using unsupervised data. APPROACH: Ten healthy subjects participated in five P300-based BCI sessions throughout a single day. First, we examined whether continuous adaptation of control parameters improved the accuracy of the asynchronous system over time. Then, we assessed the performance of the self-calibration algorithm with respect to the no-recalibration and supervised calibration conditions with regard to system accuracy and communication efficiency. MAIN RESULTS: Offline tests demonstrated that continuous adaptation of the control parameters significantly increased the communication efficiency of asynchronous P300-based BCIs. The self-calibration algorithm correctly assigned labels to unsupervised data with 95% accuracy, effecting communication efficiency that was comparable with that of supervised repeated calibration. SIGNIFICANCE: Although additional online tests that involve end-users under non-experimental conditions are needed, these preliminary results are encouraging, from which we conclude that the self-calibration algorithm is a promising solution to improve P300-based BCI usability and reliability.

A new descriptor of neuroelectrical activity during BCI-assisted motor imagery-based training in stroke patients.

In BCI applications for stroke rehabilitation, BCI systems are used with the aim of providing patients with an instrument that is capable of monitoring and reinforcing EEG patterns generated by motor imagery (MI). In this study we proposed an offline analysis on data acquired from stroke patients subjected to a BCI-assisted MI training in order to define an index for the evaluation of MI-BCI training session which is independent from the settings adopted for the online control and which is able to describe the properties of neuroelectrical activations across sessions. Results suggest that such index can be adopted to sort the trails within a session according to the adherence to the task.

Individual cortical connectivity changes after stroke: a resampling approach to enable statistical assessment at single-subject level.

One of the main limitations commonly encountered when dealing with the estimation of brain connectivity is the difficulty to perform a statistical assessment of significant changes in brain networks at a single-subject level. This is mainly due to the lack of information about the distribution of the connectivity estimators at different conditions. While group analysis is commonly adopted to perform a statistical comparison between conditions, it may impose major limitations when dealing with the heterogeneity expressed by a given clinical condition in patients. This holds true particularly for stroke when seeking for quantitative measurements of the efficacy of any rehabilitative intervention promoting recovery of function. The need is then evident of an assessment which may account for individual pathological network configuration associated with different level of patients' response to treatment; such network configuration is highly related to the effect that a given brain lesion has on neural networks. In this study we propose a resampling-based approach to the assessment of statistically significant changes in cortical connectivity networks at a single subject level. First, we provide the results of a simulation study testing the performances of the proposed approach under different conditions. Then, to show the sensitivity of the method, we describe its application to electroencephalographic (EEG) data recorded from two post-stroke patients who showed different clinical recovery after a rehabilitative intervention.

Investigating statistical differences in connectivity patterns properties at single subject level: a new resampling approach.

Methods based on the multivariate autoregressive (MVAR) approach are commonly used for effective connectivity estimation as they allow to include all available sources into a unique model. To ensure high levels of accuracy for high model dimensions, all the observations are used to provide a unique estimation of the model, and thus of the network and its properties. The unavailability of a distribution of connectivity values for a single experimental condition prevents to perform statistical comparisons between different conditions at a single subject level. This is a major limitation, especially when dealing with the heterogeneity of clinical conditions presented by patients. In the present paper we proposed a novel approach to the construction of a distribution of connectivity in a single subject case. The proposed approach is based on small perturbations of the networks properties and allows to assess significant changes in brain connectivity indexes derived from graph theory. Its feasibility and applicability were investigated by means of a simulation study and an application to real EEG data.

The effect of normalization of Partial Directed Coherence on the statistical assessment of connectivity patterns: a simulation study.

Partial Directed Coherence (PDC) is a spectral multivariate estimator for effective connectivity, relying on the concept of Granger causality. Even if its original definition derived directly from information theory, two modifies were introduced in order to provide better physiological interpretations of the estimated networks: i) normalization of the estimator according to rows, ii) squared transformation. In the present paper we investigated the effect of PDC normalization on the performances achieved by applying the statistical validation process on investigated connectivity patterns under different conditions of Signal to Noise ratio (SNR) and amount of data available for the analysis. Results of the statistical analysis revealed an effect of PDC normalization only on the percentages of type I and type II errors occurred by using Shuffling procedure for the assessment of connectivity patterns. No effects of the PDC formulation resulted on the performances achieved during the validation process executed instead by means of Asymptotic Statistic approach. Moreover, the percentages of both false positives and false negatives committed by Asymptotic Statistic are always lower than those achieved by Shuffling procedure for each type of normalization.

Aged-related changes in brain activity classification with respect to age by means of graph indexes.

Recent studies have investigated changes in the human brain network organization during the normal aging. A reduction of the connectivity between brain areas was demonstrated by combining neuroimaging technologies and graph theory. Clustering, characteristic path length and small-worldness are key topological measures and they are widely used in literature. In this paper we propose a new methodology that combine advanced techniques of effective connectivity estimation, graph theoretical approach and classification by SVM method. EEG signals recording during rest condition from 20 young subjects and 20 mid-aged adults were studied. Partial Directed Coherence was computed by means of General Linear Kalman Filter and graph indexes were extracted from estimated patterns. At last small-worldness was used as feature for the SVM classifier. Results show that topological differences of brain networks exist between young and mid-aged adults: small-worldness is significantly different between the two populations and it can be used to classify the subjects with respect to age with an accuracy of 69%.

A new statistical approach for the extraction of adjacency matrix from effective connectivity networks.

Graph theory is a powerful mathematical tool recently introduced in neuroscience field for quantitatively describing the main properties of investigated connectivity networks. Despite the technical advancements provided in the last few years, further investigations are needed for overcoming actual limitations in the field. In fact, the absence of a common procedure currently applied for the extraction of the adjacency matrix from a connectivity pattern has been leading to low consistency and reliability of ghaph indexes among the investigated population. In this paper we proposed a new approach for adjacency matrix extraction based on a statistical threshold as valid alternative to empirical approaches, extensively used in Neuroscience field (i.e. fixing the edge density). In particular we performed a simulation study for investigating the effects of the two different extraction approaches on the topological properties of the investigated networks. In particular, the comparison was performed on two different datasets, one composed by uncorrelated random signals (null-model) and the other one by signals acquired on a mannequin head used as a phantom (EEG null-model). The results highlighted the importance to use a statistical threshold for the adjacency matrix extraction in order to describe the real existing topological properties of the investigated networks. The use of an empirical threshold led to an erroneous definition of small-world properties for the considered connectivity patterns.

Advanced methods for time-varying effective connectivity estimation in memory processes.

Memory processes are based on large cortical networks characterized by non-stationary properties and time scales which represent a limitation to the traditional connectivity estimation methods. The recent development of connectivity approaches able to consistently describe the temporal evolution of large dimension connectivity networks, in a fully multivariate way, represents a tool that can be used to extract novel information about the processes at the basis of memory functions. In this paper, we applied such advanced approach in combination with the use of state-of-the-art graph theory indexes, computed on the connectivity networks estimated from high density electroencephalographic (EEG) data recorded in a group of healthy adults during the Sternberg Task. The results show how this approach is able to return a characterization of the main phases of the investigated memory task which is also sensitive to the increased length of the numerical string to be memorized.

How the statistical validation of functional connectivity patterns can prevent erroneous definition of small-world properties of a brain connectivity network.

The application of Graph Theory to the brain connectivity patterns obtained from the analysis of neuroelectrical signals has provided an important step to the interpretation and statistical analysis of such functional networks. The properties of a network are derived from the adjacency matrix describing a connectivity pattern obtained by one of the available functional connectivity methods. However, no common procedure is currently applied for extracting the adjacency matrix from a connectivity pattern. To understand how the topographical properties of a network inferred by means of graph indices can be affected by this procedure, we compared one of the methods extensively used in Neuroscience applications (i.e. fixing the edge density) with an approach based on the statistical validation of achieved connectivity patterns. The comparison was performed on the basis of simulated data and of signals acquired on a polystyrene head used as a phantom. The results showed (i) the importance of the assessing process in discarding the occurrence of spurious links and in the definition of the real topographical properties of the network, and (ii) a dependence of the small world properties obtained for the phantom networks from the spatial correlation of the neighboring electrodes.

A comparison of classification techniques for a gaze-independent P300-based brain-computer interface.

This off-line study aims to assess the performance of five classifiers commonly used in the brain-computer interface (BCI) community, when applied to a gaze-independent P300-based BCI. In particular, we compared the results of four linear classifiers and one nonlinear: Fisher's linear discriminant analysis (LDA), stepwise linear discriminant analysis (SWLDA), Bayesian linear discriminant analysis (BLDA), linear support vector machine (LSVM) and Gaussian supported vector machine (GSVM). Moreover, different values for the decimation of the training dataset were tested. The results were evaluated both in terms of accuracy and written symbol rate with the data of 19 healthy subjects. No significant differences among the considered classifiers were found. The optimal decimation factor spanned a range from 3 to 24 (12 to 94 ms long bins). Nevertheless, performance on individually optimized classification parameters is not significantly different from a classification with general parameters (i.e. using an LDA classifier, about 48 ms long bins).

Describing relevant indices from the resting state electrophysiological networks.

The "Default Mode Network" concept was defined, in fMRI field, as a consistent pattern, involving some regions of the brain, which is active during resting state activity and deactivates during attention demanding or goal-directed tasks. Several fMRI studies described its features also correlating the deactivations with the attentive load required for the task execution. Despite the efforts in EEG field, aiming at correlating the spectral features of EEG signals with DMN, an electrophysiological correlate of the DMN hasn't yet been found. In this study we used advanced techniques for functional connectivity estimation for describing the neuroelectrical properties of DMN. We analyzed the connectivity patterns elicited during the rest condition by 55 healthy subjects by means of Partial Directed Coherence. We extracted some graph indexes in order to describe the properties of the resting network in terms of local and global efficiencies, symmetries and influences between different regions of the scalp. Results highlighted the presence of a consistent network, elicited by more than 70% of analyzed population, involving mainly frontal and parietal regions. The properties of the resting network are uniform among the population and could be used for the construction of a normative database for the identification of pathological conditions.

Time-varying functional connectivity for understanding the neural basis of behavioral microsleeps.

Episodes of complete failure to respond during attentive tasks--lapses of responsiveness ('lapses')--accompanied by behavioral signs of sleep such as slow-eye-closure are known as behavioral microsleeps (BMs). The occurrence of BMs can have serious/fatal consequences, particularly in the transport sectors, and therefore further investigations on neurophysiological correlates of BMs are highly desirable. In this paper we propose a combination of High Resolution EEG techniques and an advanced method for time-varying functional connectivity estimation for reconstructing the temporal evolution of causal relations between cortical regions of BMs occurring during a visuomotor tracking task. The preliminary results highlight connectivity patterns involving parietal and fronto-parietal areas both preceding and following the onset of a BM.

Cortical activity and functional hyperconnectivity by simultaneous EEG recordings from interacting couples of professional pilots.

Controlling an aircraft during a flight is a compelling condition, which requires a strict and well coded interaction between the crew. The interaction level between the Captain and the First Officer changes during the flight, ranging from a maximum (during takeoff and landing, as well as in case of a failure of the instrumentation or other emergency situations) to a minimum during quiet mid-flight. In this study, our aim is to investigate the neural correlates of different kinds and levels of interaction between couples of professional crew members by means of the innovative technique called brain hyperscanning, i.e. the simultaneous recording of the hemodynamic or neuroelectrical activity of different human subjects involved in interaction tasks. This approach allows the observation and modeling of the neural signature specifically dependent on the interaction between subjects, and, even more interestingly, of the functional links existing between the brain activities of the subjects interacting together. In this EEG hyperscanning study, different phases of a flight were reproduced in a professional flight simulator, which allowed, on one side, to reproduce the ecological setting of a real flight, and, on the other, to keep under control the different levels of interaction induced in the crew by means of systematic and simulated failures of the aircraft instrumentation. Results of the procedure of linear inverse estimation, together with functional hyperconnectivity estimated by means of Partial Directed Coherence, showed a dense network of connections between the activity in the two brains in the takeoff and landing phases, when the cooperation between the crew is maximal, while conversely no significant links were shown during the phases in which the activity of the two pilots was independent.

Towards the time varying estimation of complex brain connectivity networks by means of a General Linear Kalman Filter approach.

One of the main limitations of the brain functional connectivity estimation methods based on Autoregressive Modeling, like the Granger Causality family of estimators, is the hypothesis that only stationary signals can be included in the estimation process. This hypothesis precludes the analysis of transients which often contain important information about the neural processes of interest. On the other hand, previous techniques developed for overcoming this limitation are affected by problems linked to the dimension of the multivariate autoregressive model (MVAR), which prevents from analysing complex networks like those at the basis of most cognitive functions in the brain. The General Linear Kalman Filter (GLKF) approach to the estimation of adaptive MVARs was recently introduced to deal with a high number of time series (up to 60) in a full multivariate analysis. In this work we evaluated the performances of this new method in terms of estimation quality and adaptation speed, by means of a simulation study in which specific factors of interest were systematically varied in the signal generation to investigate their effect on the method performances. The method was then applied to high density EEG data related to an imaginative task. The results confirmed the possibility to use this approach to study complex connectivity networks in a full multivariate and adaptive fashion, thus opening the way to an effective estimation of complex brain connectivity networks.

P300-based brain-computer interface for environmental control: an asynchronous approach.

Brain-computer interface (BCI) systems allow people with severe motor disabilities to communicate and interact with the external world. The P300 potential is one of the most used control signals for EEG-based BCIs. Classic P300-based BCIs work in a synchronous mode; the synchronous control assumes that the user is constantly attending to the stimulation, and the number of stimulation sequences is fixed a priori. This issue is an obstacle for the use of these systems in everyday life; users will be engaged in a continuous control state, their distractions will cause misclassification and the speed of selection will not take into account users' current psychophysical condition. An efficient BCI system should be able to understand the user's intentions from the ongoing EEG instead. Also, it has to refrain from making a selection when the user is engaged in a different activity and it should increase or decrease its speed of selection depending on the current user's state. We addressed these issues by introducing an asynchronous BCI and tested its capabilities for effective environmental monitoring, involving 11 volunteers in three recording sessions. Results show that this BCI system can increase the bit rate during control periods while the system is proved to be very efficient in avoiding false negatives when the users are engaged in other tasks.

Study of the functional hyperconnectivity between couples of pilots during flight simulation: an EEG hyperscanning study.

Brain Hyperscanning, i.e. the simultaneous recording of the cerebral activity of different human subjects involved in interaction tasks, is a very recent field of Neuroscience aiming at understanding the cerebral processes generating and generated by social interactions. This approach allows the observation and modeling of the neural signature specifically dependent on the interaction between subjects, and, even more interestingly, of the functional links existing between the activities in the brains of the subjects interacting together. In this EEG hyperscanning study we explored the functional hyperconnectivity between the activity in different scalp sites of couples of Civil Aviation Pilots during different phases of a flight reproduced in a flight simulator. Results shown a dense network of connections between the two brains in the takeoff and landing phases, when the cooperation between them is maximal, in contrast with phases during which the activity of the two pilots was independent, when no or quite few links were shown. These results confirms that the study of the brain connectivity between the activity simultaneously acquired in human brains during interaction tasks can provide important information about the neural basis of the "spirit of the group".

Testing the asymptotic statistic for the assessment of the significance of Partial Directed Coherence connectivity patterns.

Partial Directed Coherence (PDC) is a powerful tool to estimate a frequency domain description of Granger causality between multivariate time series. One of the main limitation of this estimator, however, has been so far the criteria used to assess the statistical significance, which have been obtained through surrogate data approach or arbitrarily imposed thresholds. The aim of this work is to test the performances of a validation approach based on the rigorous asymptotic distributions of PDC, recently proposed in literature. The performances of this method, defined in terms of percentages of false positives and false negatives, were evaluated by means of a simulation study taking into account factors like the Signal to Noise Ratio (SNR) and the amount of data available for the estimation and the use of different methods for the statistical corrections for multiple comparisons. Results of the Analysis Of Variance (ANOVA) performed on false positives and false negatives revealed a strong dependency of the performances from all the factors investigated. In particular, results indicate an amount of Type I errors below 7% for all conditions, while Type II errors are below 10% when the SNR is at least 1, the data length of at least 50 seconds and the appropriate correction for multiple comparisons is applied.

Imaging the social brain: multi-subjects EEG recordings during the "Chicken's game".

In this study we measured simultaneously by EEG hyperscannings the neuroelectric activity in 6 couples of subjects during the performance of the "Chicken's game", derived from game theory. The simultaneous recording of the EEG in couples of interacting subjects allows to observe and model directly the neural signature of human interactions in order to understand the cerebral processes generating and generated by social cooperation or competition. Results suggested that the one of the most consistently activated structure in this particular social interaction paradigm is the left orbitofrontal cortex. Connectivity results also showed a significant involvement of the orbitofrontal regions of both hemispheres across the observed population. Taken together, results confirms that the study of the brain activities in humans during social interactions can take benefit from the simultaneous acquisition of brain activity during such interaction.