Josef Faller

Autocalibration and Recurrent Adaptation: Towards a Plug and Play Online ERD-BCI

System calibration and user training are essential for operating motor imagery based brain-computer interface (BCI) systems. These steps are often unintuitive and tedious for the user, and do not necessarily lead to a satisfactory level of control. We present an Adaptive BCI framework that provides feedback after only minutes of autocalibration in a two-class BCI setup. During operation, the system recurrently reselects only one out of six predefined logarithmic bandpower features (10-13 and 16-24 Hz from Laplacian derivations over C3, Cz, and C4), specifically, the feature that exhibits maximum discriminability. The system then retrains a linear discriminant analysis classifier on all available data and updates the online paradigm with the new model. Every retraining step is preceded by an online outlier rejection. Operating the system requires no engineering knowledge other than connecting the user and starting the system. In a supporting study, ten out of twelve novice users reached a criterion level of above 70% accuracy in one to three sessions (10-80 min online time) of training, with a median accuracy of 80.2 11.3% in the last session. We consider the presented system a positive first step towards fully autocalibrating motor imagery BCIs.

An application framework for controlling an avatar in a desktop-based virtual environment via a software ssvep brain--computer interface

This paper presents a reusable, highly configurable application framework that seamlessly integrates SSVEP stimuli within a desktop-based virtual environment (VE) on standard PC equipment. Steady-state visual evoked potentials (SSVEPs) are brain signals that offer excellent information transfer rates (ITR) within braincomputer interface (BCI) systems while requiring only minimal training. Generating SSVEP stimuli in a VE allows for an easier implementation of motivating training paradigms and more realistic simulations of real-world applications. EEG measurements on seven healthy subjects within three scenarios (Button, Slalom, and Apartment) showed that moving and static software generated SSVEP stimuli flickering at frequencies of up to 29 Hz proved suitable to elicit SSVEPs. This research direction could lead to vastly improved immersive VEs that allow both disabled and healthy users to seamlessly communicate or interact through an intuitive, natural, and friendly interface.

Combining BCI with Virtual Reality: Towards New Applications and Improved BCI

Brain–Computer Interfaces (BCI) are communication systems which can convey messages through brain activity alone. Recently BCIs were gaining interest among the virtual reality (VR) community since they have appeared as promising interaction devices for virtual environments (VEs). Especially these implicit interaction techniques are of great interest for the VR community, e.g., you are imaging the movement of your hand and the virtual hand is moving, or you can navigate through houses or museums by your thoughts alone or just by looking at some highlighted objects. Furthermore, VE can provide an excellent testing ground for procedures that could be adapted to real world scenarios, especially patients with disabilities can learn to control their movements or perform specific tasks in a VE. Several studies will highlight these interactions.

A co-adaptive brain-computer interface for end users with severe motor impairment.

Co-adaptive training paradigms for event-related desynchronization (ERD) based brain-computer interfaces (BCI) have proven effective for healthy users. As of yet, it is not clear whether co-adaptive training paradigms can also benefit users with severe motor impairment. The primary goal of our paper was to evaluate a novel cue-guided, co-adaptive BCI training paradigm with severely impaired volunteers. The co-adaptive BCI supports a non-control state, which is an important step toward intuitive, self-paced control. A secondary aim was to have the same participants operate a specifically designed self-paced BCI training paradigm based on the auto-calibrated classifier. The co-adaptive BCI analyzed the electroencephalogram from three bipolar derivations (C3, Cz, and C4) online, while the 22 end users alternately performed right hand movement imagery (MI), left hand MI and relax with eyes open (non-control state). After less than five minutes, the BCI auto-calibrated and proceeded to provide visual feedback for the MI task that could be classified better against the non-control state. The BCI continued to regularly recalibrate. In every calibration step, the system performed trial-based outlier rejection and trained a linear discriminant analysis classifier based on one auto-selected logarithmic band-power feature. In 24 minutes of training, the co-adaptive BCI worked significantly (p = 0.01) better than chance for 18 of 22 end users. The self-paced BCI training paradigm worked significantly (p = 0.01) better than chance in 11 of 20 end users. The presented co-adaptive BCI complements existing approaches in that it supports a non-control state, requires very little setup time, requires no BCI expert and works online based on only two electrodes. The preliminary results from the self-paced BCI paradigm compare favorably to previous studies and the collected data will allow to further improve self-paced BCI systems for disabled users.

Brain-controlled applications using dynamic P300 speller matrices

OBJECTIVES Access to the world wide web and multimedia content is an important aspect of life. We present a web browser and a multimedia user interface adapted for control with a brain-computer interface (BCI) which can be used by severely motor impaired persons. METHODS The web browser dynamically determines the most efficient P300 BCI matrix size to select the links on the current website. This enables control of the web browser with fewer commands and smaller matrices. The multimedia player was based on an existing software. Both applications were evaluated with a sample of ten healthy participants and three end-users. All participants used a visual P300 BCI with face-stimuli for control. RESULTS The healthy participants completed the multimedia player task with 90% accuracy and the web browsing task with 85% accuracy. The end-users completed the tasks with 62% and 58% accuracy. All healthy participants and two out of three end-users reported that they felt to be in control of the system. CONCLUSIONS In this study we presented a multimedia application and an efficient web browser implemented for control with a BCI. SIGNIFICANCE Both applications provide access to important areas of modern information retrieval and entertainment.

What does clean EEG look like

Lack of a clear analytical metric for identifying artifact free, clean electroencephalographic (EEG) signals inhibits robust comparison of different artifact removal methods and lowers confidence in the results of EEG analysis. An algorithm is presented for identifying clean EEG epochs by thresholding statistical properties of the EEG. Thresholds are trained on EEG datasets from both healthy subjects and stroke/spinal cord injury patient populations via differential evolution (DE).

Random forests in non-invasive sensorimotor rhythm brain-computer interfaces: a practical and convenient non-linear classifier

Abstract There is general agreement in the brain-computer interface (BCI) community that although non-linear classifiers can provide better results in some cases, linear classifiers are preferable. Particularly, as non-linear classifiers often involve a number of parameters that must be carefully chosen. However, new non-linear classifiers were developed over the last decade. One of them is the random forest (RF) classifier. Although popular in other fields of science, RFs are not common in BCI research. In this work, we address three open questions regarding RFs in sensorimotor rhythm (SMR) BCIs: parametrization, online applicability, and performance compared to regularized linear discriminant analysis (LDA). We found that the performance of RF is constant over a large range of parameter values. We demonstrate – for the first time – that RFs are applicable online in SMR-BCIs. Further, we show in an offline BCI simulation that RFs statistically significantly outperform regularized LDA by about 3%. These results confirm that RFs are practical and convenient non-linear classifiers for SMR-BCIs. Taking into account further properties of RFs, such as independence from feature distributions, maximum margin behavior, multiclass and advanced data mining capabilities, we argue that RFs should be taken into consideration for future BCIs.

Control or non-control state: that is the question! An asynchronous visual P300-based BCI approach

OBJECTIVE Brain-computer interfaces (BCI) based on event-related potentials (ERP) were proven to be a reliable synchronous communication method. For everyday life situations, however, this synchronous mode is impractical because the system will deliver a selection even if the user is not paying attention to the stimulation. So far, research into attention-aware visual ERP-BCIs (i.e., asynchronous ERP-BCIs) has led to variable success. In this study, we investigate new approaches for detection of user engagement. APPROACH Classifier output and frequency-domain features of electroencephalogram signals as well as the hybridization of them were used to detect the users state. We tested their capabilities for state detection in different control scenarios on offline data from 21 healthy volunteers. MAIN RESULTS The hybridization of classifier output and frequency-domain features outperformed the results of the single methods, and allowed building an asynchronous P300-based BCI with an average correct state detection accuracy of more than 95%. SIGNIFICANCE Our results show that all introduced approaches for state detection in an asynchronous P300-based BCI can effectively avoid involuntary selections, and that the hybrid method is the most effective approach.

Brain-Computer Interface Systems Used for Virtual Reality Control

A Brain-Computer Interface (BCI) is a non-muscular communication channel for connecting the brain to a computer or another device. Currently, non-invasive BCIs transform thoughtrelated changes in the electroencephalogram (EEG) online and in real time into control signals. In such an EEG-based BCI, specific features are extracted from brain-signals, transformed into a control signal, and used to restore communication to patients with locked-in-syndrome or to control neuroprosthesis in patients with spinal cord injuries (Birbaumer et al., 1999; Pfurtscheller et al., 2008b; Wolpaw et al., 2002). In addition to these applications, which focus on communication and control, the related field of neurofeedback supports feedback training in people suffering from epilepsy, autism, stroke, and emotional or attentional disorders (Birbaumer & Cohen, 2007). Today the world of BCI applications is expanding and new fields are opening. One new direction involves BCIs to control virtual reality (VR), including BCIs for games, or using VR as a powerful feedback medium to reduce the need for BCI training (Leeb et al., 2007b; Scherer et al., 2008). Virtual environments (VE) can provide an excellent testing ground for procedures that could be adapted to real world scenarios, especially for patients with disabilities. If people can learn to control their movements or perform specific tasks in a VE, this could justify the much greater expense of building physical devices such as a wheelchair or robot arm that is controlled by a BCI. BCIs are more and more moving out of the laboratory and becoming also useful for healthy users in certain situations (Nijholt et al., 2008). One of the first efforts to combine VR and BCI technologies was Bayliss and Ballard (2000) and Bayliss (2003). They introduced a VR smart home in which users could control different appliances using a P300-based BCI. Pineda et al. (2003) showed that immersive feedback based on a computer game can help people learn to control a BCI based on imagined movement more quickly than mundane feedback, a finding we later validated with other immersive feedback (Leeb et al., 2006; 2007b). Lalor et al. (2005) used a steady-state visual evoked potential (SSVEP)-based BCI to control a character in an immersive 3-D gaming environment. Recently, Leeb et al. (2007b) have reported on exploring a smart virtual

Exploration of the neural correlates of cerebral palsy for sensorimotor BCI control.

Cerebral palsy (CP) includes a broad range of disorders, which can result in impairment of posture and movement control. Brain-computer interfaces (BCIs) have been proposed as assistive devices for individuals with CP. Better understanding of the neural processing underlying motor control in affected individuals could lead to more targeted BCI rehabilitation and treatment options. We have explored well-known neural correlates of movement, including event-related desynchronization (ERD), phase synchrony, and a recently-introduced measure of phase dynamics, in participants with CP and healthy control participants. Although present, significantly less ERD and phase locking were found in the group with CP. Additionally, inter-group differences in phase dynamics were also significant. Taken together these findings suggest that users with CP exhibit lower levels of motor cortex activation during motor imagery, as reflected in lower levels of ongoing mu suppression and less functional connectivity. These differences indicate that development of BCIs for individuals with CP may pose additional challenges beyond those faced in providing BCIs to healthy individuals.