Brain-computer interfaces (BCIs) allow users to control a computer application by brain activity as acquired (e.g., by EEG). In our classic machine learning approach to BCIs, the participants undertake a calibration measurement without feedback to acquire data to train the BCI system. After the training, the user can control a BCI and improve the operation through some type of feedback. However, not all BCI users are able to perform sufficiently well during feedback operation. In fact, a nonnegligible portion of participants (estimated 15%–30%) cannot control the system (a BCI illiteracy problem, generic to all motor-imagery-based BCIs). We hypothesize that one main difficulty for a BCI user is the transition from offline calibration to online feedback. We investigate adaptive machine learning methods to eliminate offline calibration and analyze the performance of 11 volunteers in a BCI based on the modulation of sensorimotor rhythms. We present an adaptation scheme that individually guides the user. It starts with a subject-independent classifier that evolves to a subject-optimized state-of-the-art classifier within one session while the user interacts continuously. These initial runs use supervised techniques for robust coadaptive learning of user and machine. Subsequent runs use unsupervised adaptation to track the features’ drift during the session and provide an unbiased measure of BCI performance. Using this approach, without any offline calibration, six users, including one novice, obtained good performance after 3 to 6 minutes of adaptation. More important, this novel guided learning also allows participants with BCI illiteracy to gain significant control with the BCI in less than 60 minutes. In addition, one volunteer without sensorimotor idle rhythm peak at the beginning of the BCI experiment developed it during the course of the session and used voluntary modulation of its amplitude to control the feedback application.
"While Dura-Bernal and colleagues  have argued that such a biomimetic neuronal model could be used for Brain- Computer Interfaces (BCIs) to translate brain activity for control of a robotic limb, we want to use this model in the opposite direction. Adaptive decoding methods, that adapt to the users brain signals, have recently been shown to improve BCI performance , , . The currently fastest noninvasive BCI uses unsupervised learning and error-related potentials  for adaptation to reach information transfer rates of more than 140 bit/min which allows to write more than 20 error-free characters per minute , showing that adaptation is a key aspect in BCI. "
[Show abstract][Hide abstract] ABSTRACT: In this paper, we present a spiking neuronal model that learns to perform a motor control task. Since the long-term goal of this project is the application of such a neuronal model to study the mutual adaptation between a Brain-Computer Interface (BCI) and its user, neurobiological plausibility of the model is a key aspect. Therefore, the model was trained using reinforcement learning similar to that of the dopamine system, in which a global reward and punishment signal controlled spike-timing dependent plasticity (STDP). Based on this method, the majority of the randomly generated models were able to learn the motor control task. Although the models were only trained on two targets, they were able to reach arbitrary targets after learning. By introducing structural synaptic plasticity (SSP), which dynamically restructures the connections between neurons, the number of models that successfully learned the task could be significantly improved.
The 2015 International Joint Conference on Neural Networks; 07/2015
"There have been a number of recent efforts to learn improved adaptive decoders specifically tailored for the closed loop setting  , including an approach relying on stochastic optimal control theory . In other contexts, emphasis has been placed on training users to improve closed-loop control . Some efforts towards modeling the co-adaptation process have sought to model properties of different decoders when used in closed-loop   , with emphasis on ensuring the stability of the decoder and tuning the adaptation rate. "
[Show abstract][Hide abstract] ABSTRACT: In a closed-loop brain-computer interface (BCI), adaptive decoders are used to learn parameters suited to decoding the user’s neural response. Feedback to the user provides information which permits the neural tuning to also adapt. We present an approach to model this process of co-adaptation between the encoding model of the neural signal and the decoding algorithm as a multi-agent formulation of the linear quadratic Gaussian (LQG) control problem. In simulation we characterize how decoding performance improves as the neural encoding and adaptive decoder optimize, qualitatively resembling experimentally demonstrated closed-loop improvement. We then propose a novel, modified decoder update rule which is aware of the fact that the encoder is also changing and show it can improve simulated co-adaptation dynamics. Our modeling approach offers promise for gaining insights into co-adaptation as well as improving user learning of BCI control in practical settings.
Neural Information Processing Systems (NIPS); 12/2013
"Another work proposed adaptive classifiers based on the expectation-maximization method . In addition, some recent studies successfully used techniques for co-adaptive learning of user and machine , . Most of the algorithms in the latter group focused on extracting invariant features by regularizing the common spatial patterns (CSP) algorithm –. "
[Show abstract][Hide abstract] ABSTRACT: A major challenge in electroencephalogram (EEG)-based brain-computer interfaces (BCIs) is the inherent nonstationarities in the EEG data. Variations of the signal properties from intra and inter sessions often lead to deteriorated BCI performances, as features extracted by methods such as common spatial patterns (CSP) are not invariant against the changes. To extract features that are robust and invariant, this paper proposes a novel spatial filtering algorithm called Kullback-Leibler (KL) CSP. The CSP algorithm only considers the discrimination between the means of the classes, but does not consider within-class scatters information. In contrast, the proposed KLCSP algorithm simultaneously maximizes the discrimination between the class means, and minimizes the within-class dissimilarities measured by a loss function based on the KL divergence. The performance of the proposed KLCSP algorithm is compared against two existing algorithms, CSP and stationary CSP (sCSP), using the publicly available BCI competition III dataset IVa and a large dataset from stroke patients performing neuro-rehabilitation. The results show that the proposed KLCSP algorithm significantly outperforms both the CSP and the sCSP algorithms, in terms of classification accuracy, by reducing within-class variations. This results in more compact and separable features.
IEEE transactions on neural networks and learning systems 04/2013; 24(4):610-619. DOI:10.1109/TNNLS.2013.2239310 · 4.29 Impact Factor
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.