Brain computer interfaces rely on machine learning (ML) algorithms to decode the brain's electrical activity into decisions. For example, in rapid serial visual presentation (RSVP) tasks, the subject is presented with a continuous stream of images containing rare target images among standard images, while the algorithm has to detect brain activity associated with target images. Here, we continue our previous work, presenting a deep neural network model for the use of single trial EEG classification in RSVP tasks. Deep neural networks have shown state of the art performance in computer vision and speech recognition and thus have great promise for other learning tasks, like classification of EEG samples. In our model, we introduce a novel spatio-temporal regularization for EEG data to reduce overfitting. We show improved classification performance compared to our earlier work on a five categories RSVP experiment. In addition, we compare performance on data from different sessions and validate the model on a public benchmark data set of a P300 speller task. Finally, we discuss the advantages of using neural network models compared to manually designing feature extraction algorithms.

Brain computer interface applications, developed for both healthy and clinical populations, critically depend on decoding brain activity in single trials. The goal of the present study was to detect distinctive spatiotemporal brain patterns within a set of event related responses. We introduce a novel classification algorithm, the spatially weighted FLD-PCA (SWFP), which is based on a two-step linear classification of event-related responses, using fisher linear discriminant (FLD) classifier and principal component analysis (PCA) for dimensionality reduction. As a benchmark algorithm, we consider the hierarchical discriminant component Analysis (HDCA), introduced by Parra, et al. 2007. We also consider a modified version of the HDCA, namely the hierarchical discriminant principal component analysis algorithm (HDPCA). We compare single-trial classification accuracies of all the three algorithms, each applied to detect target images within a rapid serial visual presentation (RSVP, 10 Hz) of images from five different object categories, based on single-trial brain responses. We find a systematic superiority of our classification algorithm in the tested paradigm. Additionally, HDPCA significantly increases classification accuracies compared to the HDCA. Finally, we show that presenting several repetitions of the same image exemplars improve accuracy, and thus may be important in cases where high accuracy is crucial.

Brain computer interfaces allow users to preform various tasks using only the electrical activity of the brain. BCI applications often present the user a set of stimuli and record the corresponding electrical response. The BCI algorithm will then have to decode the acquired brain response and perform the desired task. In rapid serial visual presentation (RSVP) tasks, the subject is presented with a continuous stream of images containing rare target images among standard images, while the algorithm has to detect brain activity associated with target images. In this work, we suggest a multimodal neural network for RSVP tasks. The network operates on the brain response and on the initiating stimulus simultaneously, providing more information for the BCI application. We present two variants of the multimodal network, a supervised model, for the case when the targets are known in advanced, and a semi-supervised model for when the targets are unknown. We test the neural networks with a RSVP experiment on satellite imagery carried out with two subjects. The multimodal networks achieve a significant performance improvement in classification metrics. We visualize what the networks has learned and discuss the advantages of using neural network models for BCI applications.

No abstract