Natural language processing with dynamic classification improves P300 speller accuracy and bit rate

Abstract

The P300 speller is an example of a brain-computer interface that can restore functionality to victims of neuromuscular disorders. Although the most common application of this system has been communicating language, the properties and constraints of the linguistic domain have not to date been exploited when decoding brain signals that pertain to language. We hypothesized that combining the standard stepwise linear discriminant analysis with a Naive Bayes classifier and a trigram language model would increase the speed and accuracy of typing with the P300 speller. With integration of natural language processing, we observed significant improvements in accuracy and 40%–60% increases in bit rate for all six subjects in a pilot study. This study suggests that integrating information about the linguistic domain can significantly improve signal classification.

Utilizing a Language Model to Improve Online Dynamic Data Collection in P300 Spellers

“…Thus, it can be inferred that the significant improvement in communication rate observed with the addition of the language model would add to the significant improvements of a dynamic stopping system when compared to a static stopping system. The approach in this study is similar to that used by Speier et al, comparing static, dynamic and natural language processing (NLP) methods to control data collection [30]. They assumed Gaussian likelihood distributions (verified via Kolmogorov-Smirnov tests) for the target and nontarget classifier scores and used a trigram model obtained from the Brown corpus.…”

“…Speier et al performed offline simulations comparing P300 speller performance using static, dynamic and natural language processing (NLP) methods to control data collection [30]. The dynamic method was similar to the Throckmorton et al approach of initializing character probabilities with uniform prior and updating these probabilities via Bayesian inference until a threshold was reached.…”

Language-Model Assisted Brain Computer Interface for Typing: A Comparison of Matrix and Rapid Serial Visual Presentation

“…A probabilistic language model can be employed to incorporate predictive word completion during the intent detection process [19], [20], [21], or to define a prior on potential target characters during the classification task [22], [23], [24]. Our system, the RSVP Keyboard TM , originally developed based on the RSVP paradigm and now also features the matrix presentation paradigm, probabilistically fuses context evidence with physiological evidence to infer user intent.…”

Review of Brain Computer Interface Components and Corresponding Research Topics

“…Reference [113] uses the area under ROC curve to find the optimal linear discriminant as the classifier. Reference [114] combines language model with the classifier to optimize the classification of P300 character speller. What's more, there are papers about the optimization of SVM [115][116][117][118] and ANN [119][120][121].…”