Adaptive LDA Classifier Enhances Real-Time Control of an EEG Brain–Computer Interface for Imagined-Syllables Decoding

Shizhe Wu,Kinkini Bhadra,Anne-Lise Giraud,Silvia Marchesotti

DOI: https://doi.org/10.3390/brainsci14030196

IF: 3.333

2024-02-22

Brain Sciences

Abstract:Brain-Computer Interfaces (BCIs) aim to establish a pathway between the brain and an external device without the involvement of the motor system, relying exclusively on neural signals. Such systems have the potential to provide a means of communication for patients who have lost the ability to speak due to a neurological disorder. Traditional methodologies for decoding imagined speech directly from brain signals often deploy static classifiers, that is, decoders that are computed once at the beginning of the experiment and remain unchanged throughout the BCI use. However, this approach might be inadequate to effectively handle the non-stationary nature of electroencephalography (EEG) signals and the learning that accompanies BCI use, as parameters are expected to change, and all the more in a real-time setting. To address this limitation, we developed an adaptive classifier that updates its parameters based on the incoming data in real time. We first identified optimal parameters (the update coefficient, UC) to be used in an adaptive Linear Discriminant Analysis (LDA) classifier, using a previously recorded EEG dataset, acquired while healthy participants controlled a binary BCI based on imagined syllable decoding. We subsequently tested the effectiveness of this optimization in a real-time BCI control setting. Twenty healthy participants performed two BCI control sessions based on the imagery of two syllables, using a static LDA and an adaptive LDA classifier, in randomized order. As hypothesized, the adaptive classifier led to better performances than the static one in this real-time BCI control task. Furthermore, the optimal parameters for the adaptive classifier were closely aligned in both datasets, acquired using the same syllable imagery task. These findings highlight the effectiveness and reliability of adaptive LDA classifiers for real-time imagined speech decoding. Such an improvement can shorten the training time and favor the development of multi-class BCIs, representing a clear interest for non-invasive systems notably characterized by low decoding accuracies.

neurosciences

What problem does this paper attempt to address?

The main goal of this paper is to address the challenges encountered in real-time decoding of imagined speech in brain-computer interfaces (BCI), particularly for binary classification BCI based on imagined syllables. Traditional methods typically use static classifiers for decoding, where the classifier parameters are computed once at the beginning of the experiment and remain unchanged throughout the BCI usage. However, this approach may be insufficient to effectively handle the non-stationarity of electroencephalogram (EEG) signals and the parameter changes during the continuous learning process associated with BCI usage.

To address the aforementioned issues, the research team developed an adaptive Linear Discriminant Analysis (LDA) classifier that can update its parameters based on the data received in real-time. This approach is expected to improve BCI control performance and reduce training time, which is particularly important for non-invasive systems, as these systems often have lower decoding accuracy.

Specifically, the research first utilized a previously recorded EEG dataset to determine the optimal parameters (update coefficient, UC) for the adaptive LDA classifier. Subsequently, the effectiveness of these optimized parameters was tested in a real-time BCI control setting. The study recruited 20 healthy participants for two BCI control sessions, using static LDA and adaptive LDA classifiers, respectively. The results showed that the adaptive classifier performed better in real-time BCI control tasks compared to the static classifier. Moreover, the optimal parameters for the adaptive classifier demonstrated consistency across both datasets, further validating the effectiveness and reliability of the adaptive LDA classifier.

In conclusion, by introducing the adaptive LDA classifier, the research significantly improved the real-time control performance of syllable-based imagined speech BCI, paving the way for the development of multi-class BCI systems and helping to improve the decoding accuracy of non-invasive systems. This advancement is particularly important for patients who have lost their speech ability due to neurological disorders, as it can provide an effective means of communication.