scholarly journals Improving P300 Speller performance by means of optimization and machine learning

2021 ◽  
Author(s):  
Luigi Bianchi ◽  
Chiara Liti ◽  
Giampaolo Liuzzi ◽  
Veronica Piccialli ◽  
Cecilia Salvatore
Keyword(s):  
Machine Learning ◽  
Signal To Noise Ratio ◽  
Event Related Potentials ◽  
Support Vector ◽  
Early Stopping ◽  
Signal To Noise ◽  
Linear Discriminant ◽  
Related Potentials ◽  
Noise Ratio ◽  
Number Of Iterations

AbstractBrain-Computer Interfaces (BCIs) are systems allowing people to interact with the environment bypassing the natural neuromuscular and hormonal outputs of the peripheral nervous system (PNS). These interfaces record a user’s brain activity and translate it into control commands for external devices, thus providing the PNS with additional artificial outputs. In this framework, the BCIs based on the P300 Event-Related Potentials (ERP), which represent the electrical responses recorded from the brain after specific events or stimuli, have proven to be particularly successful and robust. The presence or the absence of a P300 evoked potential within the EEG features is determined through a classification algorithm. Linear classifiers such as stepwise linear discriminant analysis and support vector machine (SVM) are the most used discriminant algorithms for ERPs’ classification. Due to the low signal-to-noise ratio of the EEG signals, multiple stimulation sequences (a.k.a. iterations) are carried out and then averaged before the signals being classified. However, while augmenting the number of iterations improves the Signal-to-Noise Ratio, it also slows down the process. In the early studies, the number of iterations was fixed (no stopping environment), but recently several early stopping strategies have been proposed in the literature to dynamically interrupt the stimulation sequence when a certain criterion is met in order to enhance the communication rate. In this work, we explore how to improve the classification performances in P300 based BCIs by combining optimization and machine learning. First, we propose a new decision function that aims at improving classification performances in terms of accuracy and Information Transfer Rate both in a no stopping and early stopping environment. Then, we propose a new SVM training problem that aims to facilitate the target-detection process. Our approach proves to be effective on several publicly available datasets.

scholarly journals Capturing Human Perceptual and Cognitive Activities via Event-Related Potentials Measured with Candle-Like Dry Microneedle Electrodes

Micromachines ◽  
2020 ◽  
Vol 11 (6) ◽  
pp. 556
Author(s):  
Yuri Yoshida ◽  
Takumi Kawana ◽  
Eiichi Hoshino ◽  
Yasuyo Minagawa ◽  
Norihisa Miki
Keyword(s):  
Near Infrared ◽  
Signal To Noise Ratio ◽  
Cognitive Activity ◽  
Event Related Potentials ◽  
Signal To Noise ◽  
Skin Preparation ◽  
Auditory Oddball ◽  
Related Potentials ◽  
Noise Ratio ◽  
Noise Contamination

We demonstrate capture of event-related potentials (ERPs) using candle-like dry microneedle electrodes (CMEs). CMEs can record an electroencephalogram (EEG) even from hairy areas without any skin preparation, unlike conventional wet electrodes. In our previous research, we experimentally verified that CMEs can measure the spontaneous potential of EEG from the hairy occipital region without preparation with a signal-to-noise ratio as good as that of the conventional wet electrodes which require skin preparation. However, these results were based on frequency-based signals, which are relatively robust compared to noise contamination, and whether CMEs are sufficiently sensitive to capture finer signals remained unclear. Here, we first experimentally verified that CMEs can extract ERPs as good as conventional wet electrodes without preparation. In the auditory oddball tasks using pure tones, P300, which represent ERPs, was extracted with a signal-to-noise ratio as good as that of conventional wet electrodes. CMEs successfully captured perceptual activities. Then, we attempted to investigate cerebral cognitive activity using ERPs. In processing the vowel and prosody in auditory stimuli such as /itta/, /itte/, and /itta?/, laterality was observed that originated from the locations responsible for the process in near-infrared spectroscopy (NIRS) and magnetoencephalography experiments. We simultaneously measured ERPs with CMEs and NIRS in the oddball tasks using the three words. Laterality appeared in NIRS for six of 10 participants, although laterality was not clearly shown in the results, suggesting that EEGs have a limitation of poor spatial resolution. On the other hand, successful capturing of MMN and P300 using CMEs that do not require skin preparation may be readily applicable for real-time applications of human perceptual activities.

Signal-to-noise ratio of chemosensory event-related potentials

2007 ◽  
Vol 118 (3) ◽  
pp. 690-695 ◽  
Author(s):  
Sanne Boesveldt ◽  
Antje Haehner ◽  
Henk W. Berendse ◽  
Thomas Hummel
Keyword(s):  
Signal To Noise Ratio ◽  
Event Related Potentials ◽  
Signal To Noise ◽  
Related Potentials ◽  
Noise Ratio

Quality estimation of subdurally recorded, event-related potentials based on signal-to-noise ratio

2002 ◽  
Vol 49 (1) ◽  
pp. 31-40 ◽  
Author(s):  
M.M. Rohde ◽  
S.L. BeMent ◽  
J.E. Huggins ◽  
S.P. Levine ◽  
R.K. Kushwaha ◽  
...  
Keyword(s):  
Signal To Noise Ratio ◽  
Event Related Potentials ◽  
Signal To Noise ◽  
Quality Estimation ◽  
Related Potentials ◽  
Noise Ratio

scholarly journals Assessment of chemosensory function using electroencephalographic techniques

2012 ◽  
Vol 50 (1) ◽  
pp. 13-21
Author(s):  
Ph. Rombaux ◽  
C. Huart ◽  
A. Mouraux
Keyword(s):  
Signal To Noise Ratio ◽  
Event Related Potentials ◽  
Objective Evaluation ◽  
Stimulus Onset ◽  
Signal To Noise ◽  
Brain Responses ◽  
Related Potentials ◽  
Noise Ratio ◽  
Time Domain Averaging ◽  
Chemosensory Function

Electroencephalographic techniques are widely used to provide an objective evaluation of the chemosensory function and to explore neural mechanisms related to the processing of chemosensory events. The most popular technique to evaluate brain responses to chemosensory stimuli is across trial time-domain averaging to reveal chemosensory event-related potentials (CSERP) embedded within the ongoing EEG. Nevertheless, this technique has a poor signal-to-noise ratio and cancels out stimulus-induced changes in the EEG signal that are not strictly phased-locked to stimulus onset. The fact that consistent CSERP are not systematically identifiable in healthy subjects currently constitutes a major limitation to the use of this technique for the diagnosis of chemosensory dysfunction. In this review, we will review the different techniques related to the recording and identification of CSERP, discuss some of their limitations, and propose some novel signal processing methods which could be used to enhance the signal-to-noise ratio of chemosensory event-related brain responses.

scholarly journals Ensemble Voting-Based Multichannel EEG Classification in a Subject-Independent P300 Speller

2021 ◽  
Vol 11 (23) ◽  
pp. 11252
Author(s):  
Ayana Mussabayeva ◽  
Prashant Kumar Jamwal ◽  
Muhammad Tahir Akhtar
Keyword(s):  
Healthy Subjects ◽  
Single Channel ◽  
Event Related Potentials ◽  
Support Vector ◽  
Ensemble Classifiers ◽  
K Nearest Neighbors ◽  
Linear Discriminant ◽  
P300 Speller ◽  
Related Potentials ◽  
Als Patients

Classification of brain signal features is a crucial process for any brain–computer interface (BCI) device, including speller systems. The positive P300 component of visual event-related potentials (ERPs) used in BCI spellers has individual variations of amplitude and latency that further changse with brain abnormalities such as amyotrophic lateral sclerosis (ALS). This leads to the necessity for the users to train the speller themselves, which is a very time-consuming procedure. To achieve subject-independence in a P300 speller, ensemble classifiers are proposed based on classical machine learning models, such as the support vector machine (SVM), linear discriminant analysis (LDA), k-nearest neighbors (kNN), and the convolutional neural network (CNN). The proposed voters were trained on healthy subjects’ data using a generic training approach. Different combinations of electroencephalography (EEG) channels were used for the experiments presented, resulting in single-channel, four-channel, and eight-channel classification. ALS patients’ data represented robust results, achieving more than 90% accuracy when using an ensemble of LDA, kNN, and SVM on four active EEG channels data in the occipital area of the brain. The results provided by the proposed ensemble voting models were on average about 5% more accurate than the results provided by the standalone classifiers. The proposed ensemble models could also outperform boosting algorithms in terms of computational complexity or accuracy. The proposed methodology shows the ability to be subject-independent, which means that the system trained on healthy subjects can be efficiently used for ALS patients. Applying this methodology for online speller systems removes the necessity to retrain the P300 speller.

scholarly journals Near-field sound source localization using principal component analysis–multi-output support vector regression

2020 ◽  
Vol 16 (4) ◽  
pp. 155014772091640
Author(s):  
Lanmei Wang ◽  
Yao Wang ◽  
Guibao Wang ◽  
Jianke Jia
Keyword(s):  
Principal Component Analysis ◽  
Support Vector Regression ◽  
Covariance Matrix ◽  
Signal To Noise Ratio ◽  
Near Field ◽  
Principal Component ◽  
Component Analysis ◽  
Support Vector ◽  
Signal To Noise ◽  
Noise Ratio

In this article, principal component analysis method, which is applied to image compression and feature extraction, is introduced into the dimension reduction of input characteristic variable of support vector regression, and a method of joint estimation of near-field angle and range based on principal component analysis dimension reduction is proposed. Signal-to-noise ratio and calculation amount are the decisive factors affecting the performance of the algorithm. Principal component analysis is used to fuse the main characteristics of training data and discard redundant information, the signal-to-noise ratio is improved, and the calculation amount is reduced accordingly. Similarly, support vector regression is used to model the signal, and the upper triangular elements of the signal covariance matrix are usually used as input features. Since the covariance matrix has more upper triangular elements, training it as a feature input will affect the training speed to some extent. Principal component analysis is used to reduce the dimensionality of the upper triangular element of the covariance matrix of the known signal, and it is used as the input feature of the multi-output support vector regression machine to construct the near-field parameter estimation model, and the parameter estimation of unknown signal is herein obtained. Simulation results show that this method has high estimation accuracy and training speed, and has strong adaptability at low signal-to-noise ratio, and the performance is better than that of the back-propagation neural network algorithm and the two-step multiple signal classification algorithm.

Sign in / Sign up

Export Citation Format

Share Document