Wyniki wyszukiwania - BazTech

Ograniczanie wyników

2 Biocybernetics and Biomedical Engineering

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: feature classification

Sortuj według:

Ogranicz wyniki do:

Artifact removal from EEG signals recorded in non-restricted environmentArtifact removal from EEG signals recorded in non-restricted environment

Jamil Zainab, Jamil Afshan, Majid Muhammad

Biocybernetics and Biomedical Engineering

2021

Vol. 41, no. 2

503--515

Electroencephalography (EEG) signals are always accompanied by endogenous and exogenous artifacts. Research carried out in the past few years focused on EEG artifact removal considered EEG signals recorded in a restricted lab environment. Considering the importance of EEG in daily life activities, no definitive approach is presented in removing blink artifacts from non-restricted EEG recordings. In this paper, a new supervised artifact removal method is proposed that classifies EEG chunks having eye movements and then utilizes independent component analysis and discrete wavelet transform to eliminate the ocular artifacts. The EEG data is acquired from 29 subjects in a non-restricted environment where the subject has to watch videos while walking and giving gestures and facial expressions. Thirteen morphological features are extracted from the recorded EEG signals to classify chunks with eye movements. The EEG chunks with eye movements are further processed to remove noise without distorting the morphology of signals. The proposed method is tested for eye movements and shows an improved performance in terms of correlation, mutual information, phase difference, and computational time over unsupervised modified multi-scale sample entropy and kurtosis, and wavelet enhanced independent component analysis based approaches. Moreover, the computed values of statistical parameters including sensitivity and specificity show the robustness of the proposed scheme.

A speech recognition system based on electromyography for the rehabilitation of dysarthric patients: A Thai syllable study

Sae Jong Nida, Phukpattaranont Pornchai

Biocybernetics and Biomedical Engineering

2019

Vol. 39, no. 1

234--245

The objective of this study is to develop a speech recognition system for classifying nine Thai syllables, which is used for the rehabilitation of dysarthric patients, based on five channels of surface electromyography (sEMG) signals from the human articulatory muscles. After the sEMG signal from each channel was collected, it was processed by a band-pass filter from 20–450 Hz for noise removal. Then, six features from three feature categories were determined and analyzed, namely, mean absolute value (MAV) and wavelength (WL) from amplitude based features (ABF), zero crossing (ZC) and mean frequency (MNF) from frequency based features (FBF), and L-kurtosis (L-KURT) and L-skewness (L-SKW) from statistics based features (SBF). Subsequently, a spectral regression extreme learning machine (SRELM) was used as the feature projection technique to reduce the dimension of feature vector from 30 to 8. Finally, the projected features were classified using a feed forward neural network (NN) classifier with 5-fold cross-validation. The proposed system was evaluated with the sEMG signals from seven healthy volunteers and five dysarthric volunteers. The results show that the proposed system can recognize the sEMG signals from both healthy and dysarthric volunteers. The average classification accuracies obtained from all six features in the healthy and dysarthric volunteers were 94.5% and 89.4%, respectively.