Analysis of Time – Frequency EEG Feature Extraction Methods for Mental Task Classification

Uyulan, Çağlar; Ergüzel, Türker Tekin

doi:10.2991/ijcis.10.1.87

Cited by 20 publications

(8 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, time-frequency analyses are very useful in extracting the features of these signals. Wavelet transform is one of the most practical time-frequency methods to extract features from EEG signals [20][21][22]. In this study, we applied Discrete Wavelet Transform (DWT) to EEG signals.…”

Section: Feature Extractionmentioning

confidence: 99%

A Novel Channel Selection Approach for Human Neonate’s Pain EEG Data Analysis

Talebi

Frounchi

Tazehkand

2022

Preprint

View full text Add to dashboard Cite

Quantitative measurement of pain using the Electroencephalogram (EEG) signals has received much attention, recently. Pain EEG data processing is associated with complexity and high computational cost. This study aims to propose a new method for selecting efficient EEG channels to determine the area of the scalp that contains the most information about brain activity during acute pain in neonates. Also, selecting relevant channels in pain assessment reduces computational costs. In this study, a new channel selection approach is proposed, which is a combination of filter and wrapper methods. A new pseudo-Sequential Forward Feature Selection (pseudo-SFFS) method is presented to reduce the computational complexity of wrapper methods. We preprocessed data by applying a bandpass filter. We used wavelet transform to extract features. After extracting the features, we applied two feature selection steps. In the first step, we applied the T-test to the extracted features. In the second step, we selected the effective channels based on the output of the applied pseudo-SFFS algorithm into Support Vector Machine (SVM), Decision Tree (DT), and Gaussian Naive Bayesian (GNB) classifiers. Using the proposed method two channels of the sensorimotor cortex including Cz and C4 channels have been selected from 18 EEG channels for pain stimulation through the left heel of neonates. Also, the results show that most of the acute pain information of neonates is related to the delta and theta frequency bands.

show abstract

Section: Feature Extractionmentioning

confidence: 99%

A Novel Channel Selection Approach for Human Neonate’s Pain EEG Data Analysis

Talebi

Frounchi

Tazehkand

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Generally, feature extraction methods for the MI-EEG focus on deriving time-domain features, i.e. energy and amplitude of the signal, autoregressive modelling [Fryz, 2017; Ahmad&Aqil, 2016], and on establishing frequency domain features [Lee et.al, 2014; Lu et.al, 2017] or on extracting time-frequency features [Uyulan et.al, 2019; Uyulan&Erguzel, 2017]. With its adaptive structure and the ability for analyzing the non-stationary signals wavelet transform (WT) keeps and processes both time and frequency components of the signal.…”

Section: Introductionmentioning

confidence: 99%

Development of LSTM&CNN Based Hybrid Deep Learning Model to Classify Motor Imagery Tasks

Uyulan

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Recent studies underline the contribution of brain-computer interface (BCI) applications to the enhancement process of the life quality of physically impaired subjects. In this context, to design an effective stroke rehabilitation or assistance system, the classification of motor imagery (MI) tasks are performed through deep learning (DL) algorithms. Although the utilization of DL in the BCI field remains relatively premature as compared to the fields related to natural language processing, object detection, etc., DL has proven its effectiveness in carrying out this task.In this paper, a hybrid method, which fuses the one-dimensional convolutional neural network (1D CNN) with the long short-term memory (LSTM), was performed for classifying four different MI tasks, i.e. left hand, right hand, tongue, and feet movements. The time representation of MI tasks is extracted through the hybrid deep learning model training after principal component analysis (PCA)-based artefact removal process. The performance criteria given in the BCI Competition IV dataset A are estimated. 10-folded Cross-validation (CV) results show that the proposed method outperforms in classifying electroencephalogram (EEG)-electrooculogram (EOG) combined motor imagery tasks compared to the state of art methods and is robust against data variations. The CNN-LSTM classification model reached 95.62 % (±1.2290742) accuracy and 0.9462 (±0.01216265) kappa value for datasets with four MI-based class validated using 10-fold CV. Also, the receiver operator characteristic (ROC) curve, the area under the ROC curve (AUC) score, and confusion matrix are evaluated for further interpretations.

show abstract

“…For many cases, one of the most commonly biological signals analysis using MSE is the Electroencephalogram signal (EEG). The measurements of brain functions through EEG can be used for monitoring and interpreting the brain activity, even predicting the outcomes [ 3 ]. MSE was used for the analysis of EEG signals monitoring the depth of the anesthetic process during surgery [ 4 ].…”

Section: Introductionmentioning

confidence: 99%

Sample Entropy on Multidistance Signal Level Difference for Epileptic EEG Classification

Rizal

Hadiyoso

2018

The Scientific World Journal

View full text Add to dashboard Cite

Epilepsy is a disorder of the brain's nerves as a result of excessive brain cell activity. It is generally characterized by the recurrent unprovoked seizures. This neurological abnormality can be detected and evaluated using Electroencephalogram (EEG) signal. Many algorithms have been applied to achieve high performance for the EEG classification of epileptic. However, the complexity and randomness of EEG signals become a challenge to researchers in applying the appropriate algorithms. In this research, sample entropy on Multidistance Signal Level Difference (MSLD) was applied to obtain the characteristic of EEG signals, especially towards the epilepsy patients. The test was performed on three classes of EEG data: EEG signals of epilepsy patient in ictal (seizure), interictal conditions (occurring between seizures) and normal EEG signals from healthy subjects with a closed eye condition. In this study, classification and verification were done using the Support Vector Machine (SVM) method. Through the 5-fold cross-validation, experimental results showed the highest accuracy of 97.7%.

show abstract

Analysis of Time – Frequency EEG Feature Extraction Methods for Mental Task Classification

Cited by 20 publications

References 35 publications

A Novel Channel Selection Approach for Human Neonate’s Pain EEG Data Analysis

A Novel Channel Selection Approach for Human Neonate’s Pain EEG Data Analysis

Development of LSTM&CNN Based Hybrid Deep Learning Model to Classify Motor Imagery Tasks

Sample Entropy on Multidistance Signal Level Difference for Epileptic EEG Classification

Contact Info

Product

Resources

About