Unsupervised Learning for Exploring Hidden Structures in Self-Talk DOI

Kellen Tyrrell,

Masoumeh Heidari Kapourchali

Published: Oct. 9, 2023

Innerspeech decoding from EEG data holds significant importance due to its potential revolutionize human-machine interaction and communication systems. Leveraging the power of temporal shift-invariant sparse coding, this study explores unsupervised learning inner-speech patterns using EEG, a prominent modality in body sensor networks. By analyzing data, we investigate characteristics code activities distinguish between different classes conditions. The results showcase effectiveness model, emphasizing for accurate inner speech without need explicit class labels. Furthermore, assess significance an ANOVA test, providing statistical evidence their discriminative across To discriminatory dictionaries, compare Multilayer Perceptron (MLP) Convolutional Neural Networks (CNN) classifiers on both raw dictionary outputs. findings demonstrate that accuracy does not decrease when employing approach, showcasing decoding. This research significantly contributes field signal processing networks, paving way advancements innerspeech applications diverse range domains.

Language: Английский

Lobish: Symbolic Language for Interpreting Electroencephalogram Signals in Language Detection Using Channel-Based Transformation and Pattern DOI Creative Commons
Türker Tuncer, Şengül Doğan, İrem Taşçı

et al.

Diagnostics, Journal Year: 2024, Volume and Issue: 14(17), P. 1987 - 1987

Published: Sept. 8, 2024

Electroencephalogram (EEG) signals contain information about the brain’s state as they reflect functioning. However, manual interpretation of EEG is tedious and time-consuming. Therefore, automatic translation models need to be proposed using machine learning methods. In this study, we an innovative method achieve high classification performance with explainable results. We introduce channel-based transformation, a channel pattern (ChannelPat), t algorithm, Lobish (a symbolic language). By were encoded index channels. The ChannelPat feature extractor transition between two channels served histogram-based extractor. An iterative neighborhood component analysis (INCA) selector was employed select most informative features, selected features fed into new ensemble k-nearest neighbor (tkNN) classifier. To evaluate capability language detection model, dataset comprising Arabic Turkish collected. Additionally, introduced obtain outcomes from model. engineering model applied collected dataset, achieving accuracy 98.59%. extracted meaningful cortex brain for detection.

Language: Английский

Citations

9
A State-of-the-Art Review of EEG-Based Imagined Speech Decoding DOI Creative Commons
Diego Lopez-Bernal, David Balderas, Pedro Ponce

et al.

Frontiers in Human Neuroscience, Journal Year: 2022, Volume and Issue: 16

Published: April 26, 2022

Currently, the most used method to measure brain activity under a non-invasive procedure is electroencephalogram (EEG). This because of its high temporal resolution, ease use, and safety. These signals can be Brain Computer Interface (BCI) framework, which implemented provide new communication channel people that are unable speak due motor disabilities or other neurological diseases. Nevertheless, EEG-based BCI systems have presented challenges in real life situations for imagined speech recognition difficulty interpret EEG their low signal-to-noise ratio (SNR). As consequence, order help researcher make wise decision when approaching this problem, we offer review article sums main findings relevant studies on subject since 2009. focuses mainly pre-processing, feature extraction, classification techniques by several authors, as well target vocabulary. Furthermore, propose ideas may useful future work achieve practical application toward decoding.

Language: Английский

Citations

34
A power load forecasting method in port based on VMD-ICSS-hybrid neural network DOI

Kai Ma,

Xuefeng Nie,

Jie Yang

et al.

Applied Energy, Journal Year: 2024, Volume and Issue: 377, P. 124246 - 124246

Published: Sept. 27, 2024

Language: Английский

Citations

8
Variational mode decomposition-based EEG analysis for the classification of disorders of consciousness DOI Creative Commons

Sreelakshmi Raveendran,

Raghavendra Kenchaiah,

Santhos Kumar

et al.

Frontiers in Neuroscience, Journal Year: 2024, Volume and Issue: 18

Published: Feb. 6, 2024

Aberrant alterations in any of the two dimensions consciousness, namely awareness and arousal, can lead to emergence disorders consciousness (DOC). The development DOC may arise from more severe or targeted lesions brain, resulting widespread functional abnormalities. However, when it comes classifying patients with particularly utilizing resting-state electroencephalogram (EEG) signals through machine learning methods, several challenges surface. non-stationarity intricacy EEG data present obstacles understanding neuronal activities achieving precise classification. To address these challenges, this study proposes variational mode decomposition (VMD) before feature extraction along models. By decomposing preprocessed into specified modes using VMD, features such as sample entropy, spectral kurtosis, skewness are extracted across modes. compares performance VMD-based approach frequency band-based also raw-EEG. classification process involves binary between unresponsive wakefulness syndrome (UWS) minimally conscious state (MCS), well multi-class (coma vs. UWS MCS). Kruskal-Wallis test was applied determine statistical significance a p < 0.05 were chosen for second round experiments. Results indicate that outperform other approaches, ensemble bagged tree (EBT) highest accuracy 80.5% (the best literature) 86.7% This underscores potential integrating advanced signal processing techniques improving thereby enhancing patient care facilitating informed treatment decision-making.

Language: Английский

Citations

4
Delaunay Triangulated Simplicial Complex Generation for EEG Signal Classification DOI
Srikireddy Dhanunjay Reddy, Tharun Kumar Reddy

IEEE Sensors Letters, Journal Year: 2024, Volume and Issue: 8(10), P. 1 - 4

Published: April 24, 2024

Language: Английский

Citations

4
Speech imagery decoding using EEG signals and deep learning: A survey DOI
Liying Zhang, Yueying Zhou, Peiliang Gong

et al.

IEEE Transactions on Cognitive and Developmental Systems, Journal Year: 2024, Volume and Issue: 17(1), P. 22 - 39

Published: July 19, 2024

Language: Английский

Citations

4
Adaptive multi-layer empirical Ramanujan decomposition and its application in roller bearing fault diagnosis DOI
Haiyang Pan, Ying Zhang, Jian Cheng

et al.

Measurement, Journal Year: 2023, Volume and Issue: 213, P. 112707 - 112707

Published: March 9, 2023

Language: Английский

Citations

10
A new one-dimensional testosterone pattern-based EEG sentence classification method DOI
Tuğçe Keleş, Arif Metehan Yıldız, Prabal Datta Barua

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2022, Volume and Issue: 119, P. 105722 - 105722

Published: Dec. 21, 2022

Language: Английский

Citations

10
Multiscale Domain Gradient Boosting Models for the Automated Recognition of Imagined Vowels Using Multichannel EEG Signals DOI
Shaswati Dash, Rajesh Kumar Tripathy, Dinesh Kumar Dash

et al.

IEEE Sensors Letters, Journal Year: 2022, Volume and Issue: 6(11), P. 1 - 4

Published: Nov. 1, 2022

This letter proposes the multiscale domain gradient boosting-based approach for automated recognition of imagined vowels using multichannel electroencephalogram (MCEEG) signals. The analysis MCEEG signals is performed multivariate automatic singular spectrum and fast adaptive empirical mode decomposition methods. features such as bubble entropy, energy, slope sample L1-norm are evaluated from modes extreme boosting light machine models employed vowel task //a// versus //e// //i// //o// //u// A publicly available database has been used to test performance proposed approach. results demonstrate that achieved an overall accuracy 51.47%, which higher compared other methods same comprising

Language: Английский

Citations

9
Discriminating Bacterial Infection from Other Causes of Fever Using Body Temperature Entropy Analysis DOI Creative Commons
Borja Vargas, David Cuesta–Frau, Paula González‐López

et al.

Entropy, Journal Year: 2022, Volume and Issue: 24(4), P. 510 - 510

Published: April 5, 2022

Body temperature is usually employed in clinical practice by strict binary thresholding, aiming to classify patients as having fever or not. In the last years, other approaches based on continuous analysis of body time series have emerged. These are not only absolute thresholds but also patterns and temporal dynamics these series, thus providing promising tools for early diagnosis. The present study applies three entropy calculation methods (Slope Entropy, Approximate Sample Entropy) records with bacterial infections causes search possible differences that could be exploited automatic classification. comparative analysis, Slope Entropy proved a stable robust method bring higher sensitivity realm applied this context thermometry. This was able find statistically significant between two classes analyzed all experiments, specificity above 70% most cases.

Language: Английский

Citations

7