PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
March 27, 2023MathematicsOpen Access

Automated EEG Pathology Detection Based on Significant Feature Extraction and Selection

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Existing machine learning-based EEG binary classification methods largely focus on extracting EEG-related features, which may lead to poor performance by overlooking potentially redundant information.

Does a Kruskal-Wallis test-based machine learning framework improve automated EEG pathology detection compared to existing methods?

Population

EEG data for pathology detection

Comparison

Proposed Kruskal-Wallis test-based framework vs other competing techniques

Authors

YZYunning ZhongHWHong-yu WeiLCLifei Chen

Discussion

Loading...

Member takes

Overview

May aid automated EEG detection in practice; leaves open prospective clinical validation.

Structured PICO

Does a Kruskal-Wallis test-based machine learning framework improve automated EEG pathology detection compared to existing methods?

P
Population
EEG data for pathology detection (normal vs abnormal classes)
I
Intervention
Kruskal-Wallis (KW) test-based framework using wavelet packet decomposition, piecewise aggregation approximation, and ensemble learning classifiers (random forest, CatBoost, light gradient boosting machine)
C
Comparator
Other competing techniques on the same dataset
O
Outcome
Classification performance (accuracy, F1-score, G-mean)surrogate

A novel Kruskal-Wallis test-based machine learning framework demonstrates high accuracy (89.13%) in automated EEG pathology detection.

Cite This Study

Zhong et al. (2023) studied this question.

synapsesocial.com/papers/6a10a3e301be78fe816126f7https://doi.org/10.3390/math11071619
View Full Paper
Ask AI
Bookmark
Share