Related Experiment Video
Updated: Jun 15, 2026
08:23
A Multimodal Imaging- and Stimulation-based Method of Evaluating Connectivity-related Brain Excitability in Patients with Epilepsy
Published on: November 13, 2016
11.3K
Microstate Analysis of Resting-State EEG Signals for Classifying Tinnitus from Healthy Subjects
Faezeh Mousazadeh Sarghein1, Nasser Samadzadehaghdam1, Faegheh Golabi1
1Department of Biomedical Engineering, Faculty of Advanced Medical Sciences, Tabriz University of Medical Sciences, Tabriz, Iran.
Clinical EEG and Neuroscience
|June 30, 2025
Summary
Electroencephalography (EEG) microstate analysis effectively differentiates tinnitus patients from healthy individuals. Machine learning models like SVM achieved over 96% accuracy, showing potential for improved tinnitus diagnosis.
Area of Science:
- Neuroscience
- Computational Psychiatry
- Biomedical Engineering
Background:
- Electroencephalography (EEG) offers high temporal resolution for brain electrophysiology.
- Microstate analysis segments EEG into quasi-stable states, revealing brain network activity.
- This approach is relevant for understanding neuropsychiatric disorders, including tinnitus.
Purpose of the Study:
- To differentiate tinnitus patients from healthy controls using EEG microstate features.
- To evaluate the efficacy of machine learning algorithms in classifying tinnitus based on these features.
Main Methods:
- Investigated EEG microstate differences between 16 healthy controls and 10 tinnitus patients.
- Analyzed four microstates using Multivariate Analysis of Variance (MANOVA).
- Employed machine learning algorithms (SVM, KNN) for classification based on microstate features.
Main Results:
- Significant differences in microstate A duration (auditory processing) and microstate B coverage/occurrence (visual networks) were found.
- Support Vector Machine (SVM) achieved 96.44% accuracy, 97.64% precision, and 97.24% F1-score.
- K-Nearest Neighbors (KNN) demonstrated a maximum recall of 97.24%.
Conclusions:
- EEG microstate analysis, with time-related features, shows promise for tinnitus diagnosis and classification.
- Machine learning models (SVM, KNN) accurately identify tinnitus-associated brain patterns.
- This highlights the clinical utility of EEG in managing neurological disorders.
