Related Experiment Video
Updated: May 12, 2025
07:05
A Protocol for the Administration of Real-Time fMRI Neurofeedback Training
Published on: August 24, 2017
10.9K
Enhanced classification of tinnitus patients using EEG microstates and deep learning techniques
Zahra Raeisi1, Abolfazl Sodagartojgi2, Fahimeh Sharafkhani3
1Department of Computer Science, University of Fairleigh Dickinson, Vancouver Campus, Vancouver, Canada.
Scientific Reports
|May 7, 2025
Summary
Electroencephalography (EEG) microstate analysis reveals distinct patterns in beta and gamma frequency bands, offering a potential objective diagnostic tool for tinnitus. These findings highlight EEG microstate dynamics as reliable markers for differentiating tinnitus patients and guiding personalized therapies.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Tinnitus, a prevalent auditory disorder, lacks objective diagnostic markers.
- Understanding the neural underpinnings of tinnitus is crucial for developing effective treatments.
Purpose of the Study:
- To classify tinnitus using electroencephalography (EEG) signals.
- To analyze EEG microstate dynamics in different frequency bands.
- To explore machine learning approaches for tinnitus detection.
Main Methods:
- EEG data from healthy individuals and tinnitus patients were analyzed.
- Signals were decomposed into five frequency bands (delta, theta, alpha, beta, gamma).
- Microstate features (Duration, Occurrence, Mean Global Field Power, Coverage) were extracted and analyzed using various machine learning models (SVM, DNN, etc.).
Main Results:
- Significant alterations in beta and gamma band microstates were observed in tinnitus patients.
- Microstate A duration increased, while microstate B duration decreased in tinnitus patients.
- Deep Neural Networks achieved 100% accuracy in classifying tinnitus using gamma band EEG data.
Conclusions:
- EEG microstate dynamics in beta and gamma bands are reliable markers for distinguishing tinnitus patients.
- Microstate analysis shows promise as an objective diagnostic tool for tinnitus.
- Findings can inform personalized neuromodulation therapies for tinnitus.
