Neural Efficiency and Attentional Instability in Gaming Disorder: A Task-Based Occipital EEG and Machine Learning...

Area of Science:

• Neuroscience
• Behavioral Addiction Research
• Computational Psychiatry

Background:

• Gaming Disorder (GD) is a recognized behavioral addiction with known resting-state impairments.
• Neurophysiological dynamics during active gameplay in GD remain largely unexplored.
• Understanding task-based neural activity is crucial for developing objective diagnostic tools.

Purpose of the Study:

• To identify task-based occipital electroencephalography (EEG) biomarkers specific to Gaming Disorder.
• To assess the diagnostic utility of these EEG biomarkers for GD detection.
• To explore the neurophysiological differences between individuals with and without GD during active gaming.

Main Methods:

• Collected occipital EEG (O1/O2) data from 30 participants (15 with GD, 15 controls) during mobile gaming.
• Extracted spectral, temporal, and nonlinear complexity features from EEG data.
• Utilized Random Forest for feature relevance ranking and Leave-One-Subject-Out cross-validation with five machine learning models for classification.

Main Results:

• The Gaming Disorder group exhibited 'spectral slowing' during gameplay: increased Delta/Theta power and decreased Beta activity.
• Beta variability emerged as a key biomarker, indicating altered attentional stability in GD.
• A Decision Tree classifier achieved 80.0% accuracy in distinguishing between GD and control groups.

Conclusions:

• Distinct neurophysiological patterns, including increased low-frequency power, are associated with Gaming Disorder during gameplay.
• Occipital EEG biomarkers, particularly Beta variability, show promise as objective screening metrics for GD.
• These findings suggest potential 'neural efficiency' or automatized processing in GD despite active task engagement.