Identifying Fragmented Reading and Evaluating Its Influence on Cognition Based on Single Trial Electroencephalogram

Area of Science:

• Cognitive Neuroscience
• Neuroscience
• Machine Learning Applications

Background:

• Daily social media use may foster fragmented reading (FR) habits.
• The cognitive effects and underlying electroencephalogram (EEG) alterations of FR remain largely unknown.

Purpose of the Study:

• To identify FR habits using single-trial EEG spectral features and machine learning (ML).
• To investigate potential cognitive impairments induced by FR.

Main Methods:

• Subjects were categorized into FR and noFR groups based on daily FR time.
• 64-channel EEG data were collected during a Continuous Performance Task (CPT).
• Relative power (RP) in six frequency bands was extracted from EEG epochs for ML classification (SVM, KNN, Naive Bayes).

Main Results:

• ML models achieved high accuracy (up to 96.52% with SVM) in identifying FR.
• Alpha and gamma band features were most significant for classification.
• Distinct EEG patterns (alpha and gamma RP) were observed between FR and noFR groups, indicating potential cognitive differences.

Conclusions:

• Fragmented reading can be identified via ML analysis of single-trial EEG during CPT.
• Alpha and gamma band alterations in EEG may signify attention and working memory impairments associated with FR.
• FR's potential for cognitive impairment warrants further investigation.