A study on EEG differences between active counting and focused breathing tasks for more sensitive detection of...

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Detecting consciousness in patients with disorders of consciousness is challenging using standard EEG analysis.
• Current methods comparing active and passive tasks, or single potential analysis, lack sensitivity and accuracy.

Purpose of the Study:

• To develop a novel consciousness detection paradigm using a multi-stage cognitive task.
• To enhance the sensitivity and accuracy of electroencephalography (EEG) analysis for assessing consciousness levels.

Main Methods:

• Recorded EEG responses from 20 healthy subjects under three distinct task modes.
• Analyzed event-related potentials (ERPs), event-related desynchronization/synchronization (ERD/ERS) phenomena, and employed machine learning for classification.
• Utilized a multi-feature joint analysis approach combining time and time-frequency domain data.

Main Results:

• Significant differences in EEG responses were observed across different task modes in both time and time-frequency domains.
• The breathing task mode showed reduced ERP amplitudes and weakened theta-ERS/alpha-ERD phenomena compared to passive mode.
• The breathing mode demonstrated superior distinguishability from the active mode in machine learning classifications.

Conclusions:

• Multi-feature analysis of EEG responses, particularly with cognitive tasks like attention to breathing, can significantly improve consciousness detection.
• This novel paradigm offers a promising approach for more sensitive and accurate assessment of consciousness in clinical settings.