An automated optimal engagement and attention detection system using electrocardiogram

Area of Science:

• Neuroscience and Biomedical Engineering
• Physiological Signal Analysis
• Cognitive Science

Background:

• Cognitive attention is crucial for task performance but difficult to monitor objectively.
• Electrocardiograph (ECG) and Electroencephalograph (EEG) are key physiological signals.
• Current methods for attention monitoring can be invasive or lack precision.

Purpose of the Study:

• To develop a monitoring system using ECG to analyze and predict cognitive attention.
• To investigate the correlation between attention levels and cardiac rhythm variations.
• To compare ECG-based attention detection with EEG as a benchmark.

Main Methods:

• Advanced signal processing techniques, including Stockwell-transform for ECG and Discrete Wavelet Transform (DWT) for EEG.
• Extraction of informative features from both ECG and EEG signals.
• Application of machine learning algorithms for classification of attentive vs. inattentive states.

Main Results:

• Derived significant features from ECG and EEG signals correlating with cognitive attention.
• Developed machine learning models capable of differentiating attention states.
• Demonstrated that ECG-based attention detection is comparable in accuracy to EEG.

Conclusions:

• ECG analysis provides a viable and comparable alternative to EEG for cognitive attention monitoring.
• The proposed system offers a non-invasive and potentially more accessible method for attention assessment.
• This research opens avenues for real-time cognitive state monitoring in various applications.