Mental State Assessment and Validation Using Personalized Physiological Biometrics.

Area of Science:

• Human-Computer Interaction
• Neuroscience
• Biomedical Engineering

Background:

• Current mental state monitoring relies on generic population ranges, lacking individual task performance context.
• Physiological biometrics need validation for individual-specific mental state assessment in real-world applications.

Purpose of the Study:

• To develop and validate personalized biometric models for mental fatigue, stress, and attention.
• To improve human-machine interface efficacy through individualized mental state monitoring.
• To establish a unified pipeline for mental state evaluation using EEG and ECG sensors.

Main Methods:

• Curated intake experiments to generate personalized biometric models.
• Introduction of new features to enhance existing approaches.
• Multi-level validation using k-fold cross-validation and regression testing.
• Permutation tests to correlate biometrics with task performance (accuracy, reaction time).

Main Results:

• Personalized models accurately indicate individual mental states (fatigue, stress, attention).
• Proposed validation schemes effectively assess biometric model efficacy for individual subjects.
• Demonstrated correlations between personalized biometrics and dynamic task performance fluctuations.

Conclusions:

• The developed pipeline offers a comprehensive mental state evaluation from EEG and ECG.
• Personalized biometrics provide a path for judicious integration into augmented human-machine interfaces.
• This approach enhances task performance by tailoring monitoring to individual cognitive states.