A Novel Filter for Tracking Real-World Cognitive Stress using Multi-Time-Scale Point Process Observations.

Area of Science:

• Physiological monitoring
• Neuroscience
• Signal processing

Background:

• Wearable technology enables continuous monitoring of physiological signals.
• Understanding the link between neurocognitive stress and physiological changes is crucial.
• Existing methods may not fully capture dynamic stress responses.

Purpose of the Study:

• To develop a state-space model for tracking neurocognitive stress.
• To utilize skin conductance and electrocardiography measurements for stress assessment.
• To establish a novel framework for state estimation using multiple point process observations.

Main Methods:

• Modeling individual skin conductance responses (SCRs) and heartbeats as stress-related point processes.
• Linking SCRs and heartbeats to sympathetic nervous system activation.
• Deriving Kalman-like filter equations for stress tracking.
• Employing expectation-maximization and maximum likelihood estimation for parameter recovery.

Main Results:

• Preliminary results indicate higher stress levels during unfamiliar tasks.
• Stress levels were observed to decrease with increasing task familiarity.
• External stressors can influence stress levels even with task familiarity.

Conclusions:

• The proposed state-space approach is feasible for real-world stress tracking using wearable sensors.
• The method provides a novel state estimation framework for multiple point process data.
• This work advances the understanding of physiological correlates of neurocognitive stress.