# Cognitive arousal-based measures quantify insights from self-ratings in response to sensory stimuli

## TL;DR

This paper shows that measuring skin conductance can help quantify how mentally stimulated people feel in response to different types of sensory input.

## Contribution

The study introduces a novel framework using skin conductance and Bayesian filtering to better align self-reported arousal with physiological data.

## Key findings

- Cognitive arousal measures from skin conductance align better with self-ratings than inferred autonomic nervous system events.
- Continuous skin conductance measurements effectively convey self-perceived arousal in response to sensory stimuli.
- The framework could support non-invasive mental health interventions through interactive sensory environments.

## Abstract

Reliable quantification of patients’ cognitive arousal is a challenging problem and a pertinent clinical need in various mental health applications. Recently, skin conductance-based cognitive state estimation has shown promise in inferring the cognitive arousal of individuals caused by autonomic nervous system (ANS) activation. Here, we use a physiological model of ANS-stimulated skin conductance modulations and Bayesian filtering to analyze changes in cognitive arousal induced by auditory, visual, and haptic stimuli. Our findings indicate that cognitive arousal-based measures are in better agreement with self-ratings-based metrics than inferred autonomic nervous system activation events in response to sensory stimuli. These insights on cognitive arousal increase our understanding of psychophysiology and may help diagnose, track, and treat symptoms of mental health disorders in the future by providing clinicians with a framework to estimate and modulate arousal levels in an interactive sensory stimulation environment.

In this work, we find that measures derived from continuous skin conductance measurements effectively convey insights on self-perceived arousal in response to visual, auditory and haptic stimuli. The conclusions from this analysis will contribute towards the quantification, tracking and eventually modulation of cognitive states during human-computer interactions (e.g., workplace or virtual reality settings) for entertainment as well as for non-invasive, non-pharmacological closed-loop mental health interventions.