Arousal Detection in Elderly People from Electrodermal Activity Using Musical Stimuli

Area of Science:

• Psychophysiology
• Gerontology
• Biomedical Engineering

Background:

• Emotion detection is crucial for health and well-being.
• Electrodermal activity (EDA) monitoring offers a non-invasive method for assessing physiological arousal.
• Older adults are a population group that can benefit from interventions to improve mood and reduce anxiety/depression.

Purpose of the Study:

• To identify arousal levels in older adults using electrodermal activity (EDA) monitoring.
• To explore the potential of music-based therapies for mood enhancement in the elderly.
• To investigate the relationship between music genre familiarity and emotional induction.

Main Methods:

• Commercial device used for electrodermal activity (EDA) signal acquisition.
• Signal deconvolution methods applied to EDA data.
• Statistical analysis of signal features (temporal, morphological, frequency).
• Machine learning classifiers (SVM, KNN) used to correlate EDA-based arousal with subjective reports.

Main Results:

• Flamenco and Spanish Folklore music genres showed the highest number of statistically significant EDA parameters.
• Support Vector Machines achieved 87% accuracy for Flamenco and 83.1% for Spanish Folklore in arousal detection.
• K-Nearest Neighbors demonstrated 81.4% and 81.5% accuracy for Flamenco and Spanish Folklore, respectively.

Conclusions:

• Familiar musical genres can effectively induce emotional responses and modulate arousal levels in older adults.
• EDA monitoring combined with machine learning shows promise for objective emotion assessment in gerontology.
• These findings support the development of personalized music therapies to improve emotional well-being in the elderly.