Related Experiment Video
Updated: Jan 9, 2026
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
1.7K
Music-induced physiological markers for detecting Alzheimer's disease using machine learning
Rodrigo Lima1,2,3, Gonçalo Barradas4, Sergi Bermúdez I Badia1,2,3
1Faculdade de Ciências Exatas e da Engenharia, Universidade da Madeira, Funchal, Portugal.
Frontiers in Aging Neuroscience
|December 10, 2025
Summary
Music-evoked physiological responses show promise as non-invasive biomarkers for Alzheimer's disease (AD). Machine learning models detected subtle differences in these responses, aiding in AD detection and severity assessment.
Area of Science:
- Neuroscience
- Biomarkers
- Machine Learning
Background:
- Alzheimer's disease (AD) causes cognitive and emotional decline, necessitating new non-invasive biomarkers for early detection and intervention.
- Current diagnostic methods for AD can be invasive or lack sensitivity for early-stage detection.
Purpose of the Study:
- To investigate music-evoked physiological responses as potential non-invasive biomarkers for Alzheimer's disease (AD).
- To evaluate the translational value of these physiological responses using machine learning (ML) for AD detection and severity assessment.
Main Methods:
- Recorded electrodermal activity and facial electromyography in 36 AD patients listening to music.
- Trained machine learning models (Random Forest, Naïve Bayes) on physiological signals to classify AD presence, severity, and residual emotional responses.
Main Results:
- Physiological reactivity to music decreased with AD progression.
- Machine learning models achieved 70.5% accuracy in distinguishing AD patients from controls and 65.6% accuracy in predicting AD severity.
- Subtle music-evoked physiological differences were detectable in AD patients using ML.
Conclusions:
- Music-evoked physiological signals reflect neural circuit disruption in AD and can serve as complementary biomarkers.
- Combining physiological measures with ML offers a non-invasive approach for early AD detection, monitoring, and developing stage-specific interventions.
