A Wearable Device Dataset for Mental Health Assessment Using Laser Doppler Flowmetry and Fluorescence Spectroscopy Sensors

This paper introduces a novel wearable device dataset and machine learning pipeline for predicting mental health conditions (depression, anxiety, stress) using Laser Doppler Flowmetry (LDF) and Fluorescence Spectroscopy (FS) sensors.

Abstract

We introduce a novel method to predict mental health by building machine learning models for a non-invasive wearable device equipped with Laser Doppler Flowmetry (LDF) and Fluorescence Spectroscopy (FS) sensors. The corresponding dataset, collected from 132 volunteers aged 18–94 from 19 countries, is the world’s largest and most generalized dataset for both LDF and FS studies. Wavelet analysis of the LDF signal extracts key rhythmic oscillations. LightGBM is identified as the most effective model for stress detection, achieving a ROC AUC of 0.7168 and a PR AUC of 0.8852. Explainable AI (XAI) techniques reveal that females, younger individuals, and those with higher BMI or heart rate have a greater likelihood of experiencing mental health conditions.

Key Contributions

• World’s largest LDF and FS dataset for mental health assessment (132 volunteers, 19 countries)
• Machine learning pipeline for depression, anxiety, and stress classification
• XAI-driven insights into physiological and demographic predictors of mental health conditions

Recommended citation: M. N. Nguyen, K. Le-Duc, T.-H. Pham, T. N. Nguyen et al., "A Wearable Device Dataset for Mental Health Assessment Using Laser Doppler Flowmetry and Fluorescence Spectroscopy Sensors," *Communications Medicine*, 2026, doi: 10.1038/s43856-026-01766-5.