Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

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

Minh Ngoc Nguyen, Khai Le-Duc, Tan-Hanh Pham, Trang Nguyen, Quang Minh Luu, Ba Kien Tran, Truong-Son Hy, Viktor Dremin, Sergei Sokolovsky, Edik Rafailov

TL;DR

This study addresses the need for scalable, non-invasive mental health assessment by introducing a wearable device that combines Laser Doppler Flowmetry and Fluorescence Spectroscopy to collect microcirculation and metabolic signals, paired with the DAS-21 questionnaire. The authors publish the largest LDF/FS dataset to date (132 participants across 19 countries) and evaluate multiple machine learning models, with LightGBM delivering strong binary and multi-class stress predictions, plus robust explainability via SHAP. Across various feature sets and validation schemes, top physiological factors such as BMI, age, and heart rate emerge as key predictors, and the top-10 feature subset often yields improved performance and efficiency. The work demonstrates practical potential for real-time mental health monitoring and personalized interventions, supported by open-source code and data to advance future research in wearable-based psychiatric assessment.

Abstract

In this study, 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. Besides, we present the corresponding dataset to predict mental health, e.g. depression, anxiety, and stress levels via the DAS-21 questionnaire. To our best knowledge, this is the world's largest and the most generalized dataset ever collected for both LDF and FS studies. The device captures cutaneous blood microcirculation parameters, and wavelet analysis of the LDF signal extracts key rhythmic oscillations. The dataset, collected from 132 volunteers aged 18-94 from 19 countries, explores relationships between physiological features, demographics, lifestyle habits, and health conditions. We employed a variety of machine learning methods to classify stress detection, in which 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. In addition, we also incorporated Explainable Artificial Intelligence (XAI) techniques into our analysis to investigate deeper insights into the model's predictions. Our results suggest that females, younger individuals and those with a higher Body Mass Index (BMI) or heart rate have a greater likelihood of experiencing mental health conditions like stress and anxiety. All related code and data are published online: https://github.com/leduckhai/Wearable_LDF-FS.

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

TL;DR

This study addresses the need for scalable, non-invasive mental health assessment by introducing a wearable device that combines Laser Doppler Flowmetry and Fluorescence Spectroscopy to collect microcirculation and metabolic signals, paired with the DAS-21 questionnaire. The authors publish the largest LDF/FS dataset to date (132 participants across 19 countries) and evaluate multiple machine learning models, with LightGBM delivering strong binary and multi-class stress predictions, plus robust explainability via SHAP. Across various feature sets and validation schemes, top physiological factors such as BMI, age, and heart rate emerge as key predictors, and the top-10 feature subset often yields improved performance and efficiency. The work demonstrates practical potential for real-time mental health monitoring and personalized interventions, supported by open-source code and data to advance future research in wearable-based psychiatric assessment.

Abstract

In this study, 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. Besides, we present the corresponding dataset to predict mental health, e.g. depression, anxiety, and stress levels via the DAS-21 questionnaire. To our best knowledge, this is the world's largest and the most generalized dataset ever collected for both LDF and FS studies. The device captures cutaneous blood microcirculation parameters, and wavelet analysis of the LDF signal extracts key rhythmic oscillations. The dataset, collected from 132 volunteers aged 18-94 from 19 countries, explores relationships between physiological features, demographics, lifestyle habits, and health conditions. We employed a variety of machine learning methods to classify stress detection, in which 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. In addition, we also incorporated Explainable Artificial Intelligence (XAI) techniques into our analysis to investigate deeper insights into the model's predictions. Our results suggest that females, younger individuals and those with a higher Body Mass Index (BMI) or heart rate have a greater likelihood of experiencing mental health conditions like stress and anxiety. All related code and data are published online: https://github.com/leduckhai/Wearable_LDF-FS.

Paper Structure

This paper contains 54 sections, 48 figures, 21 tables.

Table of Contents

  1. Introduction
  2. Motivation
  3. Literature Review
  4. Contribution
  5. Study Design and Dataset Description
  6. Machine Learning and Explainable Artificial Intelligence
  7. Machine Learning Models for DAS Prediction
  8. Explainable AI
  9. Experimental Results
  10. All Features with 80:20 Split
  11. Binary Classification
  12. Multi-class Classification
  13. Feature Importance
  14. All Features with Cross-Validation
  15. Binary Classification with LOPO
  16. ...and 39 more sections

Figures (48)

  • Figure 1: Data collection workflow.
  • Figure 2: Data instances collected using the wearable devices: (a) stress instance, (b) well-being instance.
  • Figure 3: Distribution of stress levels, anxiety level, and depression level.
  • Figure 4: Explanation of the reasoning behind individual class predictions using SHAP values.
  • Figure 5: Data Analysis
  • ...and 43 more figures