Multispectral sensor fusion in SmartWatch for skin hydration and body sweat loss
Activity counts measure attention, not model quality. Missing counts are unknown. Publication status follows the linked source; a listed venue alone does not confirm peer review.
This Scientific Reports paper describes a smartwatch PPG prototype that adds two infrared light sources, at 970 nm and 1450 nm, to the usual visible-light sensor so it can track skin water content and sweat loss continuously. The authors modelled how light at each wavelength travels through dry and wet skin using Monte Carlo simulations, and checked those simulations against lab spectrometer and ultrasound measurements from four volunteers. They then analysed a user study of 19 adults (aged 21 to 52) who completed 103 indoor 5 km treadmill runs wearing the prototype, using machine learning (LightGBM) on the multi-wavelength PPG signals to track hydration changes. The work is a sensing-technology and modelling pipeline rather than a validated clinical hydration measure.
- Venue
- Scientific Reports
- Published
- 2023
- arXiv
- Not reported
- Signals
- Not reported
- Form factor
- Not reported
- Subjects
- 19 adults (male and female, aged 21–52) in 103 indoor 5 km running trials
- Code
- Not reported
- Added
- Oct 13, 2026
Abstract
Post-pandemic health operations have become a near-term reality, discussions around wearables are on the rise. How do wearable health solutions effectively deploy and use this opportunity to fill the gap between wellness and healthcare? In this paper, we will talk about wearable healthcare diagnosis, with a particular focus on monitoring skin hydration using optical multi-wavelength sensor fusion. Continuous monitoring of human skin hydration is a task of paramount importance for maintaining water loss dynamics for fitness lovers as well as for skin beauty, integrity and the health of the entire body. Preserving the appropriate levels of hydration ensures consistency of weight, positively affects psychological state, and proven to result in a decrease in blood pressure as well as the levels of “bad” cholesterol while slowing down the aging processes. Traditional methods for determining the state of water content in the skin do not allow continuous and non-invasive monitoring, which is required for variety of consumer, clinical and cosmetic applications. We present novel sensing technology and a pipeline for capturing, modeling and analysis of the skin hydration phenomena and associated changes therein. By expanding sensing capabilities built into the SmartWatch sensor and combining them with advanced modeling and Machine Learning (ML) algorithms, we identified several important characteristics of photoplethysmography (PPG) signal and spectral sensitivity corresponding to dynamics of skin water content. In a hardware aspect, we newly propose the expansion of SmartWatch capabilities with InfraRed light sources equipped with wavelengths of 970 nm and 1450 nm. Evaluation of the accuracy and characteristics of PPG sensors has been performed with biomedical optics-based simulation framework using Monte Carlo simulations. We performed rigorous validation of the developed technology using experimental and clinical studies. The developed pipeline serves as a tool in the ongoing studies of the next generation of optical sensing technology.
Verbatim fromthe publisher's DOI record· Publisher licence: https://creativecommons.org/licenses/by/4.0
How it works
- Device
- wrist-worn smartwatch prototype with added 970 nm and 1450 nm infrared LEDs
- Wavelengths recorded
- 535, 940, 970 and 1450 nm PPG, sampled at 25 Hz, plus accelerometer and gyroscope
- Cohort
- 19 adults (male and female, aged 21–52) in 103 indoor 5 km running trials
- Conditions
- treadmill rooms at 10–34 °C ambient temperature and 25–60% relative humidity, with body weighing on precision scales
- Methods
- Monte Carlo light-transport simulation, validated with spectral and ultrasound skin measurements from 4 volunteers; LightGBM models on the PPG features
Cite
@article{samsungmultispectralsmar2023,
title = {{Multispectral sensor fusion in SmartWatch for skin hydration and body sweat loss}},
author = {Elena Volkova and Alexey Perchik and Konstantin Pavlov and Evgenii Nikolaev and Alexey Ayuev and Jaehyuck Park and Namseok Chang and Wonseok Lee and Justin Younghyun Kim and Alexander Doronin and Maksim Vilenskii},
year = {2023},
journal = {Scientific Reports},
doi = {10.1038/s41598-023-40339-7}
}History
Added to the ChatPPG research hub