Wearable Optical Sweat Analytics

Microfluidics and RGB colorimetry for biomarker detection

Overview

Personalized wearable system for non-invasive sweat biomarker monitoring, combining Tesla valve microfluidics, RGB colorimetry, and AI analysis (~90% accuracy). Enables real-time detection of pH, glucose, lactate, cortisol, and electrolytes for health assessment.

Problem Statement

Traditional blood/urine tests are invasive, require 24-72 hour lab turnaround, and lack continuous monitoring capability. No commercial wearable exists for real-time multi-biomarker sweat analysis despite strong clinical demand (cystic fibrosis, diabetes, stress monitoring).

Methodology

Hardware Design

  • Tesla Valve Microfluidics: Passive one-way flow control without pumps, reduces contamination
  • RGB Photoelectric Colorimetry: 100× cheaper than spectrometers, captures biomarker-reagent color changes
  • Wireless Transmission: Bluetooth/Wi-Fi streaming to smartphone/PC

AI Analysis Pipeline

  1. RGB signal acquisition from test strips
  2. Color space transformation (HSV/Lab normalization)
  3. Neural network inference (3-layer fully connected, dropout regularization)
  4. Personalized baseline calibration for user-specific accuracy

Biomarkers Detected: pH, glucose, lactate, cortisol, electrolytes (Na⁺, K⁺, Cl⁻)

Results

Performance:

  • Detection accuracy: ~90% agreement with laboratory methods
  • Response time: <5 minutes (sweat contact to readout)
  • Detection ranges: pH 4-8, glucose 0-200 mg/dL, lactate 5-50 mM
  • Reproducibility: Inter-device CV <8%, intra-device CV <5%
  • Wearability: 20+ hours continuous operation tested

Applications

  • Cystic fibrosis screening (elevated sweat chloride)
  • Diabetes monitoring (non-invasive glucose tracking)
  • Athletic performance optimization (hydration, lactate)
  • Stress assessment (cortisol monitoring)
  • Stroke rehabilitation (autonomic function mapping)

Achievements & Recognition

Awards

  • 3rd Prize - 10th National Student Optoelectronic Design Competition (2022)
  • NCHU ‘Three Small Projects’ Award
  • National Innovation Project - Completed with prototype

Publications

Talanta (Q1 journal, IF ~6.0) - peer-reviewed publication on sweat biomarker detection

Key Metrics

  • ~90% detection accuracy
  • ~$50 device cost (vs. >$5,000 lab equipment)
  • ~$0.50 test strip cost (vs. $20-100 lab analysis)

Team & Collaboration

Principal Investigators: A/Prof. Shi Huanhuan (石环环), Prof. Zhang Weiwei (张巍巍)
Institution: Department of Biomedical Engineering, Nanchang Hangkong University
Core Contributors: Zhou Yanuo (system integration, AI development), Xie Zhihao (circuit design), Cao Yu (data analysis, Talanta co-author)
Collaborators: Nie Daosheng, Xin Jiliang, Yan Yuwei, Liu Yujie
Funding: Outstanding Young Scientist Project (20192BCB23011)

Timeline

Duration: September 2020 - July 2022 (22 months)
Competition: October 2022