YOLOv5-aided paper-based microfluidic intelligent sensing platform for multiplex sweat biomarker analysis

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Point-of-Care Diagnostics

Background:

• Sweat analysis offers non-invasive health monitoring potential.
• Conventional sweat analysis methods face accuracy and complexity challenges.
• Existing platforms struggle with scalability and resource limitations.

Purpose of the Study:

• To develop an intelligent, paper-based microfluidic sensing platform for sweat biomarker detection.
• To overcome limitations of conventional methods in accuracy, cost, and complexity.
• To enable portable and reproducible point-of-care diagnostics.

Main Methods:

• Integration of a paper-based microfluidic chip with smartphone imaging.
• Utilization of a deep learning framework (YOLOv5) for image analysis.
• Colorimetric detection of multiplex biomarkers in sweat samples.

Main Results:

• The YOLOv5 deep learning model achieved a mean average precision of 99.5%.
• Successful validation of the platform for detecting iron ions, chloride ions, and glucose in sweat.
• Demonstration of a cost-effective, portable, and reproducible sensing system.

Conclusions:

• The developed platform offers a promising solution for non-invasive sweat analysis.
• This intelligent sensing system addresses key challenges in current point-of-care diagnostics.
• The technology facilitates accessible health monitoring, particularly in resource-limited settings.