Deep-Learning-Assisted Microfluidic Immunoassay via Smartphone-Based Imaging Transcoding System for On-Site and...

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Artificial Intelligence

Background:

• Clinical diagnostics and food safety demand point-of-care testing (POCT) with multiplexing, high sensitivity, affordability, and ease of use.
• Current POCT methods often lack the multiplexed capability and sensitivity required for comprehensive analysis.

Purpose of the Study:

• To develop a deep-learning-assisted microfluidic immunoassay platform for advanced POCT applications.
• To create a user-friendly, smart toolkit for rapid, multiplexed detection of various analytes.

Main Methods:

• Utilized a microfluidic chip with polystyrene microsphere-based encoding for multiprobe generation.
• Integrated a smartphone-based imaging transcoding system with artificial intelligence (AI) for decoding and quantification.
• Implemented a novel "encoding-decoding" strategy within the microfluidic chip, eliminating the need for immunomagnetic separation.

Main Results:

• Successfully detected multiple respiratory viruses and antibiotics in real samples with ultrahigh sensitivity.
• Achieved rapid detection results within 30 minutes.
• Demonstrated accurate classification, counting, and concentration calculations via AI-assisted decoding.

Conclusions:

• The developed platform offers a smart, universal toolkit for next-generation POCT.
• The AI-assisted microfluidic immunoassay shows significant potential for clinical diagnostics and food safety.
• The system's user-friendly design and multiplexed capability address critical needs in rapid diagnostics.