Deep Learning Enabled Universal Multiplexed Fluorescence Detection for Point-of-Care Applications

Area of Science:

• Analytical Chemistry
• Biotechnology
• Machine Learning Applications

Background:

• Multiplexed fluorescence sensing is crucial but hindered by bulky, expensive optical systems.
• Existing methods require complex spectrum separation, limiting portability and cost-effectiveness.

Purpose of the Study:

• To develop a compact, lens-free, and affordable fluorescence sensing system for scalable multiplexed detection.
• To enable multiplexed detection without optical adjustments using machine learning.

Main Methods:

• Utilized low-cost optical components and a pretrained machine learning (ML) model.
• Developed a lens-free setup for fluorescence sensing.
• Demonstrated scalability through ML model updates without hardware changes.

Main Results:

• Successfully demonstrated multiplexed fluorescence sensing without optical adjustments.
• Achieved simultaneous detection of three respiratory viruses using a multiplexed Loop-Mediated Isothermal Amplification (LAMP) assay.
• Validated the system's potential for real-world point-of-care applications.

Conclusions:

• The ML-enabled, compact fluorescence sensing system offers a cost-effective and scalable solution for multiplexed detection.
• This technology has significant potential for widespread adoption in clinical diagnostics and field settings.
• The approach bypasses the need for expensive optical components, democratizing advanced sensing capabilities.