Image-based fuzzy logic control for pressure-driven droplet microfluidics as autosampler for multimodal imaging...

Area of Science:

• Microfluidics
• Control Systems
• Biotechnology

Background:

• Controlling droplet position in microfluidic devices is crucial for applications like high-throughput screening.
• Existing methods may lack precision or adaptability to varying chip designs and droplet characteristics.

Purpose of the Study:

• To develop and validate a customizable, image-based fuzzy logic control (FLC) method for precise droplet positioning in pressure-driven microfluidics.
• To assess the system's performance across diverse microfluidic chip geometries and droplet compositions.

Main Methods:

• Implementation of two parallel Multiple Input Single Output (MISO) fuzzy logic controllers (FLCs) for real-time image-based droplet position correction.
• Utilizing pre-determined, constant pressure values to maintain droplet stability within the region of interest (ROI).
• Testing the FLC system on various microfluidic chips (different materials, channel layouts, cross-sections) and pressure controllers.

Main Results:

• Achieved 95.1% droplet retention within the ROI with an average displacement of only 2.5 μm over 10-second intervals.
• Demonstrated successful control across a range of droplet volumes (picoliters to microliters) and compositions (deionized water, S. cerevisiae solutions).
• The system exhibited an average processing time of 12.5 seconds per droplet.

Conclusions:

• The developed image-based FLC method offers a highly customizable and effective solution for precise droplet manipulation in microfluidics.
• This technique is well-suited for analyzing hundreds of pre-sorted droplets from high-throughput screening experiments, improving data reliability and throughput.
• The system's robustness across different experimental setups highlights its broad applicability in microfluidic research and diagnostics.