Lensless imaging for droplet identification towards visual feedback-based pressure controlled droplet microfluidic platforms

Abstract

The usability and robustness of droplet microfluidic devices constitute significant challenges to the widespread adoption of droplet microfluidics despite the field’s immense promise. Active droplet manipulation via visual feedback-based pressure driven control has addressed some of the concerns; however, the use of large and costly optical microscopes for acquiring visual feedback limits such systems for many applications. We present a compact and cost-effective (200 [USD]) droplet sensing system that leverages lensless imaging techniques using a single LED and a CMOS sensor. The imaging system can detect nearly all droplet interfaces (0.05 < miss rate, > 0.95 positive predictive value). Its accuracy (mean errors less than 10 [µm]), Field of View (7.9 [mm2]), resolution (2.2 [µm]) and sensing rate (40.1 [Hz]) are similar to existing optical microscope-based droplet identification systems. The lensless imaging system has acceptable performance while having a significantly smaller footprint and cost, which presents tremendous potential to miniaturize and modularize droplet microfluidic systems paving the road for widespread adoption.