Flow measurement in microfluidic chips through optical trapping and deep learning

Mechanobiology is an emerging multidisciplinary field that involves the study of the mechanisms by which biological organisms sense and respond to mechanical stimuli. In recent years, this field has seen significant advancements through the application of microfluidics and optical manipulation. Microfluidics enables precise control of channel content and flow with great precision, while optical trapping allow for manipulation of microscopic objects. Combining these disciplines offers new opportunities for studying biological phenomena with reduced scale experiments. However, challenges remain in coordinating microfluidics with optical manipulation within confined spaces, in particular when working with biological entities. To address these limitations, an integrated approach is proposed, using 3D optical manipulation setup, microfluidics and deep learning image recognition to estimate forces experienced by optically trapped objects. By analyzing the bead’s displacement within the flow, forces are quantified using a deep learning algorithm. Experimental results demonstrate force variations based on the position within the chip, revealing the potential for improved understanding of biological mechanisms through characterization of local forces. This study facilitates the establishment of optofluidics manipulations, paving the way for future explorations in mechanobiology.