Deep learning-assisted concentration gradient generation for the study of 3D cell cultures in hydrogel beads of varying stiffness

Frontiers in Bioengineering and Biotechnology, Год журнала: 2024, Номер 12

Опубликована: Апрель 11, 2024

The study of dose-response relationships underpins analytical biosciences. Droplet microfluidics platforms can automate the generation microreactors encapsulating varying concentrations an assay component, providing datasets across a large chemical space in single experiment. A classical method consists flow rate multiple solutions co-flowing into microchannel (producing different volume fractions) before contents water-in-oil droplets. This process be automated through controlling pumping elements but lacks ability to adapt unpredictable experimental scenarios, often requiring constant human supervision. In this paper, we introduce image-based, closed-loop control system for assessing and adjusting fractions, thereby generating unsupervised, uniform concentration gradients. We trained shallow convolutional neural network assess position laminar interface between two fluids used model adjust rates real-time. apply generate alginate microbeads which HEK293FT cells could grow three dimensions. stiffnesses ranged from 50 Pa close 1 kPa Young modulus were encoded with fluorescent marker. deep learning models based on YOLOv4 object detector efficiently detect both multicellular spheroids high-content screening images. allowed us map hydrogel stiffness spheroid growth.

Язык: Английский

Biophysical Reviews— 2021, the year that was

Biophysical Reviews, Год журнала: 2021, Номер 13(6), С. 803 - 811

Опубликована: Ноя. 22, 2021

Язык: Английский