“Image-based microfluidic cell sorting”

Cells show remarkable heterogeneity, even between cells with the same genome which are grown in the same conditions. To further understand these differences, it is often required to separate subpopulations of cells for analysing individually. One of the most widely used systems in cell sorting is Fluorescence Activated Cell Sorting (FACS). FACS is able to sort thousands of cells per second but relies on fluorescent labels and encapsulating the cells into droplets, in air, thus creating potentially harmful aerosols which may contaminate the samples.
Microfluidics is a powerful technique which can overcome this issue as cells can be manipulated within microscopic channels while remaining entirely surrounded by liquid. I am currently combining microfluidic cell sorting with high-speed, label-free image analysis for image-based cell sorting. Cells are imaged in flow using high speed microscopy imaging. These images are processed using machine learning algorithms to classify each cell into subpopulations based on features present in the image. This decision making is used to control the microfluidics to guide the cells into the relevant outlets, separating each subpopulation. This system opens the door to using morphological features to identify and separate subpopulations of cells for detailed study of cellular heterogeneity.