Automated Image-Based Cell Sorting by Targeted Photopolymerization

We present an automated image-based cell sorting method capable of a through-put of hundreds of cells per second. Our microscopy platform integrates automated image acquisition, machine-learning based classification and sub-sequent depletion of up to 99.98 % of negative cells by spatially controlled photo-encapsulation, preserving target cells for collection. Applied to peripheral blood mononuclear cells, the method achieves the label-free, morphology-based enrichment of activated T cells for downstream applications in biomedicine.