COSMOS, a platform integrating AI and microfluidics, characterizes and sorts single cells based on real-time deep learning analysis of high-resolution brightfield images. Supervised deep learning models analyze morphology to sort cell lines and primary tissue without labels or stains. The platform's ability to capture deep visual characteristics enables efficient purification of viable cells with desired morphological traits, resolving limitations of existing real-time cell sorting technologies.

Mahyar Salek, Nianzhen Li, Hou-Pu Chou, Kiran Saini, Andreja Jovic, Kevin B. Jacobs, Chassidy Johnson, Vivian Lu, Esther J. Lee, Christina Chang, Phuc Nguyen, Jeanette Mei, Krishna P. Pant, Amy Y. Wong-Thai, Quillan F. Smith, Stephanie Huang, Ryan Chow, Janifer Cruz, Jeff Walker, Bryan Chan, Thomas J. Musci, Euan A. Ashley, Maddison (Mahdokht) Masaeli