With the creation of human induced pluripotent stem cells (iPS cells) and other advances, the movement toward utilization of cultured human cells is in full swing. At present, passage operations involving identification and removal (purification) of undesired cells and subdivision of cell monolayers are basically performed manually. However, the need to automate these operations is increasing rapidly in order to support anticipated growth in demand and stricter quality control for cell processed products.

Through joint research with the Institute of Physical and Chemical Research (RIKEN), Kataoka Corporation, Meijo University, and iPS PORTAL Inc., the researchers have developed a technology that uses a photoresponsive polymer and a laser to automatically process cultured cells at high speed based on deep learning, which is a machine learning method.

The developed technology automatically performs operations such as identification and purification of cultured cells based on machine learning, and cutting and uniform subdivision of cell monolayers at high-speed in order to meet anticipated future needs for mass utilization of human-derived cells. The technology is expected to contribute to drug discovery and regenerative medicine through automation of precision processing and quality control of cultured cells.