iPSC Image Recognition

Reprogramming somatic cells to pluripotent stem cells (iPSCs) became a critical method in basic biological research, drug screening, and drug discovery since their discovery in 2006. A number of technologies have been used to induce pluripotency in somatic cells, each with some variation in process and effectiveness. While the effectiveness of these molecular toolkits have made huge improvements, selecting iPSC colonies during reprogramming often relies on human subjectivity.

With automated imaging, computer vision, and image recognition algorithms, this collaborative research sought to facilitate the iPSC colony selection process to optimize pluripotent cell line outcomes, and thereby enabling iPSC culture growth to be used in high-throughput processes, and reducing the cost and labor requirements for iPSC production.

This project was conducted at the Harvard Stem Cell Institute in collaboration with DR Vision Technologies and Nikon Corporation. This work was presented at presented at several international conferences and used to create cell biology imaging and analysis products. If you would like to learn more about this project visit some publications here, and here for information about the automated imaging systems used for this research.

*Please note that this work is not reported in full here to respect potentially sensitive IP.