[3S06a] 16. Cell and Tissue Engineering for Biomedical Process Innovation - Part 2

2:40 PM - 3:20 PM JST(5:40 AM - 6:20 AM UTC)

[3CTBP-07] Image-Based AI for Advanced Quality Control in the Manufacturing of Regenerative Medicine Products

○Ryuji Kato^1,2 (1. Graduate School of Pharmaceutical Sciences, Nagoya University (Japan), 2. Institute of Nano-Life-Systems, Nagoya University (Japan))

Keywords:

Cell manufacturing,Image-based AI,Lable-free,Quality control,CDMO,Non-invasive

In recent years, advances in cell science have enabled the reliable generation and utilization of diverse cell types. In particular, stem cell technologies, including induced pluripotent stem (iPS) cells, have made it possible to obtain human cells tailored to specific applications. Combined with rapid progress in genome editing and single-cell analysis, cellular functions and properties can now be precisely controlled at the genetic level, enabling the optimization of cells for defined purposes. At the same time, the integration of automated culture systems and artificial intelligence (AI)-driven data processing has transformed cell culture into a more controlled and scalable process. These developments have led to the emergence of large-scale manufacturing technologies that treat cells as producible materials. The commercialization of scale-up and automation technologies has further accelerated this trend, driving rapid growth in the global market for contract manufacturing and development organizations (CMOs/CDMOs) and intensifying international competition. Despite these advances, quality assessment and control in cell manufacturing remain insufficiently established. In particular, non-destructive evaluation methods for living cells are still limited, and robust quality criteria beyond safety and efficacy are not well defined. As a result, cell quality evaluation continues to rely heavily on empirical approaches, such as microscopic observation, reflecting a historical dependence on destructive, molecular-based assays. However, as the cell manufacturing industry matures, there is an increasing demand for technologies that enable non-invasive, quantitative evaluation of living cells. Over the past two decades, our research group has developed morphology-based cell analysis using advanced informatics approaches. This technology enables quantitative, label-free, and non-invasive evaluation of cell morphology from imaging data and has been applied to quality assessment and control in various contexts. In this presentation, we introduce our AI-powered image-based analysis framework and discuss its potential as a next-generation quality control strategy for cell manufacturing.