2. Early prediction of cellular quality using cell morphology informatics

Early prediction of cellular quality using cell morphology informatics

  • Graduate School of Pharmaceutical Sciences
  • Department of Basic Medicinal Sciences
  • Laboratory of Cell and Molecular Bioengineering
  • Division of Biosciences

Ryuji Kato [Associate Professor]


Outline of Seeds

In regenerative medicine products, safe and efficient production technology is important for both patients and cell-producing companies. This technology is a technology of training artificial intelligence with phase contrast microscopic image-derived cellular morphological information and their experimentally determined quality. By our analysis, we can quantitatively predict the future cellular potential, such as their growth or differentiation rate, only from the label-free early images of cultured cells. This method enables to control the quality of cellular-based materials.

Novelty and originality of this research

In most of the cell image analysis works for cellular evaluation uses florescent labeling. However, our technology only uses label-free phase contrast images for cellular evaluation, which is essential for production of regenerative medicine products. Moreover, since we have shown in publications about the effective usage of artificial intelligence/machine learning algorithms for using such cellular morphological information to predict their future potentials with high accuracy. This technology is a collaborative technology developed with Nikon corporation, therefore there are many experiences in actual instrumental/software setting. Therefore this technology can feasibly be introduced in new facilities.

Application and research area for Industry collaboration

We have been applying this technology for various applications;
- Quality prediction of cells manufactured for cell therapy
- Process optimization of cell manufacturing for cell-based products
- Evaluation and optimization of culture protocols in regenerative medicine
We are welcome to collaborate with following companies;
- Company that relate to regenerative medicine or cell therapies

Key Takeaway

We are not simple "image analysis group". We have a strength of wide ability that enable to support the whole analysis from the data acquisition (cell culture) to the algorithm development as total developer.


Regenerative medicine, Cell culture, Quality control, Image analysis, Cell morphology data, Optimization research, Efficiency research


  • Cell image analysis technology
  • Cell morphology informatics
  • Cell quality prediction algorithms


  • Automated cell imaging system

Monographs, Papers and Articles

  • Kato R, Matsumoto M, Sasaki H, Joto R, Okada M, Ikeda Y, Kanie K, Suga M, Kinehara M, Yanagihara K, Liu Y, Uchio-Yamada K, Fukuda T, Kii H, Uozumi T, Honda H, Kiyota Y, Fueur K M. Parametric analysis of colony morphology of non-labelled live human pluripotent stem cells for cell quality control., Sci. Rep. 2016, 6, 34009
  • Sasaki H, Takeuchi I, Okada M, Sawada R, Kanie K, Kiyota Y, Honda H, Kato R. Label-Free Morphology-Based Prediction of Multiple Differentiation Potentials of Human Mesenchymal Stem Cells for Early Evaluation of Intact Cells. PLoS One, 2014, 9(4), e93952.
  • Matsuoka F, Takeuchi, I, Agata H, Kagami H, Shiono H, Kiyota Y, Honda H, Kato R. Characterization of time-course morphological features for efficient prediction of osteogenic potential in human mesenchymal stem cells., Biotechnol. Bioeng. 2014, 111(7), 1430-1439.