Related Experiment Video
Updated: Dec 3, 2025
05:10
Purification, Expansion, and Flow Cytometry-Based Phenotyping of Mouse Derived Bone Marrow Mesenchymal Stem Cells
Published on: July 11, 2025
730
Predicting quality decay in continuously passaged mesenchymal stem cells by detecting morphological anomalies
Yuto Takemoto1, Yuta Imai1, Kei Kanie2
1Graduate School of Pharmaceutical Sciences, Nagoya University, Furocho, Chikusa-ku, Nagoya 464-8601, Japan.
Journal of Bioscience and Bioengineering
|October 30, 2020
Summary
Image-based analysis of cell morphology can predict human mesenchymal stem cell (hMSC) quality decline. This method enhances cell manufacturing by detecting anomalies early, improving consistency and efficiency in cell therapies.
Area of Science:
- Biotechnology
- Cell Biology
- Data Science
Background:
- Cell therapy advancements necessitate consistent and efficient manufacturing technologies.
- Current morphological monitoring relies on manual skill, limiting reproducibility.
- Image-based analysis offers a data-driven alternative for quality evaluation in cell manufacturing.
Purpose of the Study:
- To develop a practical, morphology-based method for detecting quality anomalies in cell manufacturing.
- To adapt machine learning models for situations with limited anomaly data.
- To assess the performance of morphological parameters in discriminating quality decay in human mesenchymal stem cells (hMSCs).
Main Methods:
- Utilized time-course imaging to extract morphological parameters from hMSCs.
- Developed machine learning algorithms incorporating visualization and asymmetric statistical discrimination.
- Investigated the detection of anomalous quality decay during continuous hMSC passaging.
Main Results:
- Morphological parameters reflecting cellular population heterogeneity effectively predicted hMSC quality decay.
- Quality decay was predictable within 6 hours after cell seeding.
- The developed concept enables morphology-based, in-process quality monitoring.
Conclusions:
- A practical, morphology-based concept for in-process quality monitoring in cell manufacturing was established.
- This approach effectively detects quality anomalies, addressing data variation challenges in real-world facilities.
- The findings support the use of image-based morphological analysis for enhanced reproducibility and data-driven quality control in cell therapy production.
