Integration and harmonization of cell shape images for generative modeling.

Area of Science:

• Biomaterials Science
• Cell Biology
• Computational Biology

Background:

• Cell imaging data integration is crucial for understanding cell-material interactions.
• Variability in experimental conditions and imaging formats hinders quantitative analysis of cellular responses to mechanical cues like stiffness and topography.

Purpose of the Study:

• To develop a framework for harmonizing 2D and 3D immunofluorescent cell imaging datasets.
• To enable quantitative analysis of cellular responses to biomaterial-derived mechanical cues.
• To generate realistic synthetic cell data for augmenting machine learning analyses.

Main Methods:

• Combined compact, interpretable cell shape models with generative artificial intelligence.
• Developed a method to harmonize 2D and 3D immunofluorescent datasets.
• Utilized AI to capture cell morphology and associated biological features.

Main Results:

• Successfully harmonized diverse cell imaging datasets within defined experimental contexts.
• Enabled the generation of realistic synthetic cells, including rare phenotypes.
• Demonstrated the potential for scalable in silico studies.

Conclusions:

• The proposed framework advances data-driven investigation of cellular responses to biomaterial mechanics.
• This approach facilitates more robust and scalable research in cell-material interactions.
• AI-powered data harmonization and synthetic data generation are key for future cell imaging studies.