Classification of differentially activated groups of fibroblasts using morphodynamic and motile features

Area of Science:

• Oncology
• Cell Biology
• Biophysics

Background:

• Fibroblasts significantly influence cancer progression, with varying activation states impacting tumor growth.
• Current molecular markers for cancer-associated fibroblasts lack specificity, hindering precise identification of their activation states.
• Fibroblast morphology and motility are indicative of their activation status and gene expression.

Purpose of the Study:

• To develop an artificial intelligence (AI)-based framework for classifying differentially activated fibroblasts.
• To leverage morphodynamic and motile features for fibroblast characterization.
• To overcome limitations of traditional molecular markers in identifying fibroblast subtypes.

Main Methods:

• Utilized label-free live-cell imaging of fibroblasts co-cultured with breast cancer cells.
• Applied deep learning and machine learning algorithms to extract morphodynamic and motile features.
• Developed an AI classification framework based on these biophysical cellular features.

Main Results:

• Morphodynamic and motile features provide robust insights into fibroblast activation states.
• The AI framework successfully distinguished between different fibroblast activation profiles.
• Biophysical features complement and overcome limitations of molecular markers.

Conclusions:

• AI-driven analysis of fibroblast morphodynamics and motility offers a novel approach to characterize activation states.
• This biophysical classification framework enhances understanding of the tumor microenvironment.
• The findings hold significant potential for advancing targeted cancer therapies.