Towards a comprehensive approach for characterizing cell activity in bright-field microscopic images

Area of Science:

• Cell Biology
• Biophysics
• Computational Biology

Background:

• Quantitative measurement of cellular physical responses via light microscopy is challenging, often relying on subjective qualitative descriptions.
• Cellular migration and morphology are key indicators of physical cell activity but are difficult to quantify objectively.
• Existing methods lack precision in localizing adhesive cells and minimizing contaminant interference.

Purpose of the Study:

• To develop a comprehensive, quantitative approach for estimating physical cell activity based on migration and morphology.
• To statistically analyze cell populations for drug-induced changes in physical properties.
• To validate a novel deep learning method for precise cell detection and tracking.

Main Methods:

• Developed a customized encoder-decoder deep learning model for cell detection and tracking.
• Applied statistical analysis to quantify cell migration and morphology within a defined field of view and timespan.
• Investigated the effects of cytochalasin D and taxol on human neuroblastoma SH-SY5Y cell populations.

Main Results:

• Successfully quantified the influence of cytochalasin D and taxol on SH-SY5Y cell physical activity.
• Demonstrated precise cell localization, even for adhesive cells, and minimized confusion from contaminants.
• Validated the accuracy of the deep learning approach in quantifying cell activity.

Conclusions:

• The developed deep learning method provides an accurate and robust approach for quantitative analysis of cell physical activity.
• This method overcomes the limitations of subjective assessment, enabling precise measurement of drug effects on cell behavior.
• The approach is viable for studying cellular responses in various biological and pharmacological contexts.