Testing Equality of Cell Populations Based on Shape and Geodesic Distance

Area of Science:

• Cell Biology
• Biophysics
• Computational Biology

Background:

• Image cytometry and automated microscopy enable large-scale analysis of cellular image data.
• Evaluating cell population equality is crucial for various biological and toxicological studies.
• Traditional methods often rely on simple metrics like cell size and shape, which may not capture complex population differences.

Purpose of the Study:

• To demonstrate the effectiveness of cell shape analysis for testing the equality of cell populations using image data.
• To adapt and apply multivariate nonparametric statistical hypothesis tests for this purpose.
• To compare the efficacy of shape space theory-derived geodesic distance with traditional metrics like cell area and roundness.

Main Methods:

• Review of shape space theory to quantify shape differences using geodesic distance.
• Adaptation of several multivariate nonparametric statistical hypothesis tests.
• Application of these tests to image cytometry data to compare cell populations.
• Comparison of geodesic distance with cell spread area and roundness as distinguishing features.

Main Results:

• Geodesic distance is a more effective feature than cell spread area and roundness for distinguishing between cell populations.
• Statistical tests based on geodesic distance successfully detect natural cellular perturbations.
• Kolmogorov-Smirnov tests using area and roundness failed to detect these perturbations.

Conclusions:

• Cell shape analysis, particularly using geodesic distance, offers a robust method for assessing cell population equality from image data.
• This approach provides greater sensitivity in detecting subtle cellular changes compared to traditional metrics.
• Geodesic distance-based hypothesis testing is a valuable advancement for in vitro screening and biological analysis.