Automation and improvement of WBC mechanical profiling in deformability cytometry

Area of Science:

• Biophysics
• Cell Biology
• Medical Diagnostics

Background:

• Deformability cytometry (DC) offers high-throughput mechanical profiling of blood cells.
• Label-free classification of white blood cells (WBCs) is challenging.
• Current analysis relies on manual gating, limiting scalability and reproducibility.

Purpose of the Study:

• To develop a fully automated and generalizable framework for WBC classification in DC experiments.
• To overcome limitations of supervised models and manual gating.
• To enable accurate mechanophenotyping of WBCs, including activated cells.

Main Methods:

• Utilized box filters and unsupervised clustering for cell population analysis.
• Employed cell shape features from high-accuracy segmentation and texture features from bright-field images.
• Derived shape features from the original contour, not the convex hull, for improved accuracy.

Main Results:

• Validated the automated method against manual gating across diverse experimental conditions (donors, anticoagulants, collection sources, brightness).
• Demonstrated consistent reliability and robustness of the automated classification.
• Enabled inclusion and morphological characterization of WBCs with membrane protrusions.

Conclusions:

• The automated framework enhances accuracy, consistency, and scalability of WBC mechanophenotyping.
• This approach facilitates high-throughput analysis across large patient cohorts.
• The method improves reproducibility and reduces processing time compared to manual gating.