Cell Topography Regression Learner (CTRL)

Cell Topography Regression Learner (CTRL) estimates single-cell volumes in mammalian cells from label-free differential interference contrast (DIC) microscopy images by reconstructing three-dimensional cell topography with deep learning.

Key Features:

• Label-Free Measurement: Uses differential interference contrast (DIC) microscopy images alone to measure cell volume without fluorescent labels.
• Deep Learning Integration: Employs a deep learning algorithm to reconstruct cell topography from DIC images for precise volume estimation.
• Fluorescence Exclusion Method: Incorporates the fluorescence exclusion method to delineate cell boundaries and improve volume estimates.
• Quantitative Accuracy: Produces high quantitative accuracy in single-cell volume measurements.
• Versatility Across Conditions: Applicable across a broad spectrum of biological and experimental conditions.

Scientific Applications:

• Single-Cell Volume Dynamics: Tracks single-cell volume dynamics over extended periods.
• Cell Growth Control Studies: Applied to studies of cell growth control and cellular behavior analysis.
• HT1080 Fibrosarcoma Experiments: Used with HT1080 fibrosarcoma cells to observe correlations between cell size at division and birth (sizer principle).
• Cell Cycle Fluctuation Analysis: Detects reductions in cell size fluctuations during specific phases of the cell cycle.

Methodology:

DIC microscopy images are captured and processed by a deep learning algorithm that reconstructs three-dimensional cell topography, from which volumes are computed by calculating spatial dimensions derived from the reconstruction.