Skip to content

Workflow: Cell Counting

This workflow demonstrates how to count cells in a fluorescence microscopy image.

Pipeline Overview

Image Reader → Gaussian Blur → Binary Threshold → Remove Small Obj → Watershed → Data Table

Step-by-Step

1. Load the Image

Add an Image Reader node and select your microscopy image.

2. Pre-process

  • Gaussian Blur (sigma = 1-2): Reduces noise
  • Equalize Adapthist (optional): Enhances contrast for uneven illumination

3. Segment

  • Binary Threshold: Separate cells from background. Adjust the threshold until cells are cleanly separated.
  • Remove Small Obj: Eliminate noise artifacts (set minimum size in pixels)

4. Split Touching Cells

  • Watershed: Separates touching/overlapping cells using distance-based markers
  • Set Min Object Sep. to the approximate minimum distance between cell centers

5. Measure

The Watershed node outputs both a label image and a table with per-object measurements:

  • Area, perimeter, circularity
  • Centroid coordinates
  • Eccentricity, orientation
  • Intensity statistics (if image is connected)

6. Visualize

Connect the table to a Data Table Node for inline viewing, or to plot nodes for visualization.