Spot Count

This tutorial guides you through the Spot Count workflow - detecting and counting fluorescent spots (e.g. extracellular vesicles) in one image channel, with optional removal of calibration bead artefacts.

The workflow uses two pipelines:

1. EV Detection - detects spots in the EV channel.
2. Tetraspeck (optional) - detects calibration beads and removes any EV detections that coincide with them.

Prerequisites

• Fluorescence microscopy images containing EVs/spots in at least one channel.
• Tetraspeck bead images in a separate channel (optional).

Step 1: Project Setup

1. Open EVAnalyzer and set the Image directory to your image folder.
2. Set a Job name (e.g. spot-count-01).
3. If your images are from a plate experiment, configure Grouping and the filename regex.

Step 2: Define Classes

In the Classification tab, add:

• ch1@spot - the EV/spot class (adjust the prefix to match your fluorophore)
• tetraspeck@spot - the calibration bead class (only if using the Tetraspeck pipeline)

Step 3: EV Detection Pipeline

Create a new pipeline named EV Detection and add the following steps in order:

Step	Settings
Rolling Ball	Radius: 4, Type: Paraboloid - removes uneven background
Gaussian Blur or Blur	Kernel: 3, repeat 2× - reduces noise artefacts
Threshold	Method: Manual - adapt to your images; start with Min: 200
Connected Components	Min size: 3–10 px² recommended to suppress single-pixel noise
Watershed	Tolerance: 0.5 - separates closely touching spots
Extract ROIs	No settings
Classify ROIs	Target: ch1@spot; Min area: 3 px²; Min circularity: 0.1
Save Image	Path: images/${imageName} - save a control image (remove if not needed)

Threshold selection

Open the advanced preview to display the image histogram alongside the segmentation result. Set the minimum threshold just above the highest background-noise peak.

Step 4: Tetraspeck Pipeline (optional)

If you have a dedicated calibration-bead channel, create a second pipeline named Tetraspeck:

Step	Settings
Rolling Ball	Radius: 4, Type: Paraboloid
Gaussian Blur	Kernel: 3, repeat 2×
Threshold	Method: Manual - adapt to bead channel
Connected Components	-
Watershed	Tolerance: 0.5
Extract ROIs	-
Classify ROIs	Target: tetraspeck@spot; Min area: 5 px²
Classify ROIs (second)	Source: ch1@spot; Intersects with: tetraspeck@spot; Target: tetraspeck@spot - moves any EV spot that overlaps a bead to the tetraspeck class, effectively removing it from the EV count

Step 5: Run and Inspect

Click Play to run the analysis. When complete:

• Open the results viewer and check the ch1@spot count per image.
• Open control images in the images/ subfolder to visually verify segmentation.
• Adjust the threshold and rerun if objects appear over- or under-segmented.

Tips

• The Tetraspeck pipeline should be disabled if no calibration bead channel is available.
• Add a Classify ROIs step after the EV classifier with additional filters (e.g. Min area: 10) to gate the population further without rerunning the full analysis.