How to detect sub-cellular spots

This guide shows you how to count and measure spots inside your cell masks.

Reference keys: single-cell measurements

Prerequisite: you have segmented your cell population of interest accurately.

  1. Go to the MEASURE section and click on the icon to enter measurement settings.

  2. Scroll down to the SPOT DETECTION section.

  3. Tick the Perform spot detection option.

  4. Press the icon on the right side to set up spot detection visually.

  5. Set up the channel interest using the controls on the Right Panel. You can change the displayed frame and adjust contrast to see the spots clearly.

    Note

    The viewer is split into two panels: - Left Panel: Contains all detection settings (Channels, Thresholds, Preprocessing). - Right Panel: Displays the image and visualization controls.

  6. In the Left Panel, set the detection channel to the same channel as above.

    Tip

    If the image is noisy or the background is uneven, or if the spots are dark (e.g., RICM), use the Preprocessing options below the channel selection.

    • For noisy images: Add a gaussian or median filter (e.g., sigma=1 or size=3).

    • For uneven background: Add a tophat filter (white tophat) to isolate bright spots.

    • For dark spots: Add an invert filter roughly at the bit-depth max (e.g., 255 or 65535) to make spots bright.

    You can check the Preview box (below the Preprocessing list) to see the effect of your filters on the image (e.g. smoothing, inversion). This preview does not show the detected spots, only the enhanced image.

  7. Estimate visually the average spot diameter (in pixels). You can zoom in on the image.

  8. Set the Detection threshold to 0 initially.

  9. Press Set (next to Diameter or Threshold) to run the spot detection.

    • Visual Feedback: Detected spots will appear as red circles.

    • Note: At threshold 0, you will likely see many false positives (background noise detected as spots). This is normal.

  10. Gradually increase the detection threshold and press Set again to update the preview.

    • The goal is to filter out the false positives until only the real spots remain circled.

    • If spots are not detected even at threshold 0, try adjusting the diameter or checking your preprocessing.

    • Note: The detection uses the preprocessed image if filters are listed, regardless of whether the “Preview” checkbox is ticked.

  11. Once the detection is satisfactory, press Add measurement.

  12. Save the new measurement settings.

  13. Check the MEASURE option and press Submit to measure.

See celldetective.measure.extract_blobs_in_image() for more information about the algorithm used for single-spot detection.