Finding and Selecting Individual worms

I am currently using pipeline 2 from the worm toolbox to select individual c elegans worms . However when I analyse images all I am able to detect are the channels of my microfluidic chamber and no worms.


The example pipelines make the assumption of worms contained within a well of a microtiter plate, with the detection parameters set accordingly. Therefore, changing both the type of image and the image quality will resulting in the pipeline not working as expected.

The original pipeline works to identify the well first, mask it out, and then identify the worms from the remainder. Your assay will most likely need to work similarly but the modules/settings will need to be adjusted appropriately.

Some suggestions:

  • Try to acquire your images such that the chamber is placed in same location in all images. Also, try to make sure that only one chamber is present per image.

  • Make sure to keep all acquisition settings the same across your experiment.

  • With the above in mind, I would suggest a workflow similar to the following:[list]
    *]Taking an image of an empty chamber, and using it as a template image.

  • With that, you can identify the chamber itself minus the worms using IdentifyPrimaryObejcts.

  • Then you can use the Align module to line up each incoming image with the template image

  • Mask the aligned image with the chamber object (using MaskImage).

  • Then you can use CorrectIlluminateCalculate/CorrectIlluminateApply to remove background heterogeneities.

  • Use IdentfyPrimaryObjects to identify the worms from the masked, corrected image.

I should also mention that it appears that the worms are in agar in the 4HG3 image (since I see worm tracks). The pipelines were designed with worms in water in mind; we’ve found that worms in agar are much harder to deal with from an image processing standpoint.