The current approach for segmenting different sizes in C. elegans in wells with non-uniform backgrounds (i.e., contamination, poor illumination, leftover HB101) isn't good (to be frank). Cellpose could offer a substantial improvement.
Tasks:
[x] Generate and evaluate a model using representative images (normal sized 48 hr worms, IVM-treated worms, large worms that have been grown on a plate)
[x] Develop a CellProfiler pipeline that will use the Cellpose output
[x] Update the file list R script generator
[x] Choose dx images
This will be first pipeline that requires mixing and matching of Python/CellProfiler modules/pipelines and would be best served by waiting until we move from module --> pipeline logic. However, a more immediate bandaid could be to run the cellpose command in the wrapper immediately before the running the file list generator.
The cellpose branch has been created to work on this (69065d3e85c099c943fb6677180b376d7ca4c96b)
The current approach for segmenting different sizes in C. elegans in wells with non-uniform backgrounds (i.e., contamination, poor illumination, leftover HB101) isn't good (to be frank). Cellpose could offer a substantial improvement.
Tasks:
This will be first pipeline that requires mixing and matching of Python/CellProfiler modules/pipelines and would be best served by waiting until we move from module --> pipeline logic. However, a more immediate bandaid could be to run the cellpose command in the wrapper immediately before the running the file list generator.
The
cellposebranch has been created to work on this (69065d3e85c099c943fb6677180b376d7ca4c96b)