Rl/process dic timelapses

[lanery]

This PR is a sort of crude and minimal implementation for processing DIC time lapses of agar microchambers such that it could be used to (quickly) process old in-house datasets.

The modalityFlag tag is parsed from the nd2 metadata to determine the imaging modality of the time lapse -- or at least whether or not it is brightfield. The only real change in processing this triggers is how the thresholding is done for the time lapses (see below). This approach was tested on a handful of in-house DIC time lapses acquired previously, but is not expected to be very robust. The DIC data was previously processed and analyzed using a different pipeline for the pub: "Phenotypic differences between interfertile Chlamydomonas species". While this dataset will not be included or discussed in the pub associated with this repository, the idea is to still be able to use this repository to process DIC time lapses for the sake of convenience.

• Brightfield --> Li thresholding (same as before)
• DIC --> Otsu thresholding

This PR also includes some changes to default parameter values and some minor refactoring changes.

PR checklist

• Describe the changes you've made.
• If you've added new functionality, make sure that the documentation is updated accordingly.

[keithchev]

Just a few minor comments and caught what I think might be some bugs/typos.

[lanery]

After seeing how many issues/quirks you found in pool_processor.has_cells() I felt the need to provide some evidence that it worked at all....

Code to generate above plot

from pathlib import Path
import skimage as ski
import numpy as np
import matplotlib.pyplot as plt
from mpl_interactions import hyperslicer
from chlamytracker.timelapse import Timelapse
from chlamytracker.pool_finder import PoolFinder

nd2_file = Path(
    "/Users/ryanlane/Projects/2024-unicellular-tracking/pub_videos/pools_data/s1-cr3-t-timelapse_#6.nd2"
)

poolfinder = PoolFinder(nd2_file)
pools = poolfinder.extract_pools()

%matplotlib ipympl

ncols, nrows = max(pools)
fig, axes = plt.subplots(
    nrows=nrows + 1,
    ncols=ncols + 1,
    figsize=(2.5 * ncols, 2.5 * nrows),
)

for (ix, iy), pool in pools.items():

    has_cells = pool.has_cells()

    if ix == iy == 0:
        slider = hyperslicer(
            pool.raw_data,
            vmin=14786,
            vmax=35060,
            cmap="Greys_r",
            display_controls=True,
            ax=axes[iy, ix],
        )

    else:
        slider = hyperslicer(
            pool.raw_data,
            vmin=14786,
            vmax=35060,
            cmap="Greys_r",
            controls=slider,
            display_controls=False,
            ax=axes[iy, ix],
        )

    axes[iy, ix].set_title(f"Cells :: {has_cells}")
    axes[iy, ix].axis("off")

[keithchev]

After seeing how many issues/quirks you found in pool_processor.has_cells() I felt the need to provide some evidence that it worked at all....

Lol, nice. I would guess that the cells are so dark compared to the rest of the image that exactly how std or var are called won't matter that much...