Quick look at MACSIMA data from https://cloud.irc.ugent.be/public/index.php/s/WPRWfQi6MKSF3Ex

macsima_liver.py

from pathlib import Path
import tifffile
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.backends.backend_pdf import PdfPages
from stardist.models import StarDist2D
from csbdeep.utils import normalize

_dir = Path("HumanLiverH35")
files = list(
    filter(lambda x: not x.name.startswith("."), sorted(_dir.glob("*.ome.tif")))
)

_dapi = list()
_stack = list()
channels = list()
for f in tqdm(files):
    name = f.stem.split("ROI-01_A-")[-1].replace(".ome", "")
    if name.startswith("DAPI"):
        _dapi.append(tifffile.imread(f))
    else:
        _stack.append(tifffile.imread(f))
        channels.append(name)
dapi = np.asarray(_dapi).mean(0)
del _dapi
# stack = np.asarray(_stack)

model = StarDist2D.from_pretrained("2D_versatile_fluo")
mask = model.predict_instances(normalize(dapi))[0]

fig, ax = plt.subplots()
ax.imshow(np.log1p(dapi))
ax.contour(mask, cmap="Reds", linewidths=0.25)
ax.axis("off")
ax.set(title="DAPI")
fig.savefig("mask.pdf")
plt.close(fig)

with PdfPages("stack.pdf") as pdf:
    fig, ax = plt.subplots()
    ax.imshow(np.log1p(dapi))
    ax.axis("off")
    ax.set(title="DAPI")
    pdf.savefig(fig)
    plt.close(fig)

    for channel, image in tqdm(zip(channels, _stack), total=len(channels)):
        fig, ax = plt.subplots()
        ax.imshow(np.log1p(image))
        ax.axis("off")
        ax.set(title=channel)
        pdf.savefig(fig)
        plt.close(fig)