Multi-color cell tracking#
Now you are on your own! Try to apply the concepts you learned in the previous sections to a new dataset.
After we go through the dataset, we will provide some cues on how we would approach the problem.
Feel free to ask for help if you get stuck or want to discuss your results.
Download data#
We will download the data from our public server. The dataset is metastic breat cancer cells with RGB lentiviral marking, data from Lammerding lab.
!wget -nc https://public.czbiohub.org/royerlab/ultrack/multi-color-cytoplasm.tif
Import libraries#
We will start with the minimal import of napari and tifffile, update this chunk as needed.
import napari
from napari.utils import nbscreenshot
from tifffile import imread
Setting up napari viewer#
Napari is setup the same way as the previous tutorial.
viewer = napari.Viewer()
viewer.window.resize(1800, 1200)
def screenshot() -> None:
display(nbscreenshot(viewer))
Data loading#
We will load the data and display in the viewer as an RGB image.
image = imread("multi-color-cytoplasm.tif")
print("Image array shape", image.shape)
viewer.add_image(image, rgb=True, name="multi-color-cytoplasm")
screenshot()
Image array shape (300, 1440, 1920, 3)
Tip
The image stack have 3 channels, since we don’t have multi-color segmentation model (or at least I don’t), we could process each channel individually and then combine them using Ultrack’s to obtain a final segmentation and tracking without complicated heuristics for multi-color tracking.