Image Semgmentation¶

This is a minimal tutorial on how to use ngio for image segmentation.

Step 1: Setup¶

We will first implement a very simple function to segment an image. We will use skimage to do this.

In [1]:

# Setup a simple segmentation function
import numpy as np
import skimage


def otsu_threshold_segmentation(image: np.ndarray, max_label: int) -> np.ndarray:
    """Simple segmentation using Otsu thresholding."""
    threshold = skimage.filters.threshold_otsu(image)
    binary = image > threshold
    label_image = skimage.measure.label(binary)
    label_image += max_label
    label_image = np.where(binary, label_image, 0)
    return label_image

# Setup a simple segmentation function import numpy as np import skimage def otsu_threshold_segmentation(image: np.ndarray, max_label: int) -> np.ndarray: """Simple segmentation using Otsu thresholding.""" threshold = skimage.filters.threshold_otsu(image) binary = image > threshold label_image = skimage.measure.label(binary) label_image += max_label label_image = np.where(binary, label_image, 0) return label_image

Step 2: Open the OmeZarr container¶

In [2]:

from pathlib import Path

from ngio import open_ome_zarr_container
from ngio.utils import download_ome_zarr_dataset

# Download the dataset
download_dir = Path(".").absolute().parent.parent / "data"
hcs_path = download_ome_zarr_dataset("CardiomyocyteTiny", download_dir=download_dir)
image_path = hcs_path / "B" / "03" / "0"

# Open the ome-zarr container
ome_zarr = open_ome_zarr_container(image_path)

from pathlib import Path from ngio import open_ome_zarr_container from ngio.utils import download_ome_zarr_dataset # Download the dataset download_dir = Path(".").absolute().parent.parent / "data" hcs_path = download_ome_zarr_dataset("CardiomyocyteTiny", download_dir=download_dir) image_path = hcs_path / "B" / "03" / "0" # Open the ome-zarr container ome_zarr = open_ome_zarr_container(image_path)

Downloading data from 'https://zenodo.org/records/13305156/files/20200812-CardiomyocyteDifferentiation14-Cycle1.zarr.zip' to file '/home/runner/work/ngio/ngio/data/20200812-CardiomyocyteDifferentiation14-Cycle1.zarr.zip'.

Unzipping contents of '/home/runner/work/ngio/ngio/data/20200812-CardiomyocyteDifferentiation14-Cycle1.zarr.zip' to '/home/runner/work/ngio/ngio/data/'

Step 3: Segment the image¶

For this example, we will not segment the image all at once. Instead we will iterate over the image FOVs and segment them one by one.

In [3]:

# First we will need the image object and the FOVs table
image = ome_zarr.get_image()
roi_table = ome_zarr.get_table("FOV_ROI_table", check_type="roi_table")

# Second we need to derive a new label image to use as target for the segmentation

label = ome_zarr.derive_label("new_label", overwrite=True)

max_label = 0  # We will use this to avoid label collisions
for roi in roi_table.rois():
    image_data = image.get_roi(roi=roi, c=0)  # Get the image data for the ROI

    image_data = image_data.squeeze()  # Remove the channel dimension
    roi_segmentation = otsu_threshold_segmentation(
        image_data, max_label
    )  # Segment the image

    max_label = roi_segmentation.max()  # Get the max label for the next iteration

    label.set_roi(
        roi=roi, patch=roi_segmentation
    )  # Write the segmentation to the label image

# First we will need the image object and the FOVs table image = ome_zarr.get_image() roi_table = ome_zarr.get_table("FOV_ROI_table", check_type="roi_table") # Second we need to derive a new label image to use as target for the segmentation label = ome_zarr.derive_label("new_label", overwrite=True) max_label = 0 # We will use this to avoid label collisions for roi in roi_table.rois(): image_data = image.get_roi(roi=roi, c=0) # Get the image data for the ROI image_data = image_data.squeeze() # Remove the channel dimension roi_segmentation = otsu_threshold_segmentation( image_data, max_label ) # Segment the image max_label = roi_segmentation.max() # Get the max label for the next iteration label.set_roi( roi=roi, patch=roi_segmentation ) # Write the segmentation to the label image

Step 4: Consolidate the segmentation¶

The new_label has data only at a single resolution lebel. To consolidate the segmentation to all other levels we will need to call the consolidate method.

In [4]:

label.consolidate()

label.consolidate()

Plot the segmentation¶

In [5]:

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import ListedColormap

rand_cmap = np.random.rand(1000, 3)
rand_cmap[0] = 0
rand_cmap = ListedColormap(rand_cmap)

fig, axs = plt.subplots(2, 1, figsize=(8, 4))
axs[0].set_title("Original image")
axs[0].imshow(image.get_array(c=0, z=1).squeeze(), cmap="gray")
axs[1].set_title("Final segmentation")
axs[1].imshow(label.get_array(z=1).squeeze(), cmap=rand_cmap)
for ax in axs:
    ax.axis("off")
plt.tight_layout()
plt.show()

import matplotlib.pyplot as plt import numpy as np from matplotlib.colors import ListedColormap rand_cmap = np.random.rand(1000, 3) rand_cmap[0] = 0 rand_cmap = ListedColormap(rand_cmap) fig, axs = plt.subplots(2, 1, figsize=(8, 4)) axs[0].set_title("Original image") axs[0].imshow(image.get_array(c=0, z=1).squeeze(), cmap="gray") axs[1].set_title("Final segmentation") axs[1].imshow(label.get_array(z=1).squeeze(), cmap=rand_cmap) for ax in axs: ax.axis("off") plt.tight_layout() plt.show()