Examples: Working with Zarr and NIfTI

This section provides practical workflows for using ZarrNii with OME-Zarr and NIfTI datasets.

Loading Datasets

From OME-Zarr

Load a dataset from an OME-Zarr file and inspect its metadata:

from zarrnii import ZarrNii

# Load OME-Zarr dataset
znimg = ZarrNii.from_ome_zarr("path/to/dataset.zarr")

# Inspect data
print("Shape:", znimg.darr.shape)
print("Affine matrix:\n", znimg.affine.matrix)

From NIfTI

Load a NIfTI dataset and inspect its attributes:

# Load NIfTI dataset
znimg = ZarrNii.from_nifti("path/to/dataset.nii")

# Inspect data
print("Shape:", znimg.darr.shape)
print("Affine matrix:\n", znimg.affine.matrix)

Performing Transformations

Downsampling

Reduce the resolution of the dataset using the downsample method:

# Downsample by level
downsampled = znimg.downsample(level=2)
print("Downsampled shape:", downsampled.darr.shape)

Cropping

Extract a specific region from the dataset using bounding boxes:

Voxel Space:

cropped = znimg.crop_with_bounding_box((10, 10, 10), (50, 50, 50))
print("Cropped shape:", cropped.darr.shape)

RAS Space:

cropped_ras = znimg.crop_with_bounding_box(
    (-20, -20, -20), (20, 20, 20), ras_coords=True
)
print("Cropped shape:", cropped_ras.darr.shape)

Combining Affine Transformations

Apply multiple transformations to the dataset in sequence:

from zarrnii.transforms import AffineTransform

# Define transformations
scale = AffineTransform.from_scaling((2.0, 2.0, 1.0))
translate = AffineTransform.from_translation((10.0, -5.0, 0.0))

# Apply transformations
transformed = znimg.apply_transform(scale, translate, ref_znimg=znimg)
print("Transformed affine matrix:\n", transformed.affine.matrix)

Saving Results

To OME-Zarr

Save the dataset to OME-Zarr format:

znimg.to_ome_zarr("output.zarr", max_layer=3, scaling_method="local_mean")

To NIfTI

Save the dataset to NIfTI format:

znimg.to_nifti("output.nii")

Advanced Example: Full Workflow

Combine multiple operations in a single workflow:

from zarrnii import ZarrNii
from zarrnii.transforms import AffineTransform

# Load an OME-Zarr dataset
znimg = ZarrNii.from_ome_zarr("path/to/dataset.zarr")

# Crop the dataset
cropped = znimg.crop_with_bounding_box((10, 10, 10), (100, 100, 100))

# Downsample the dataset
downsampled = cropped.downsample(level=2)

# Apply an affine transformation
scale = AffineTransform.from_scaling((1.5, 1.5, 1.0))
transformed = downsampled.apply_transform(scale, ref_znimg=downsampled)

# Save the result as a NIfTI file
transformed.to_nifti("final_output.nii")

Summary

In this section, you learned how to: - Load datasets from OME-Zarr and NIfTI formats. - Perform transformations like downsampling, cropping, and affine transformations. - Save results back to OME-Zarr or NIfTI.

Next: - Explore the API Reference for in-depth details about ZarrNii's classes and methods. - Check the FAQ for answers to common questions.