Skip to content

Walkthrough: Overview

This page provides an overview of the core concepts behind ZarrNii. It’s the starting point for understanding how to work with OME-Zarr, NIfTI, and ZarrNii’s transformation tools.

Core Concepts

1. Zarr and OME-Zarr

  • Zarr is a format for chunked, compressed N-dimensional arrays.
  • OME-Zarr extends Zarr with metadata for multidimensional microscopy images, supporting axes definitions and multiscale pyramids.

Key Features of OME-Zarr:

  • Axes Metadata: Defines spatial dimensions (e.g., x, y, z).
  • Multiscale Pyramids: Stores image resolutions at multiple scales.
  • Annotations: Includes OME metadata for visualization and analysis.

2. NIfTI

  • NIfTI is a neuroimaging file format, commonly used for MRI and fMRI data.
  • It supports spatial metadata, such as voxel sizes and affine transformations, for anatomical alignment.

3. ZarrNii

  • ZarrNii provides tools to bridge these formats while preserving spatial metadata and enabling transformations.

Main Features:

  • Read and write OME-Zarr and NIfTI formats.
  • Apply transformations like cropping, downsampling, and interpolation.
  • Apply segmentation algorithms through an extensible plugin system.
  • Convert between ZYX (OME-Zarr) and XYZ (NIfTI) axes orders.

Data Model

ZarrNii wraps datasets using the ZarrNii class, which has the following attributes:

  • darr: The dask array containing image data.
  • affine: An affine transformation matrix for spatial alignment.
  • axes_order: Specifies the data layout (ZYX or XYZ).
  • OME-Zarr Metadata:
  • axes: Defines the dimensions and units.
  • coordinate_transformations: Lists scaling and translation transformations.
  • omero: Contains channel and visualization metadata.

Example Workflow

Here’s a high-level example workflow using ZarrNii:

  1. Read Data: python from zarrnii import ZarrNii znimg = ZarrNii.from_ome_zarr("path/to/dataset.ome.zarr")

  2. Apply Transformations: python znimg_downsampled = znimg.downsample(level=2) znimg_cropped = znimg_downsampled.crop((0, 0, 0), (100, 100, 100))

  3. Apply Segmentation: python znimg_segmented = znimg.segment_otsu(nbins=256)

  4. Convert Formats: python znimg_cropped.to_nifti("output.nii") znimg_segmented.to_nifti("segmented.nii")

What’s Next?