med-blosc2 0.0.5

A standardized blosc2 image reader and writer for medical images.

MedBlosc2

A standardized Blosc2 image reader and writer for medical images. The MedBlosc2 file format (".mb2nd") is a Blosc2-compressed container with standardized metadata support for N-dimensional medical images. Plain ".b2nd" files are also supported, but they do not participate in the MedBlosc2 metadata standard.

Installation

You can install med-blosc2 via pip:

pip install med-blosc2

API

See API.md for the full MedBlosc2 api, including argument descriptions and types.

Metadata schema

See SCHEMA.md for the full MedBlosc2 metadata schema, including field descriptions and types.

Usage

Below are common usage patterns for loading, saving, and working with metadata.

Default usage

import numpy as np
from med_blosc2 import MedBlosc2, Meta

array = np.random.random((128, 256, 256)).astype(np.float32)
image = MedBlosc2(array)
image.save("sample.mb2nd")

Memory-mapped loading

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd", mmap=True)
# image.array is a blosc2.ndarray.NDArray when mmap=True, otherwise a np.ndarray

Loading and saving

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd")
image.save("copy.mb2nd")

Metadata inspection and manipulation

import numpy as np
from med_blosc2 import MedBlosc2

array = np.random.random((64, 128, 128)).astype(np.float32)
image = MedBlosc2(
    array,
    spacing=(1.0, 1.0, 1.5),
    origin=(10.0, 10.0, 30.0),
    direction=[[1, 0, 0], [0, 1, 0], [0, 0, 1]],
    meta=Meta(image={"patient_id": "123", "modality": "CT"}, is_seg=False),
)

print(image.spacing)  # [1.0, 1.0, 1.5]
print(image.origin)  # [10.0, 10.0, 30.0]
print(image.meta.image)  # {"patient_id": "123", "modality": "CT"}

image.spacing[1] = 5.3
image.meta.image["study_id"] = "study-001"
image.save("with-metadata.mb2nd")

Copy metadata with overrides

import numpy as np
from med_blosc2 import MedBlosc2

base = MedBlosc2("sample.mb2nd")
array = np.random.random(base.shape).astype(np.float32)

image = MedBlosc2(
    array,
    spacing=(0.8, 0.8, 1.0),
    copy=base,  # Copies all non-explicitly set arguments from base
)

image.save("copied-metadata.mb2nd")

Patch size variants

Default patch size (192):

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd")
image.save("default-patch.mb2nd")

Custom isotropic patch size (512):

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd")
image.save("patch-512.mb2nd", patch_size=512)

Custom non-isotropic patch size:

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd")
image.save("patch-non-iso.mb2nd", patch_size=(128, 192, 256))

Manual chunk/block size:

from med_blosc2 import MedBlosc2

image = MedBlosc2("sample.mb2nd")
image.save("manual-chunk-block.mb2nd", patch_size=None,
           chunk_size=(1, 128, 128), block_size=(1, 32, 32))

Contributing

Contributions are welcome! Please open a pull request with clear changes and add tests when appropriate.

Issues

Found a bug or have a request? Open an issue at https://github.com/Karol-G/med-blosc2/issues.

License

Distributed under the MIT license. See LICENSE for details.