This release is a pre-release and may not be stable for production use.
imagedata
Imagedata is a python library to read and write medical image data into numpy arrays. Imagedata will handle multi-dimensional data. In particular, imagedata will read and sort DICOM 3D and 4D series based on defined tags. Imagedata will handle geometry information between the formats.
The following formats are included:
DICOM
NIfTI
ITK (MetaIO)
Matlab
PDF (write only)
Non-medical image formats through the Pillow imaging library
Other formats can be added through a plugin architecture.
Install
pip install imagedataDocumentation
See the Documentation page for info.
Example code
A simple example reading two time series from dirA and dirB, and writing their mean to dirMean:
from imagedata import Series
a = Series('dirA', 'time')
b = Series('dirB', 'time')
assert a.shape == b.shape, "Shape of a and b differ"
# Notice how a and b are treated as numpy arrays
c = (a + b) / 2
c.write('dirMean')Sorting
Sorting of DICOM slices is considered a major task. Imagedata will sort slices into volumes based on slice location. Volumes may be sorted on a number of DICOM tags:
‘time’: Dynamic time series, sorted on acquisition time
‘b’: Diffusion weighted series, sorted on diffusion b value
‘fa’: Flip angle series, sorted on flip angle
‘te’: Sort on echo time TE
In addition, volumes can be sorted on user defined tags.
Non-DICOM formats usually don’t specify the labelling of the 4D data. In this case, you can specify the sorting manually.
Viewing
A simple viewer. Scroll through the image stack, step through the tags of a 4D dataset. These operations are possible:
Read-out voxel value: Move mouse over.
Window/level adjustment: Move mouse with left key pressed.
Scroll through slices of an image stack: Mouse scroll wheel, or up/down array keys.
Step through tags (time, b-values, etc.): Left/right array keys.
Move through series when many series are displayed: PageUp/PageDown keys.
# View a Series instance
a.show()
# View both a and b Series
a.show(b)
# View several Series
a.show([b, c, d])Converting data from DICOM and back
In many situations you need to process patient data using a tool that do not accept DICOM data. In order to maintain the coupling to patient data, you may convert your data to e.g. Nifti and back.
Example using the command line utility image_data:
image_data --of nifti niftiDir dicomDir
# Now do your processing on Nifti data in niftiDir/, leaving the result in niftiResult/.
# Convert the niftiResult back to DICOM, using dicomDir as a template
image_data --of dicom --template dicomDir dicomResult niftiResult
# The resulting dicomResult will be a new DICOM series that could be added to a PACS
# Set series number and series description before transmitting to PACS using DICOM transport
image_data --sernum 1004 --serdes 'Processed data' \
dicom://server:104/AETITLE dicomResultThe same example using python code:
from imagedata import Series
a = Series('dicomDir')
a.write('niftiDir', formats=['nifti']) # Explicitly select nifti as output format
# Now do your processing on Nifti data in niftiDir/, leaving the result in niftiResult/.
b = Series('niftiResult', template=a) # Or template='dicomDir'
b.write('dicomResult') # Here, DICOM is default output format
# Set series number and series description before transmitting to PACS using DICOM transport
b.seriesNumber = 1004
b.seriesDescription = 'Processed data'
b.write(' dicom://server:104/AETITLE')Series fields
The Series object is inherited from numpy.ndarray, adding a number of useful fields:
- Axes
a.axes defines the unit and size of each dimension of the matrix
- Addressing
4D: a[tags, slices, rows, columns]
3D: a[slices, rows, columns]
2D: a[rows, columns]
RGB: a[…, rgb]
- patientID, patientName, patientBirthDate
Identifies patient
- accessionNumber
Identifies study
- seriesNumber, seriesDescription, imageType
Labels DICOM data
- slices
Returns number of slices
- spacing
Returns spacing for each dimension. Units depend on dimension, and could e.g. be mm or sec.
- tags
Returns tags for each slice
- timeline
Returns time steps for when a time series
- transformationMatrix
The transformation matrix to calculate physical coordinates from pixel coordinates
Series instancing
From image data file(s):
a = Series('in_dir')From a list of directories:
a = Series(['1', '2', '3'])From a numpy array:
e = np.eye(128)
a = Series(e)Series methods
- write()
Write the image data as a Matlab file to out_dir:
a.write('out_dir', formats=['mat'])- slicing
The image data array can be sliced like numpy.ndarray. The axes will be adjusted accordingly. This will give a 3D b image when a is 4D.
b = a[0, ...]Archives
The Series object can access image data in a number of archives. Some archives are:
- Filesystem
Access files in directories on the local file system.
a = Series('in_dir')- Zip
Access files inside zip files.
# Read all files inside file.zip:
a = Series('file.zip')
# Read named directory inside file.zip:
b = Series('file.zip?dir_a')
# Write the image data to DICOM files inside newfile.zip:
b.write('newfile.zip', formats=['dicom'])Transports
- file
Access local files (default):
a = Series('file:in_dir')- dicom
Access files using DICOM Storage protocols. Currently, writing (implies sending) DICOM images only:
a.write('dicom://server:104/AETITLE')Command line usage
The command line program image_data can be used to convert between various image data formats:
image_data --order time out_dir in_dirs1 maintainer
Metadata
1 maintainer
Release files for imagedata 3.10.0rc2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| imagedata-3.10.0rc2.tar.gz | 150.8 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| imagedata-3.10.0rc2-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 325.5 kB
Release files / imagedata-3.10.0rc2.tar.gz
| Download URL | imagedata-3.10.0rc2.tar.gz |
|---|---|
| Size | 150.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
74234000cddba8b37d7a84b5ffac752f66d77d00ff65bc763a91d51f4f88478a
|
|
BLAKE2b-256 checksum How to use checksums |
df9a35a6c55051980cfc6291a0199a8affd6143c298d69d95986cb87f1ae3e5e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 21, 2026.
Transparency logRelease files / imagedata-3.10.0rc2-py3-none-any.whl
| Download URL | imagedata-3.10.0rc2-py3-none-any.whl |
|---|---|
| Size | 174.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
53e64eef60c4a5a587e53f52ec0be3cfd2f3ad867a6e4c3acbb52ff759b195f6
|
|
BLAKE2b-256 checksum How to use checksums |
fecd3ec27d50cc2002e7df2cb1f82f9ff6fb198a20992f71097573f9ff5b522c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 21, 2026.
Transparency log
