imagedata 3.8.14__tar.gz → 3.9.0.dev3__tar.gz

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/CHANGELOG.md

@@ -6,6 +6,60 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 <!--next-version-placeholder-->
+## [v3.9.0-dev3] - 2026-05-21
+### Fixed
+* Series.__getitem__(): Slice the tags even when there is no tag axis.
+* DICOMPLUGIN._join_axes(): Only accept tag axes of length > 1.
+
+## [v3.9.0-dev2] - 2026-05-20
+* Support for python 3.14
+
+## [v3.9.0-rc1] - 2026-05-20
+### Added
+* VariableAxis: Accept list of tuples as values.
+* DICOMPlugin: Implement setting DTI tuple (b, bvector).
+### Fixed
+* Sorting av duplicate tags fixed.
+
+## [v3.9.0-rc0] - 2026-05-19
+### Added
+* DICOMPlugin: Sort DTI series where tags are (b, bvector) tuple.
+### Changed
+* DICOMPlugin: Refactored sorting of n-dimensional datasets.
+
+## [v3.8.16-dev0] - 2026-05-15
+### Changed
+* DICOMPlugin: Refactored sorting of n-dimensional datasets.
+
+## [v3.8.15] - 2026-05-15
+
+## [v3.8.15-dev4] - 2026-05-15
+### Fixed
+* Viewer.connect_draw(): Traversing image tags is done using np.ndindex.
+* Viewer.connect_draw(): Accept legacy ROI where tag is number.
+* Viewer.disconnect_draw(): Traversing image tags is done using np.ndindex.
+
+## [v3.8.15-dev3] - 2026-04-21
+### Added
+* DICOMPlugin: Document data types and processes.
+
+## [v3.8.15-dev2] - 2026-04-14
+### Fixed
+* DICOMPlugin: Corrected writing dtype np.bool_
+* DICOMPlugin.write_slice(): Drop `enforce_file_format` argument to pydicom.save_as().
+
+## [v3.8.15-dev1] - 2026-03-18
+### Added
+* anonymize() methods on Series, Study, Patient and Cohort levels.
+* DICOMPlugin.write_slice(): properly encode date and time data elements when given as string.
+
+## [v3.8.15-dev0] - 2026-03-16
+### Fixed
+* Fixed parsing windows drive letter
+* Header.__set_axes_from_template: Catch missing axes property.
+
+## [v3.8.15-dev0] - 2026-04-14
+* Series ufunc: Fixed problem with np.mean().
 
 ## [v3.8.14] - 2025-11-27
 * Release 3.8.14

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/LICENSE.txt

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2013-2025 Erling Andersen, Haukeland University Hospital, Bergen, Norway
+Copyright (c) 2013-2026 Erling Andersen, Haukeland University Hospital, Bergen, Norway
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: imagedata
-Version: 3.8.14
+Version: 3.9.0.dev3
 Summary: Read/write medical image data
 Author-email: Erling Andersen <Erling.Andersen@Helse-Bergen.NO>
 Maintainer-email: Erling Andersen <Erling.Andersen@Helse-Bergen.NO>
 License: MIT License
         
-        Copyright (c) 2013-2025 Erling Andersen, Haukeland University Hospital, Bergen, Norway
+        Copyright (c) 2013-2026 Erling Andersen, Haukeland University Hospital, Bergen, Norway
         
         Permission is hereby granted, free of charge, to any person obtaining a copy
         of this software and associated documentation files (the "Software"), to deal
@@ -63,6 +63,7 @@ Requires-Dist: imageio
 Requires-Dist: pillow>=10.0.0
 Requires-Dist: xnat<0.7
 Requires-Dist: matplotlib>=3.8.0
+Requires-Dist: dicom-anonymizer
 Requires-Dist: opencv-python-headless
 Dynamic: license-file
 

imagedata-3.9.0.dev3/VERSION.txt

@@ -0,0 +1 @@
+3.9.0-dev3

imagedata-3.9.0.dev3/docs/Anonymization.rst

@@ -0,0 +1,16 @@
+.. _Anonymization:
+
+Anonymization
+===============
+
+DICOM data can be anonymized on the Series, Study, Patient or Cohort level.
+Each object class provides an anonymize() method.
+
+Common parameters include:
+
++---------------+------------------------------------------------------------+
+|uid_table      |Translation table for UIDs                                  |
++---------------+------------------------------------------------------------+
+|kwargs         |dict or parameters giving mapping rules for specific        |
+|               |DICOM attributes                                            |
++---------------+------------------------------------------------------------+

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/docs/Sorting.rst

@@ -36,6 +36,7 @@ Several sort criteria are predefined:
 * triggertime: Sort on Trigger Time
 * b: Sort on MR diffusion b-value
 * bvector: Sort on MR diffusion b vector
+* dti: Sort on MR DTI (b, b vector) tuple
 * fa: Sort on MR Flip Angle
 * te: Sort on MR Echo Time
 
@@ -103,6 +104,25 @@ Trigger Time is the sorting criteria:
         trigger=get_TriggerTime)
     )
 
+MRI DTI sorting
+---------------
+
+MRI diffusion tensor images (DTI) are sorted on diffusion b-value and b-vectors.
+The tags (b, bvector) are tuples.
+
+.. code-block:: python
+
+    img = Series('dti', input_order='dti', input_format='dicom')
+    tags = img.tags[0]
+    for idx in np.ndindex(tags.shape):
+        try:
+            b, bvector = tags[idx]
+        except TypeError:
+            continue
+        dti = img[idx]
+        print(b, bvector, dti.shape)
+
+
 N-dimensional sorting
 ---------------------
 
@@ -115,7 +135,8 @@ A dynamic dual-echo MR acquisition can be sorted on time and echo time into a 5D
 
     img = Series('dyn_dual_echo', input_order='time,te')
 
-In particular, MR RSI diffusion data can be sorted on `b` value and `b` vector:
+In particular, MR RSI diffusion data can be sorted on `b` value and `b` vector
+(remember that the `dti` sort criteria can be used to sort these data as 4D):
 
 .. code-block:: python
 

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/docs/UserGuide.rst

@@ -14,6 +14,7 @@ User Guide
   Colors
   Introduction
   Input_output
+  Anonymization
   Sorting
   DICOMAttributes
   Segmentation

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/docs/conf.py

@@ -21,7 +21,7 @@ sys.path.insert(0, os.path.join(root, 'src'))
 # -- Project information -----------------------------------------------------
 
 project = 'imagedata'
-copyright = '2013-2023, Erling Andersen, Haukeland University Hospital, Bergen, Norway'
+copyright = '2013-2026, Erling Andersen, Haukeland University Hospital, Bergen, Norway'
 author = 'Erling Andersen'
 
 

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/pyproject.toml

@@ -44,6 +44,7 @@ dependencies = [
     "pillow >= 10.0.0",
     "xnat < 0.7",
     "matplotlib >= 3.8.0",
+    "dicom-anonymizer",
     "opencv-python-headless"
 ]
 classifiers = [

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/src/imagedata/__init__.py

@@ -37,24 +37,24 @@ except TypeError:
 if len(_plugins) == 0:
     # Fallback to known built-in plugins
     try:
-        from src.imagedata.archives.filesystemarchive import FilesystemArchive
-        from src.imagedata.archives.zipfilearchive import ZipfileArchive
+        from imagedata.archives.filesystemarchive import FilesystemArchive
+        from imagedata.archives.zipfilearchive import ZipfileArchive
         plugins['archive'] = [
             ('filesystemarchive', FilesystemArchive.name, FilesystemArchive),
             ('zipfilearchive', ZipfileArchive.name, ZipfileArchive)
         ]
-        from src.imagedata.transports.filetransport import FileTransport
-        from src.imagedata.transports.dicomtransport import DicomTransport
-        from src.imagedata.transports.xnattransport import XnatTransport
+        from imagedata.transports.filetransport import FileTransport
+        from imagedata.transports.dicomtransport import DicomTransport
+        from imagedata.transports.xnattransport import XnatTransport
         plugins['transport'] = [
             ('filetransport', FileTransport.name, FileTransport),
             ('dicomtransport', DicomTransport.name, DicomTransport),
             ('xnattransport', XnatTransport.name, XnatTransport)
         ]
-        from src.imagedata.formats.dicomplugin import DICOMPlugin
-        from src.imagedata.formats.itkplugin import ITKPlugin
-        from src.imagedata.formats.matplugin import MatPlugin
-        from src.imagedata.formats.niftiplugin import NiftiPlugin
+        from imagedata.formats.dicomplugin import DICOMPlugin
+        from imagedata.formats.itkplugin import ITKPlugin
+        from imagedata.formats.matplugin import MatPlugin
+        from imagedata.formats.niftiplugin import NiftiPlugin
         plugins['format'] = [
             ('dicomformat', DICOMPlugin.name, DICOMPlugin),
             ('itkformat', ITKPlugin.name, ITKPlugin),

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/src/imagedata/archives/__init__.py

@@ -3,12 +3,10 @@
 Standard plugins provide support for local filesystem and zip archives.
 """
 
-# Copyright (c) 2018-2024 Erling Andersen, Haukeland University Hospital, Bergen, Norway
+# Copyright (c) 2018-2026 Erling Andersen, Haukeland University Hospital, Bergen, Norway
 
-import os.path
 import logging
 import urllib.parse
-import fnmatch
 from ..transports import Transport
 
 logger = logging.getLogger(__name__)
@@ -40,19 +38,8 @@ def find_mimetype_plugin(mimetype, url, mode="r", read_directory_only=False, opt
     if opts is None:
         opts = {}
     from .. import plugins
-    if os.name == 'nt' and \
-            fnmatch.fnmatch(url, '[A-Za-z]:\\*'):
-        # Windows: Parse without /x:, then re-attach drive letter
-        urldict = urllib.parse.urlsplit(url[2:], scheme="file")
-        # _path = url[:2] + urldict.path
-    else:
-        urldict = urllib.parse.urlsplit(url, scheme="file")
-        # _path = urldict.path if len(urldict.path) > 0 else urldict.netloc
-    # if urldict.scheme == 'xnat':
-    #     mimetype = 'application/zip'
-    # if mimetype is None:
-    #     logger.debug("imagedata.archives.find_mimetype_plugin: filesystem")
-    #     return find_plugin('filesystem', url, mode, opts=opts)
+    # Windows: Drive letter will be parsed as scheme
+    urldict = urllib.parse.urlsplit(url, scheme="file")
     transport = None
     if urldict.scheme:
         transport = Transport(
@@ -74,8 +61,6 @@ def find_mimetype_plugin(mimetype, url, mode="r", read_directory_only=False, opt
             logger.debug("{}: {}, mode: {}".format(
                 _name, ptype, mode))
             return pclass(url=url, transport=transport, mode=mode, opts=opts)
-    # if os.path.isfile(_path):
-    # if os.path.exists(_path):
     if urldict.scheme == "file":
         logger.debug("{}: filesystem".format(_name))
         try:

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/src/imagedata/axis.py

@@ -258,7 +258,11 @@ class VariableAxis(Axis):
 
     def __init__(self, name: str, values: Sequence) -> None:
         super(VariableAxis, self).__init__(name)
-        if issubclass(type(values[0]), Sequence):
+        if issubclass(type(values[0]), tuple):
+            self._values = np.ndarray(len(values), dtype=tuple)
+            for i, value in enumerate(values):
+                self._values[i] = value
+        elif issubclass(type(values[0]), Sequence):
             self._values = np.ndarray(len(values), dtype=np.ndarray)
             for i, value in enumerate(values):
                 self._values[i] = np.array(value)

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/src/imagedata/collection.py

@@ -8,17 +8,20 @@ The Cohort class is a collection of Patient objects.
 
 """
 
-# Copyright (c) 2023-2025 Erling Andersen, Haukeland University Hospital, Bergen, Norway
+# Copyright (c) 2023-2026 Erling Andersen, Haukeland University Hospital, Bergen, Norway
 
 import logging
 from datetime import datetime, date, time
 import argparse
 from pathlib import Path
+from pydicom.uid import UID
 from typing import Union
+# from dicomanonymizer.dicomfields_selector import dicom_anonymization_database_selector
 
 from .series import Series
 from .readdata import read as r_read
-from .formats import UnknownInputError
+from .formats import UnknownInputError, get_uid
+from .formats.dicomlib import anonymization_rules
 
 logger = logging.getLogger(__name__)
 
@@ -309,7 +312,10 @@ class Study(IndexedDict):
                         else:
                             _value = datetime.strptime(_value, "%H%M%S").time()
                     except ValueError:
-                        _value = datetime.time(_value)
+                        try:
+                            _value = datetime.time(_value)
+                        except TypeError:
+                            pass
                 # Update self property if None from series
                 if getattr(self, _attr, None) is None:
                     # _series = self[_seriesInstanceUID]
@@ -349,6 +355,7 @@ class Study(IndexedDict):
 
     def __init__(self, data, opts: dict = None, **kwargs):
         super(Study, self).__init__()
+        self.__uid_generator = get_uid()
         for _attr in self._attributes:
             setattr(self, _attr, None)
 
@@ -374,6 +381,8 @@ class Study(IndexedDict):
             _series_dict = data
         elif issubclass(type(data), Study):
             raise ValueError("Why here?")
+        elif data is None:
+            _series_dict = {}
         else:
             # Assume data is URL
             try:
@@ -396,7 +405,12 @@ class Study(IndexedDict):
         return \
             "Study: {} {}".format(datetime.combine(_date, _time), _descr)
 
-    def write(self, url, opts=None, formats=None):
+    def new_uid(self) -> UID:
+        """Return the next available UID from the UID generator.
+        """
+        return self.__uid_generator.__next__()
+
+    def write(self, url, opts=None, formats=None, **kwargs):
         """Write image data, calling appropriate format plugins
 
         Args:
@@ -412,6 +426,13 @@ class Study(IndexedDict):
             imagedata.formats.WriteNotImplemented: Cannot write this image format.
         """
 
+        if opts is None:
+            opts = {}
+        elif issubclass(type(opts), argparse.Namespace):
+            opts = vars(opts)
+        for key, value in kwargs.items():
+            opts[key] = value
+
         _used_urls = []
         for _seriesUID in self.keys():
             _series = self[_seriesUID]
@@ -431,6 +452,28 @@ class Study(IndexedDict):
             except Exception as e:
                 raise Exception(_url) from e
 
+    def anonymize(self, uid_table: dict = {}, actions: dict = {}, **kwargs):
+        _actions = {} | actions
+        rules = anonymization_rules | kwargs
+        if rules['studyTime'] is None:
+            rules['studyTime'] = self.studyTime
+        _copy = Study(None)
+        for _rule in rules:
+            try:
+                _ = getattr(_copy, _rule)
+                setattr(_copy, _rule, rules[_rule])
+            except AttributeError:
+                pass
+        if self.studyInstanceUID is None:
+            self.studyInstanceUID = _copy.new_uid()
+        if self.studyInstanceUID not in uid_table:
+            uid_table[self.studyInstanceUID] = _copy.new_uid()
+        _copy.studyInstanceUID = uid_table[self.studyInstanceUID]
+        for _seriesUID in self.keys():
+            _series = self[_seriesUID].anonymize(uid_table, **rules)
+            _copy[_series.seriesInstanceUID] = _series
+        return _copy
+
 
 class Patient(IndexedDict):
     """Patient -- Read and sort images into a collection of Study objects.
@@ -500,6 +543,9 @@ class Patient(IndexedDict):
             # _study_dict = data
             # for _studyInstanceUID in data:
             #     self[_studyInstanceUID] = _study_dict[_studyInstanceUID]
+        elif data is None:
+            _series_dict = {}
+            _study_dict = {}
         else:
             _series_dict = _sort_in_series(data, _in_opts)
             _study_dict = _sort_in_studies(_series_dict, _in_opts)
@@ -618,6 +664,21 @@ class Patient(IndexedDict):
                 except Exception as e:
                     raise Exception(_url) from e
 
+    def anonymize(self, uid_table: dict = {}, actions: dict = {}, **kwargs):
+        _actions = {} | actions
+        rules = anonymization_rules | kwargs
+        _copy = Patient(None)
+        for _rule in rules:
+            try:
+                _ = getattr(_copy, _rule)
+                setattr(_copy, _rule, rules[_rule])
+            except AttributeError:
+                pass
+        for _studyUID in self.keys():
+            _study = self[_studyUID].anonymize(uid_table, **rules)
+            _copy[_study.studyInstanceUID] = _study
+        return _copy
+
 
 class Cohort(IndexedDict):
     """Cohort -- Read and sort images into a collection of Patient objects.
@@ -687,6 +748,10 @@ class Cohort(IndexedDict):
             for _patientID in _patient_dict:
                 # self[_patientID] = Patient(_patient_dict[_patientID])
                 self[_patientID] = _patient_dict[_patientID]
+        elif data is None:
+            _series_dict = {}
+            _study_dict = {}
+            _patient_dict = {}
         else:
             raise ValueError('Unexpected cohort data type {}'.format(type(data)))
 
@@ -782,3 +847,19 @@ class Cohort(IndexedDict):
                         _series.write(_url, opts=opts, formats=formats)
                     except Exception as e:
                         raise Exception(_url) from e
+
+    def anonymize(self, uid_table: dict = {}, actions: dict = {}, **kwargs):
+        _actions = {} | actions
+        rules = anonymization_rules | kwargs
+        _copy = Cohort(None)
+        for _rule in rules:
+            try:
+                _ = getattr(_copy, _rule)
+                setattr(_copy, _rule, rules[_rule])
+            except AttributeError:
+                pass
+        for _patientID in self.keys():
+            _patient = self[_patientID].anonymize(uid_table, **rules)
+            _copy[_patientID] = _patient
+        return _copy
+

{imagedata-3.8.14 → imagedata-3.9.0.dev3}/src/imagedata/formats/__init__.py

@@ -3,7 +3,7 @@
 Standard plugins provides support for DICOM and Nifti image file formats.
 """
 
-# Copyright (c) 2013-2024 Erling Andersen, Haukeland University Hospital, Bergen, Norway
+# Copyright (c) 2013-2026 Erling Andersen, Haukeland University Hospital, Bergen, Norway
 
 import os
 import sys
@@ -23,11 +23,13 @@ INPUT_ORDER_TIME = 'time'
 INPUT_ORDER_TRIGGERTIME = 'triggertime'
 INPUT_ORDER_B = 'b'
 INPUT_ORDER_BVECTOR = 'bvector'
+INPUT_ORDER_DTI = 'dti'
 INPUT_ORDER_RSI = 'rsi'
 INPUT_ORDER_FA = 'fa'
 INPUT_ORDER_TE = 'te'
 INPUT_ORDER_FAULTY = 'faulty'
 input_order_set = {INPUT_ORDER_NONE, INPUT_ORDER_TIME, INPUT_ORDER_B, INPUT_ORDER_BVECTOR,
+                   INPUT_ORDER_DTI,
                    INPUT_ORDER_RSI, INPUT_ORDER_FA, INPUT_ORDER_TRIGGERTIME,
                    INPUT_ORDER_TE, INPUT_ORDER_FAULTY}
 

imagedata-3.9.0.dev3/src/imagedata/formats/dicomlib/__init__.py

@@ -0,0 +1,17 @@
+anonymization_rules = {
+    'patientName': 'ANONYMOUS',
+    'patientID': 'ANONYMOUS',
+    'patientBirthDate': '19000101',
+    'patientSex': 'O',
+    'patientIdentityRemoved': 'YES',
+    'deidentificationMethod': 'imagedata',
+    'studyDate': '19700101',
+    'studyTime': None,
+    'studyDescription': '',
+    'studyID': '1',
+    'referringPhysicianName': '',
+    'instanceCreationDate': '19700101',
+    'seriesDate': '19700101',
+    'acquisitionDate': '19700101',
+    'performedProcedureStepStartDate': '19700101',
+}