Add support for Apparent Diffusion Coefficient images.

[alexzwanenburg]

mirp supports various modalities, including conventional magnetic resonance imaging. Apparent diffusion coefficient (ADC) maps can be computed from diffusion weighted imaging. Implementing import of ADC maps is not trivial, as these are typically stored in one of two DICOM formats: MR and Parametric Maps (PM), see Malyarenko et al.. This requires two new import classes: one for ADC maps stored as DICOM MR, and another for ADC maps stored as DICOM PM. These classes yield the same image class (inheriting from GenericImage).

The following is required:

• [x] Find a (public) dataset that contains ADC maps as an MR file (with the required tags set).
• [ ] Find a (public) dataset that contains ADC maps as an PM file (with the required tags set).
• [ ] Write unit tests to start implementation (in test/dicom_mr_adc_test.py).
• [ ] Implement a new mirp.importData.imageDICOMFileParametericMap.ImageDicomFileParametericMap class to support import of PM images in general. This may also help resolve #33.
  • [ ] Implement a new mirp.importData.imageDICOMFileParametericMap.ImageDicomFileParametericMapADC class that inherits from mirp.importData.imageDICOMFileParametericMap.ImageDicomFileParametericMap.
• [ ] Implement a new mirp.importData.imageDICOMFileMR.ImageDicomFileMRADC class that inherits from mirp.importData.imageDICOMFileMR.ImageDicomFileMR.
• [ ] Implement new mirp.images.mrADCImage.py contents.
• [ ] Add additional unit tests for testing import, reading and processing of ADC files.
  • [ ] test/extract_image_parameters_test.py
  • [ ] test/import_image_and_mask_test.py
  • [ ] test/import_image_test.py
  • [ ] test/read_image_and_mask_test.py
• [ ] Update long-form documentation to mention the support of MR-ADC maps.