Highdicom

highdicom is a pure Python package built on top of pydicom to provide a higher-level application programming interface (API) for working with DICOM files. Its focus is on common operations required for machine learning, computer vision, and other similar computational analyses. Broadly speaking, the package helps with three types of task:

1. Reading existing DICOM image files of a wide variety of modalities (covering radiology, pathology, and more) and selecting and formatting their frames for computational analysis. This includes considerations such as spatial arrangements of frames and application of pixel transforms, which are not handled by pydicom.

2. Storing image-derived information, for example from computational analyses or human annotation, in derived DICOM objects for communication and storage. This includes:

   • Annotations
   • Parametric Map images
   • Segmentation images
   • Structured Report documents (containing numerical results, qualitative evaluations, and/or vector graphic annotations)
   • Secondary Capture images
   • Key Object Selection documents
   • Legacy Converted Enhanced CT/PET/MR images (e.g., for single frame to multi-frame conversion)
   • Softcopy Presentation State instances (including Grayscale, Color, and Pseudo-Color)

3. Reading existing derived DICOM files of the above types and filtering and accessing the information contained within them.

Documentation

Please refer to the online documentation at highdicom.readthedocs.io, which includes installation instructions, a user guide with examples, a developer guide, and complete documentation of the application programming interface of the highdicom package.

Citation

For more information about the motivation of the library and the design of highdicom's API, please see the following article:

Highdicom: A Python library for standardized encoding of image annotations and machine learning model outputs in pathology and radiology C.P. Bridge, C. Gorman, S. Pieper, S.W. Doyle, J.K. Lennerz, J. Kalpathy-Cramer, D.A. Clunie, A.Y. Fedorov, and M.D. Herrmann. Journal of Digital Imaging, August 2022

If you use highdicom in your research, please cite the above article.

Support

The developers gratefully acknowledge their support:

• The Alliance for Digital Pathology
• The MGH & BWH Center for Clinical Data Science, now MGB AI
• Quantitative Image Informatics for Cancer Research (QIICR)
• Radiomics
• The NCI Imaging Data Commons
• The APOLLO Network

This software is maintained in part by the NCI Imaging Data Commons project, which has been funded in whole or in part with Federal funds from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, under task order no. HHSN26110071 under contract no. HHSN261201500003I. The statements do not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.