Overview
The DicomAnonymizer module within the PYCAD library provides users with the capability to anonymize DICOM files. Anonymization of medical images is a critical process in ensuring the privacy of patient’s personal information, especially when sharing medical data for research or educational purposes. The DicomAnonymizer focuses on customizable anonymization, allowing users to select specific DICOM tags for anonymization according to their requirements or institutional policies.
Features
- Anonymize any specified DICOM tags.
- Simple CLI interface for selecting tags to anonymize.
- Support for batch processing of multiple DICOM files.
- Verification step before the anonymization process is executed.
Prerequisites
- PYCAD library installed.
- Input directory with DICOM files ready for anonymization.
- Output directory where anonymized DICOM files will be saved.
Parameters
input_dir: Directory containing original DICOM files.output_dir: Directory where anonymized DICOM files will be saved.
Usage
Step 1: Importing and Initializing
from pycad.datasets import DicomAnonymizer
input_dir = 'path/to/input/dicom/dir'
output_dir = 'path/to/output/anonymized/dir'
anonymizer = DicomAnonymizer(input_dir, output_dir)
Step 2: Listing Fields
Upon initialization, DicomAnonymizer can provide a list of commonly anonymized DICOM tags.
anonymizer.list_anonymization_fields()Step 3: Running Anonymization
anonymizer.run()When run() is called, the user is prompted to enter the tags they wish to anonymize. After confirmation, the anonymization process begins.
Anonymization Fields
By default, the anonymization process can include the following fields, but it can be customized as needed:
- PatientName
- PatientID
- PatientBirthDate
- PatientSex
- StudyInstanceUID
- SeriesInstanceUID
- StudyID
- InstitutionName
- ReferringPhysicianName
Example
# Create an instance of the DicomAnonymizer
anonymizer = DicomAnonymizer('path/to/dicom_files', 'path/to/save_anonymized_files')
# Print out the fields available for anonymization
anonymizer.list_anonymization_fields()
# Run the anonymization process
anonymizer.run()
Note
Anonymization is irreversible; it is recommended to keep a backup of the original data before proceeding. The responsibility for ensuring that all necessary DICOM tags are anonymized according to relevant regulations and ethical guidelines lies with the user.
License
This DicomAnonymizer module is part of the PYCAD library and is released under the MIT License. For more details, see the LICENSE file.
