Grand-Challenge DICOM De-Identifier

This Python-based package uses Grand-Challenge De-Identification Procedure to de-identify DICOM files.

It follows the procedure's prescribed actions of keeping, replacing, rejecting, or replacing elements in the files.

Basic Usage

Install via:

$ pip install grand-challenge-dicom-de-identifier

In your Python script then initiate the deidentifier and use either the deidentify_file or deidentify_file methods to process data:

from grand_challenge_dicom_de_identifier.deidentifier import DicomDeidentifier

deidentifier = DicomDeidentifier()

# Deidentify a single file
deidentifier.deidentify_file(
    "input.dcm",
    output="anom/output.dcm"
)

Or by providing a pydicom.Dataset directly:

from grand_challenge_dicom_de_identifier.deidentifier import DicomDeidentifier
import pydicom

deidentifier = DicomDeidentifier()
dataset = pydicom.Dataset()

# Deidentify a pydicom Dataset
deidentifier.deidentify_dataset(
    dataset
)

Advanced Usage

The following arguments can be provided to the deidentifier:

assert_unique_value_for

A collection of element keywords (e.g. ["PatientName"]) that ensures input files all have the same value for these elements. If a file has a different value for any of these elements compared to previous files, a RejectedDICOMFileError is raised. By default no such check is performed.

study_instance_uid_suffix / series_instance_uid_suffix

A specific suffix to overwrite the respective StudyInstanceUID and SeriesInstanceUID with. These will be prefixed with the ROOT uid of Grand Challenge: "1.2.826.0.1.3680043.10.1666.".