Pseudonym Dicom de-identification

Description

This pipeline anonymizes DICOM metadata by replacing personal identifiers with pseudonyms instead of removing them. It ensures that PHI is no longer traceable while maintaining data integrity for longitudinal studies and collaborations.

Obfuscation mode: Removes PII from images and replaces sensitive metadata values (e.g., patient names, IDs) with randomized or pseudonymized data, preserving the overall structure and usability of the metadata.

Predicted Entities

Live Demo Open in Colab

How to use

dicom_df = spark.read.format("binaryFile").load(dicom_path)

pipeline = PretrainedPipeline("dicom_deid_generic_augmented_pseudonym", "en", "clinical/ocr")

result = pipeline.transform(dicom_df).cache()

val dicom_df = spark.read.format("binaryFile").load(dicom_path)

val pipeline = new PretrainedPipeline("dicom_deid_generic_augmented_pseudonym", "en", "clinical/ocr")

val result = pipeline.transform(dicom_df).cache()

Example

Input:

Output:

Model Information