Simple script to anonymize dicoms in parallel

anonymize_dicom.py

import argparse
import logging
import os
import time
from concurrent.futures import ThreadPoolExecutor, ProcessPoolExecutor
from datetime import timedelta
from functools import partial
from pathlib import Path

import pydicom
import warnings

warnings.filterwarnings("ignore", category=UserWarning)

logger = logging.getLogger("anonymizer")


def parse_args() -> argparse.Namespace:
    parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
    parser.add_argument(
        "input_dicom_dir",
        type=Path,
        help="path to the directory with dicom files",
    )

    parser.add_argument(
        "output_dicom_dir",
        type=Path,
        help="path to the directory fro anonymized files",
    )

    parser.add_argument("-v", "--verbose", help="increase output verbosity", action="store_true")
    parser.add_argument("-c", "--concurrency", help="number of processes to run in parallel", type=int, default=None)

    return parser.parse_args()


def configure_logging(is_verbose: bool = False):
    handler = logging.StreamHandler()
    handler.setFormatter(logging.Formatter("%(asctime)-15s %(levelname)-8s %(message)s"))
    logger.addHandler(handler)
    logger.setLevel(level=logging.DEBUG if is_verbose else logging.INFO)


def scan_input_dir_or_path(path_to_dicoms: Path) -> list[Path]:
    dicom_paths: list[Path] = []

    if path_to_dicoms.is_dir():
        logger.info("Scanning dicoms in %s", path_to_dicoms)
        for dir_path, _, filenames in os.walk(path_to_dicoms):
            for filename in filenames:
                if filename.endswith(".dcm"):
                    dicom_paths.append(Path(dir_path) / filename)
    else:
        if path_to_dicoms.suffix == ".dcm":
            dicom_paths.append(path_to_dicoms)

    logger.info("Found %s files", len(dicom_paths))
    return dicom_paths


def _anonymize(dataset):
    dataset.PatientID = "ANON"
    dataset.PatientName = "ANON"
    dataset.PatientBirthDate = "19000101"

    if "OtherPatientIDs" in dataset:
        del dataset.OtherPatientIDs

    if "OtherPatientIDsSequence" in dataset:
        del dataset.OtherPatientIDsSequence


def anonymize(filename: Path, input_dicom_dir: Path, output_dicom_dir: Path):
    dataset = pydicom.dcmread(filename)

    _anonymize(dataset)

    output_filename = output_dicom_dir / filename.relative_to(input_dicom_dir)
    Path(output_filename.parent).mkdir(parents=True, exist_ok=True)
    dataset.save_as(output_filename)


def main(args: argparse.Namespace) -> None:
    tic = time.monotonic()

    input_dicom_dir = args.input_dicom_dir
    output_dicom_dir = args.output_dicom_dir

    dicom_paths = scan_input_dir_or_path(input_dicom_dir)

    with ProcessPoolExecutor(args.concurrency) as executor:
        executor.map(
            partial(anonymize, input_dicom_dir=input_dicom_dir, output_dicom_dir=output_dicom_dir), dicom_paths
        )
    toc = time.monotonic()
    logger.info("Anonymized dicoms saved to %s", output_dicom_dir)
    logger.debug("Anonymization of %s files finished at %s", len(dicom_paths), timedelta(seconds=toc - tic))


if __name__ == "__main__":
    args = parse_args()
    configure_logging(args.verbose)
    main(args)