TidyDicom

TidyDicom is a simple program based on Python . It can categorize all DICOM files in the root directory into specific folders.

Welcome any issues and PR.

Release

The Windows version based on embeddable Python is uploaded. It can run without python pre-installed.

Requirement

The below modules must be installed first to make the TidyDicom work.

- pydicom=2.4.3
- tqdm=4.66.4

Quick Start

Suppose that we have a root folder full of DICOM files and we need to categorize these DICOM files into specific folders. For example, if we want the first layer of the directory tree to be the patient's name and patient's ID, the second layer to be the study date, and the second layer to be the series description. DICOM files with the same patient's name, patient's ID, study date, and series description will go to the same folder.

customize the code

Open the custom.py file:

Firstly, we need to tell the program which DICOM data element to read to get the informations.

INFO_DICT = {
    'PatientName':{
        'standard_name':'PatientName', 
        'default_value':'NoPatientNameTag'
        },
    'PatientID':{
        'standard_name':'PatientID', 
        'default_value':'NoPatientIDTag'
        },
    'StudyDate':{
        'group_id':'0008', 
        'element_id':'0020', 
        'default_value':'NoStudyDateTag'
        },
    'SeriesDescription':{
        'group_id':'0008', 
        'element_id':'103E', 
        'default_value':'NoSeriesDescriptionTag'
        }
    }

We defined a Python dictionary in which the keyword is a string of anything and the value is used to get the tag information.

Standard DICOM data elements can be accessed by standard keyword like PatientName for (0010,0010) Patient's Name and PatientID for (0010,0020) | Patient ID . It can also be accessed by group number and element number like group_id=0008, element_id=0020 for (0008,0020) Study Date and group_id=0008, element_id=103E for (0008,103E) Series Description .

Private DICOM data elements can only be accessed by group number and element number.

A list of keywords for all standard elements can be found here.

The program uses the INFO_DICT and returns a Python dictionary info containing the element informations.

Secondly, we need to tell the program what is the rule to categorize the DICOM files.

def GetPathByInfo(info:Dict[str, Any]) -> str:
    pid = info['PatientID'] # no change

    pname = info['PatientName'].components # PearsonName object, return tuple
    pname = ''.join(pname).lower().replace(' ', '')

    date = info['StudyDate'] # no change

    series_description = info['SeriesDescription'] # no change

    return os.path.join(id+'_'+pname, date, series_name) # using os.path.join is a safe method for all OS.

Suppose a DICOM file is named 1234.dcm and the program read it and returns the info = {'PatientName':'Jackson', 'PatientID':'111111', 'StudyDate':20211202, 'SeriesDescription':'arterial'}. The function above will return '111111_Jackson/20211202/arterial'. Finally the file will be move to './111111_Jackson/20211202/arterial/1234.dcm'.

Set the arguments

Open the config.json file:

we need to tell the program where is the root folder and where we want to move the DICOM files to.

{
    "in_dir": "C:\\Users\\J\\Desktop\\mess",
    "out_dir": "C:\\Users\\J\\Desktop\\clean",
    "copy_or_cut": "copy", 
    "custom_file":"custom", 
    "batch_size":2000
}

The in_dir is the root folder of your unsorted DICOM files. The out_dir is the root folder of your sorted DICOM files. copy_or_cut defines the method when moving the files, by either shutil.copy or shutil.move. custom_file is the filename of the custom.py. You can write different custom files (e.g., custom_a for custom_a.py, custom_b for custom_b.py) to handle different DICOMs. batch_size is the number of files processed in one batch.

Run

Run the main.py and the program will start to work.

The terminal will display the process bar batch by batch.

When any error occurs, a failures.csv will be stored in out_dir logging the filepath and error information.

Embeddable Version

download TidyDicom.zip and unzip it.
modify custom.py and config.json.
double click run.bat

License

This project is licensed under the GPL 3.0 License

DOCT-Y/TidyDicom