pfdicom_tagSub

Table of Contents

• Quick Overview
• Overview
• Installation
  • Dependencies
  • Using PyPI
• Command line arguments
• Examples
• NOTE:

Quick Overview

• pfdicom_tagSub reads/edits/saves DICOM meta information. It can be used to anonymize DICOM header data.

Overview

pfdicom_tagSub replaces a set of <tag, value> pairs in a DICOM header with values passed in a JSON structure. Individual DICOM tags can be explicitly referenced in the JSON structure, as well as a regular expression construct to capture all tags satisfying that expression (allowing for idiomatic bulk substitution of <tag, value> pairs).

Tag regular expression constructs are python string expressions and are prefixed by "re:<pythonRegex>". For example, "re:.*hysician" will perform some substitution on all tags that contain the letters hysician. The value substitution has access to a special lookup, #tag, which is the current tag hit. It is possible to apply built in functions to the tag hit, for example md5 hashing, using "%_md5|4_#tag",

{
    "re:.*hysician":                "%_md5|4_#tag"
}

will be expanded to

{
    "PerformingPhysiciansName" :    "%_md5|4_PerformingPhysiciansName"
    "PhysicianofRecord"        :    "%_md5|4_PhysicianofRecord"
    "ReferringPhysiciansName"  :    "%_md5|4_ReferringPhysiciansName"
    "RequestingPhysician"      :    "%_md5|4_RequestingPhysician"
}

The tag regular expression construct allows for simple and powerful bulk substition of <tag, value> pairs.

The script accepts an <inputDir>, and then from this point an os.walk() is performed to extract all the subdirs. Each subdir is examined for DICOM files (in the simplest sense by a file extension mapping) are passed to a processing method that reads and replaces specified DICOM tags, saving the result in a corresponding directory and filename in the output tree.

Installation

Dependencies

The following dependencies are installed on your host system/python3 virtual env (they will also be automatically installed if pulled from pypi):

• pfmisc (various misc modules and classes for the pf* family of objects)
• pftree (create a dictionary representation of a filesystem hierarchy)
• pfdicom (handle underlying DICOM file reading)

Using PyPI

The best method of installing this script and all of its dependencies is by fetching it from PyPI

pip3 install pfdicom_tagSub

Command line arguments

[--inputDir <inputDir>
Input DICOM directory to examine. By default, the first file in this
directory is examined for its tag information. There is an implicit
assumption that each <inputDir> contains a single DICOM series.

[--inputFile <inputFile>]
An optional <inputFile> specified relative to the <inputDir>. If
specified, then do not perform a directory walk, but convert only
this file.

[--fileFilter <fileFilter>]
An optional extension to filter the DICOM files of interest from the
<inputDir>.

[--outputDir <outputDir>
The output root directory that will contain a tree structure identical
to the input directory, and each "leaf" node will contain the analysis
results.

[--outputLeafDir <outputLeafDirFormat>]
If specified, will apply the <outputLeafDirFormat> to the output
directories containing data. This is useful to blanket describe
final output directories with some descriptive text, such as
'anon' or 'preview'.

This is a formatting spec, so

    --outputLeafDir 'preview-%s'

where %%s is the original leaf directory node, will prefix each
final directory containing output with the text 'preview-' which
can be useful in describing some features of the output set.

[--tagFile <JSONtagFile>]
Parse the tags and their "subs" from a JSON formatted <JSONtagFile>.

[--tagStruct <JSONtagStructure>]
Parse the tags and their "subs" from a JSON formatted <JSONtagStucture>
string passed directly in the command line. Note that sometimes protecting
a JSON string can be tricky, especially when used in scripts or as variable
expansions. If the JSON string is problematic, use the [--tagInfo <string>]
instead.

[--tagInfo <delimited_parameters>]
A token delimited string that is reconstructed into a JSON structure by the
script. This is often useful if the [--tagStruict] JSON string is hard to
parse in scripts and variable passing within scripts. The format of this
string is:

        "<tag1><splitKeyValue><value1><split_token><tag2><splitKeyValue><value2>"

for example:

        --splitToken ","
        --splitKeyValue ':'
        --tagInfo "PatientName:anon,PatientID:%_md5|7_PatientID"

or more complexly (esp if the ':' is part of the key):

        --splitToken "++"
        --splitKeyValue "="
        --tagInfo "PatientBirthDate = %_strmsk|******01_PatientBirthDate ++
                   re:.*hysician"   = %_md5|4_#tag"


[--splitToken <split_token>]
The token on which to split the <delimited_parameters> string.
Default is '++'.

[--splitKeyValue <keyValueSplit>]
The token on which to split the <key> <value> pair. Default is ':'
but this can be problematic if the <key> itself has a ':' (for example
in the regular expression expansion).

[--outputFileStem <outputFileStem>]
The output file stem to store data. This should *not* have a file
extension, or rather, any "." chars. Dots in the name are considered
part of the stem and are *not* considered extensions.

[--removePrivateTags]
If specified, remove all the private tag elements from the input DICOMs

[--threads <numThreads>]
If specified, break the innermost analysis loop into <numThreads>
threads.

[--man]
Show full help.

[--synopsis]
Show brief help.

[--json]
If specified, output a JSON dump of final return.

[--followLinks]
If specified, follow symbolic links.

[--verbosity <level>]
Set the app verbosity level.

    0: No internal output;
    1: Run start / stop output notification;
    2: As with level '1' but with simpleProgress bar in 'pftree';
    3: As with level '2' but with list of input dirs/files in 'pftree';
    5: As with level '3' but with explicit file logging for
            - read
            - analyze
            - write

Examples

Perform a DICOM anonymization by processing specific tags:

pfdicom_tagSub                                      \
    --fileFilter dcm                                \
    --inputDir /var/www/html/normsmall              \
    --outputDir /var/www/html/anon                  \
    --tagStruct '
    {
        "PatientName":              "%_name|patientID_PatientName",
        "PatientID":                "%_md5|7_PatientID",
        "AccessionNumber":          "%_md5|8_AccessionNumber",
        "PatientBirthDate":         "%_strmsk|******01_PatientBirthDate",
        "re:.*hysician":            "%_md5|4_#tag",
        "re:.*stitution":           "#tag",
        "re:.*ddress":              "#tag"
    }
    ' --threads 0 --printElapsedTime

-- OR equivalently --

pfdicom_tagSub                                      \
    --fileFilter dcm                                \
    --inputDir /var/www/html/normsmall              \
    --outputDir /var/www/html/anon                  \
    --splitToken ","                                \
    --splitKeyValue "="                             \
    --tagInfo '
        PatientName         =  %_name|patientID_PatientName,
        PatientID           =  %_md5|7_PatientID,
        AccessionNumber     =  %_md5|8_AccessionNumber,
        PatientBirthDate    =  %_strmsk|******01_PatientBirthDate,
        re:.*hysician       =  %_md5|4_#tag,
        re:.*stitution      =  #tag,
        re:.*ddress         =  #tag
    ' --threads 0 --printElapsedTime

will replace the explicitly named tags as shown:

• the PatientName value will be replaced with a Fake Name, seeded on the PatientID;
• the PatientID value will be replaced with the first 7 characters of an md5 hash of the PatientID;
• the AccessionNumber value will be replaced with the first 8 characters of an md5 hash of the AccessionNumber;
• the PatientBirthDate value will set the final two characters, i.e. the day of birth, to 01 and preserve the other birthdate values;
• any tags with the substring hysician will have their values replaced with the first 4 characters of the corresponding tag value md5 hash;
• any tags with stitution and ddress substrings in the tag contents will have the corresponding value simply set to the tag name.

NOTE:

Spelling matters! Especially with the substring bulk replace, please make sure that the substring has no typos, otherwise the target tags will most probably not be processed.

_-30-_