Package for extracting data from DICOM RDSR files
pip install rdsr-navigator==0.3.1
A package for extracting data from DICOM RDSR files. The focus of this project is to extract data. It is not possible to modify files. The project is currently in early development and things might not work exactly how you expect. The public api is not yet stable, so please do not use this for any critical projects.
These instructions will get you a copy of the project up and running on your local machine.
RDSR Navigator is written in Python 3 and uses pydicom. Therefore make sure you are running Python 3 and make sure pydicom is installed. If pydicom is not installed, use the command below to install pydicom.
$ pip install pydicom
RDSR Navigator is available on pypi and can be installed using the following command.
$ pip install rdsr_navigator
Now, you are ready to start using RDSR navigator.
To open an RDSR file, type the following.
>>> import rdsr_navigator as nav >>> rdsr_obj = nav.read_file('C:\rdsr_file.dcm')
The input argument to
read_file is a
str containing the path to an RDSR file. Other supported data types are
Data is extracted by first navigating through the RDSR hierarchy using the concept names. All concept names are given in lower case, separated by underscores ("_"). In the example below we are extracting the value from "Procedure Reported".
>>> rdsr_obj['procedure_reported'].value 'Projection X-Ray'
When the square brackets are used, the first matched concept name is returned. This is inappropriate if several entries with the same concept name exist on the same level. To iterate entries with the same concept name, use the
>>> for irr_event in sr_obj.get_all('irradiation_event_x-ray_data')): print(irr_event['dose_area_product'].value) [(1.9632189e-7, 'Gy.m2')] [(1.1173212e-5, 'Gy.m2')] [(8.566802e-7, 'Gy.m2')]
Drill deeper into the hierarchy by adding more concept names in the square brackets.
>>> rdsr_obj['procedure_reported', 'has_intent'].value 'Combined Diagnostic and Therapeutic Procedure'
The RDSR navigator classes are integrated with jupyter notebooks.
Currently, only a few value types are supported. The supported value types are listed below.
This project is licensed under the MIT License - see the LICENSE file for details.