Welcome to pediatricIQphantom documentation!¶
This documentation provides information regarding how to download, install, and use the pediatricIQphantoms tools which are designed to simulate computed tomography (CT) image characteristics in a range of pediatric sizes ranging from newborn through adolescence.
Introduction¶
Digital Pediatric Image Quality Phantoms for Evaluating CT Denoising Methods are a set of digital phantoms and simulation methods for generating CT images of standard image quality (IQ) phantoms designed to match the effective diameter of pediatric patients ranging from newborns to teenagers. This repository has tools for generating MITA-LCD phantom and a multi-contrast sensitometry module similar to the CTP404 module of the Catphan 600 phantom. Functions are also provided to simulate different acquisition parameters and CT scanner models.
Size is one of the most important patient factors influencing CT performance as it determines the overall x-ray attenuation and noise properties. New deep learning-based denoisers have shown potential to improve image quality for a fixed radiation dose or maintain image quality while reducing dose Brady 2023. Performance Assessment consists of analytical quality assurance phantom models and interfaces to CT simulation frameworks to generate simulated CT images representing different diameters of each phantom.
Installation¶
Installation is only required to generate new datasets, a pregenerated dataset can be downloaded from Zenodo, only proceed if you want to generate new simulated datasets.
Requirements
Mac, Linux, or Windows Subsystem for Linux (WSL) operating systems described on the Octave Conda Forge page. This package currently uses the Octave-based Michigan Image Reconstruction Toolbox (MIRT)
Installation
git clone https://github.com/DIDSR/pediatricIQphantoms
cd pediatricIQphantoms
conda env create --file environment.yml
conda activate pediatricIQphantoms
The code block above does the following in 4 lines:
Git clones the pediatricIQphantoms repository
Changes the active directory to the repo
Creates a new conda environment called “pediatricIQphantoms”
Activates the conda environment. This makes the phantom creation library pediatricIQphantoms accessible in scripts (see examples) and via command line calls (see demo 01 and demo 02).
Test the Installation
pytest
This runs the unit tests to verify that installation was successful.
Running Notebooks
Example computational notebooks have been provided to demonstrate
To run these Notebooks you will need to have jupyter installed
conda install jupyterlab -y
Running Shell Scripts
Example shell scripts and config files have been provided to demonstrate non-intertactive use of the tool to generate simulated datasets.
Command line usage for phantom creation and simulation
demonstrates command line usage including simulating different scanner configurations and acquisition protocols
run with bash demo_01_phantom_creation.sh
Command line usage for more complex simulation experiments with changing acquisition parameters
After Installation
Check out the Usage section for detailed information on customizing dataset running_simulations.
- Computational notebooks have also been provided to demonstrate how to use pediatricIQphantoms dataset including:
Contents¶
Developers¶
If you’d like to contribute to the code or documentation of this project, please check out our Contributing Guide page.