iqid-alphas

A Python-based framework for listmode data processing, image processing, and dosimetry using the iQID camera digital autoradiograph, as described in the paper published here: https://doi.org/10.1038/s41598-022-22664-5.

The permanent DOI of this repository is:

Author: Robin Peter

iQID camera expert: Brian Miller

Please contact the authors with any questions or for access to data samples.

Python Dependencies

Please run script "check_dependencies.py" to check these against your versions.

Required:

• numpy (1.20.2)
• opencv / cv2 (4.0.1)
• skimage (0.18.1)
• scipy (1.6.2)
• PyStackReg (0.2.5) (https://github.com/glichtner/pystackreg)

Recommended for visualization and Jupyter notebook demos:

• Jupyter (4.7.1) notebook (6.4.0) or lab (3.0.14)
• matplotlib (3.3.4)

Installation

To install the iQID processing functions: 1. clone git repo 2. python check_dependencies.py to check for the required Python packages. 3. (optional) run the demo notebooks to get a sense of the workflow. 4. import functions from iqid.process, iqid.dpk, or iqid.align as needed in your scripts.

Important note Please mind the file structure if you are running the demo notebooks. See below.

Repository Structure

Below is a description of the current files in the main (stable) branch of the iqid-alphas repo.

iqid-alphas/ (main project directory)

• check_dependencies.py - setup script for checking Python dependencies. See Installation section above.
• iqid/ (Python source code)
  • process.py (processing iQID listmode data)
  • align.py (alignment and registration)
  • dpk.py (dose kernel and DPK convolution)
• demo_notebooks/ (demo Jupyter notebook tutorials)
  • Please use these as templates for your own analysis if you like.
  • Please explore the rest of the source code, as not everything is represented by the tutorials.
  • No demo is currently available for the iqid.process (preprocessing) module, as the data that was published in the Scientific Reports manuscript used a legacy iQID data format. Authors are working to update this for modern iQID devices.
  • Important note each notebook starts with "cd .." to move the working directory up a level to access the source files. This MUST be run in order to import the iQID source code from iqid/ unless you move the .py or .ipynb files. I opted for this solution to avoid adding the iQID package to your system or Python path. However, for more permanent setup, please see discussion on the following thread: https://stackoverflow.com/questions/34478398/import-local-function-from-a-module-housed-in-another-directory-with-relative-im
• parameters/ (shareable data files)
  • at211_10E6_151_1um_water.txt (DPK for At-211 simulated in GEANT4 using $10^7$ alpha particles in water.)
  • scale_data.csv (Spreadsheet containing iQID and H&E scale information for all studies.)
  • activity_correction.csv (Spreadsheet containing small correction values introduced by transformations.)
• data/ (NOT PUBLIC: canine lymph node iQID data from Fred Hutchinson)
  • For data security reasons, these are not currently uploaded to the repository.
  • However, until they are, you should be able to see the file structure of the folders.
  • Additionally, select images are visible in the demo notebook previews. Please feel free to make one copy of the notebooks for viewing and another to experiment using your own data.
  • Finally, please contact the authors if you would like to request access to the data.