Machine Learning Tutorial for Tomography

This repository contains tutorials that cover different topics related to Machine Learning applications in tomography. The tutorials have been designed for the Integrating Acquisition and AI in Tomography workshop at the Lorentz Center.

Tutorials Covered

1. Post-processing for CT Images (MLtutorial_postproc.ipynb)
2. Deep Image Prior for CT Reconstruction (MLtutorial_dip.ipynb)
3. Implicit Neural Representations for CT Reconstruction (MLtutorial_inr.ipynb)

Dataset

• Training Data for Post-processing: recon_low.npy, recon_high.npy
• Phantom for Training Data Generation: train_phantom.h5
• Object for Reconstruction Tasks: lung_small.png

Additional utility:

• generate_projs.py: Python script to generate projections for the phantom.

Requirements

• For local machines: Please install the packages listed in the environment.yml file. You can create a new conda environment with the specified packages using:

conda env create -f environment.yml

• For Google Colab: The necessary steps to install the required packages are already included in the notebooks.

Workshop Details

For more details about the workshop and additional resources, please visit the official workshop website.