Differentiable Forward Projector for X-ray Computed Tomography

TL;DR

This paper introduces a differentiable forward projector for X-ray CT that ensures reconstructed images are consistent with measured data, supporting various geometries and integrating seamlessly with deep learning workflows.

Contribution

It presents an accurate, memory-efficient differentiable projection library for X-ray CT that improves consistency with measurements and supports multiple geometries.

Findings

Supports various projection geometries.

Minimizes GPU memory usage.

Enables better integration with deep learning models.

Abstract

Data-driven deep learning has been successfully applied to various computed tomographic reconstruction problems. The deep inference models may outperform existing analytical and iterative algorithms, especially in ill-posed CT reconstruction. However, those methods often predict images that do not agree with the measured projection data. This paper presents an accurate differentiable forward and back projection software library to ensure the consistency between the predicted images and the original measurements. The software library efficiently supports various projection geometry types while minimizing the GPU memory footprint requirement, which facilitates seamless integration with existing deep learning training and inference pipelines. The proposed software is available as open source: https://github.com/LLNL/LEAP.