Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Proton Computed Tomography Image Reconstruction Based on the Richardson-Lucy Algorithm

Gábor Bíró, Ákos Sudár, Zsófia Jólesz, Gábor Papp, Gergely Gábor Barnaföldi

TL;DR

An iterative image reconstruction algorithm, based on the Richardson-Lucy iteration is proposed for the first time for proton CT image reconstruction, and provides a promising proof-of-concept candidate for compromise between accuracy and speed with several further development directions.

Abstract

Proton therapy is an emerging method in cancer therapy. One of the main developments is to increase the accuracy of the Bragg-peak position calculation, which requires more precise relative stopping power (RSP) measurements. A promising choice is the application of proton computed tomography (pCT) systems which takes the images under similar conditions, as they use the same irradiation device and hadron beam for imaging and treatment. A key aim is to develop a precise image reconstruction algorithm for pCT systems to reach their maximal performance. In this work, an iterative image reconstruction algorithm, based on the Richardson-Lucy iteration is proposed for the first time for proton CT image reconstruction. Monte Carlo (MC) simulations of CTP528 and CTP404 phantoms were used to benchmark the proposed method. In the case of an idealized detector setup, using a 1 mm pitch grid, 4.88 lp/cm spatial resolution and 0.66% average RSP uncertainty was achieved. The present method provides a promising proof-of-concept candidate for compromise between accuracy and speed with several further development directions.

Proton Computed Tomography Image Reconstruction Based on the Richardson-Lucy Algorithm

TL;DR

An iterative image reconstruction algorithm, based on the Richardson-Lucy iteration is proposed for the first time for proton CT image reconstruction, and provides a promising proof-of-concept candidate for compromise between accuracy and speed with several further development directions.

Abstract

Proton therapy is an emerging method in cancer therapy. One of the main developments is to increase the accuracy of the Bragg-peak position calculation, which requires more precise relative stopping power (RSP) measurements. A promising choice is the application of proton computed tomography (pCT) systems which takes the images under similar conditions, as they use the same irradiation device and hadron beam for imaging and treatment. A key aim is to develop a precise image reconstruction algorithm for pCT systems to reach their maximal performance. In this work, an iterative image reconstruction algorithm, based on the Richardson-Lucy iteration is proposed for the first time for proton CT image reconstruction. Monte Carlo (MC) simulations of CTP528 and CTP404 phantoms were used to benchmark the proposed method. In the case of an idealized detector setup, using a 1 mm pitch grid, 4.88 lp/cm spatial resolution and 0.66% average RSP uncertainty was achieved. The present method provides a promising proof-of-concept candidate for compromise between accuracy and speed with several further development directions.
Paper Structure (27 sections, 14 equations, 8 figures, 2 tables)

This paper contains 27 sections, 14 equations, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Image Reconstruction Techniques
  3. Analytical Reconstruction Methods
  4. Iterative (Algebraic) Reconstruction Methods
  5. Acceleration Strategies and GPU Implementations
  6. Motivation for a Richardson--Lucy–Based Approach
  7. Classical Richardson--Lucy Iteration
  8. Linear Forward Model in Proton CT
  9. Adapted Richardson--Lucy Iteration
  10. Cost Function and Fixed-Point Properties
  11. Mathematical Properties and Practical Implications
  12. Sources of Uncertainty in Proton CT Reconstruction
  13. Data Acquisition and Detector-Related Uncertainties
  14. Physics-Driven Uncertainties
  15. Reconstruction and Model-Dependent Uncertainties
  16. ...and 12 more sections

Figures (8)

  • Figure 1: Detector design of the list mode imaging concept.
  • Figure 2: Single-sided, list mode detector design.
  • Figure 3: The structure of the investigated layers.
  • Figure 4: Simulation steps of the Richardson -- Lucy algorithm evaluation.
  • Figure 7: The ground truth and ideal reconstructions of the CTP404 phantom with the area averaged RSP regions.
  • ...and 3 more figures