Proton CT Image Reconstruction Using Back-Projection Filtered Algorithm

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current computerized tomography (CT) methods, particularly in proton therapy, face challenges in accurate image reconstruction due to systematic range errors in water-equivalent path length (WEPL) data conversion from X-ray CT to proton stopping power, leading to uncertainties in treatment planning for hadron and ion beam therapy.

Innovation Solution

Implementing a Back-Projection Filtered (BPF) method in proton CT that records individual proton trajectories and uses a band-limited convolution kernel with a correction factor to improve quantitative accuracy and reduce residual errors, allowing for efficient and effective use of computational resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If X-ray CT is used for water-equivalent path length (WEPL) data conversion to proton stopping power, then treatment planning can be performed, but systematic range errors of 3-5% occur for soft tissue and even higher for bone and lung tissues

Engineering Contradiction:
ImproveWEPL data accuracyVSAvoidtreatment planning reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent changes the fundamental parameter being measured from X-ray attenuation (in X-ray CT) to proton transmission (in proton CT). This parameter change eliminates the conversion step from Hounsfield units to proton stopping power, directly measuring proton interaction with tissue and achieving sub-1% range uncertainty compared to 3-5% error in conventional X-ray CT methods

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces the X-ray photon-based measurement system with a proton beam-based measurement system. By substituting the probing mechanism from electromagnetic radiation (X-rays) to hadronic radiation (protons), the system directly measures proton stopping power and water-equivalent path length with superior accuracy for proton therapy planning

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Manufacturing precision

If a ramp filter is used in Filtered Back-Projection to remove artefacts and blurring, then image quality improves, but noise is accentuated

Engineering Contradiction:
Improveimage reconstruction accuracyVSAvoidnoise amplification
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The patent implements an iterative reconstruction algorithm that uses feedback loops to progressively refine the image solution. The algorithm cycles through measured data multiple times, adjusting the reconstructed image based on the difference between measured and calculated projections, and converges to an optimal solution that balances artifact removal with noise suppression

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent applies preprocessing steps to the measured data before reconstruction, including normalization and correction for systematic effects. The iterative algorithm also performs preliminary estimations and refines them progressively, preparing the data and solution in advance to achieve better final image quality with reduced noise

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If iterative or statistical approaches are used for image reconstruction, then quantitative accuracy improves, but computational demand increases significantly

Engineering Contradiction:
Improvequantitative accuracyVSAvoidcomputational efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent implements a limited-memory iterative algorithm that performs a fixed number of iterations (e.g., 10-20 iterations) rather than continuing until full convergence. This partial action provides sufficient quantitative accuracy for clinical applications while dramatically reducing computational time and resource requirements compared to exhaustive iterative methods

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent applies different reconstruction strategies to different regions of the image or different types of data. The iterative algorithm focuses computational effort on regions with higher uncertainty or greater impact on treatment planning, rather than uniformly processing all data with equal computational intensity

Inventive Principle:
Principle #3Local quality

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The BPF method enhances the accuracy of proton CT image reconstruction by minimizing systematic range errors and improving spatial resolution, leading to more precise treatment planning and patient outcomes.

Implementation Method 1

a beam of protons passes from a source through the object being imaged from various angles. Each beam is detected on the side of the body away from the beam source, and its detected intensity, energy or some other property is compared to that property at the source

Methodology Applied
Scientific EffectEnergy loss of protons through matter: Ionisation

Data Source

PatentEP3174465B1Method and apparatus for proton computerised tomography
Publication Date: 2021.10.20 THE UNIV OF LINCOLN
  • EP3174465B1 patent drawingFigure 1~2
  • EP3174465B1 patent drawingFigure 3i~4
  • EP3174465B1 patent drawingFigure 5~6

AI summary

A method of reconstructing a3-dimensionalimage in a proton transmission computerised tomography (CT) apparatus is disclosed. The method comprises the creation of a reconstruction matrix. The matrix is created by directing a plurality of particles to traverse the object; and for each particle, measuring the trajectory and energy of each particle before and after it has traversed the object; for each particle, calculating the water-equivalent path length within the object; and for each particle, calculating the positions at which it entered and exited the object; and adding the water-equivalent path length, entry and exit positions to the reconstruction matrix. This procedure is repeated from a plurality of angular positions surrounding an object to be imaged. Then, a spatially varying 2- dimensional filter function is applied to the reconstruction matrix. Subsequently, a correction factor is applied to the filtered reconstruction matrix to at least partially correct for the finite extent of the matrix.