Photon-Counting X-Ray Detector Pileup Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Photon-counting x-ray detectors face challenges with pulse pileup, leading to reduced contrast-to-noise ratio, spectral distortion, and inefficiencies in image correction processes due to the loss of counts and spectral migration, especially in non-paralyzable energy-discriminating detectors, which are complex to correct and not optimized for real-time medical imaging systems.
Innovation Solution
A method for pileup correction in non-paralyzable energy-discriminating photon-counting x-ray detectors involves adding a correction term to the detected signal, composed of two separable parameterized functions, one depending on the weighted sum of detected signals across energy bins and the other on the spectral response, reducing algorithm complexity and facilitating early implementation in the imaging chain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If photon-counting x-ray detectors are used to measure energy for each photon, then additional information about object composition is obtained, but pulse pileup occurs leading to loss of counts and spectral distortion
Solution Approach 1:
The patent applies spectral pileup correction that models the deterministic parts of pileup effects to convert the harmful spectral distortion into correctable data. By modeling how pileup distorts the spectral response and applying inverse correction, the system recovers accurate compositional information despite pileup occurring at high count rates.
2Ease of manufacture
If standard x-ray imaging correction processes are applied, then image correction is performed, but these processes assume linear detector response which is invalid under pileup conditions
Solution Approach 1:
The patent applies pileup correction early in the image chain before other correction processes or data reduction operations. This preliminary correction establishes accurate spectral response and count rates that subsequent correction processes can then rely upon, preventing error propagation through the imaging pipeline.
3Adaptability or versatility
If material decomposition is performed with different x-ray tube currents, then compositional imaging is achieved, but spectral response varies with count rate causing bias in material-decomposed images
Solution Approach 1:
The patent models spectral response as a function of count rate parameters, allowing the system to adapt the spectral response model to the actual operating conditions. By parameterizing the spectral response variation with count rate and applying appropriate corrections, accurate material decomposition is achieved across different x-ray tube currents.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided a method and corresponding system for correction of the pileup effect in energy-discriminating photon-counting detectors. Correcting for pileup improves linearity and the spectral response of the detector and is crucial for many applications of photon-counting detectors. The method and corresponding system are low in complexity and can be implemented early in the imaging chain, possibly directly in the detector firmware to improve speed and data handling performance of the x-ray imaging system. According to a first aspect, there is provided a method for pileup correction in a non-paralyzable energy-discriminating photon-counting x-ray detector operating based on a number of energy bins. The method comprises adding, for each of a number of energy bins, a correction term to the detected signal of the energy bin, said correction term being a product of two separable parameterized functions, each of which includes at least one parameter, where a first parameterized function depends on a weighted sum of the detected signal over the energy bins, and where a second parameterized function depends on the detected signal(s) in one or several energy bin(s). By assuming separability and ignoring any cross correlations, the number of parameters and the complexity of the pileup correction algorithm are reduced substantially.