Photon-Counting Detector Output Model Calibration for CT Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for characterizing the output of photon-counting detectors (PCDs) in X-ray computed tomography (CT) systems face challenges such as non-linear responses to X-ray energy levels, spectral distortion due to charge sharing and pulse pileup, and difficulties in developing models that accurately represent PCD data, leading to model-data mismatch and pixel-to-pixel variations.
Innovation Solution
A method and apparatus for developing a highly accurate PCD output model that incorporates count-rate-dependent threshold energy functions, addressing global and local mismatches by optimizing global and local threshold energies, and estimating model parameters based on calibration data from various combinations of basis materials and X-ray tube currents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If model-based methods are used to characterize PCD output, then the ability to handle untested conditions is improved, but model-data mismatch and pixel-to-pixel variations occur
Solution Approach 1:
The patent applies local quality by introducing pixel-specific parameter sets that allow each pixel to have its own calibrated characteristics. This resolves the contradiction by maintaining the model's adaptability to untested conditions while improving measurement precision through pixel-level customization that accounts for manufacturing variations and local detector responses.
Solution Approach 2:
The patent changes parameters by introducing count-rate-dependent threshold energy functions and multiple parameter sets that adapt to different operating conditions. This resolves the contradiction by allowing the model to maintain accuracy across varying conditions (adaptability) while using calibrated parameters to reduce model-data mismatch (precision).
2Measurement precision
If data-based approaches are used for PCD characterization, then simplicity and accuracy under matched conditions are improved, but labor intensity increases due to large calibration data requirements
Solution Approach 1:
The patent applies preliminary action by performing comprehensive calibration across multiple count rates and material combinations in advance. This creates a library of pre-calibrated parameter sets that can be quickly selected during operation, maintaining high accuracy under matched conditions while reducing the time required during actual use.
Solution Approach 2:
The patent introduces dynamic parameter selection based on count rate conditions. The system automatically selects the appropriate parameter set based on the current operating conditions, maintaining accuracy without requiring manual recalibration for each condition, thus reducing time loss while preserving precision.
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 resulting PCD output model demonstrates exceptional agreement with measured data, achieving mean absolute percentage errors (MAPEs) of less than 5% across all test data, effectively addressing both global model-data mismatch and pixel-to-pixel variations, ensuring precise and reliable measurement results.
Implementation Method 1
photon-counting detector (PCD) used in a computed tomography (CT) system
Implementation Method 2
the responses of a PCD to different X-ray energy levels are not linear, and the X-ray spectrum is distorted due to several factors, such as charge sharing (CS)
Implementation Method 3
the X-ray spectrum is distorted due to several factors, such as charge sharing (CS) and pulse pileup (PP)
Data Source
AI summary
A method for determining a model characterizing an output from a photon-counting detector (PCD) used in a computed tomography (CT) system is provided. The PCD has a plurality of pixels. The method includes constructing a PCD output model that has a plurality of model parameters including a first model parameter set. The first model parameter set is dependent on an incident count rate on the PCD, and dependent on a pixel position in the PCD. The method also includes receiving calibration data acquired by scanning a plurality of combinations of basis materials, under a plurality of incident count rates. The method further includes estimating the plurality of model parameters based on the received calibration data.