Photon Counting Detector Depth Segmentation for Pile-Up Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Photon counting x-ray imaging systems face challenges with signal pile-up, energy resolution degradation, and noise counts due to high x-ray flux rates, particularly in clinical CT applications, where silicon detectors suffer from Compton interactions and low atomic number limitations.

Innovation Solution

The method involves allocating detected events into energy bins, inverting the detector response function, and integrating energy bin functions to extract original x-ray quantum energies, and resetting the shaper filter output to enhance count efficiency and energy resolution, while allowing selection between image reconstruction modes based on energy information to manage noise counts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If silicon detectors are used for photon counting at high x-ray flux rates, then signal pile-up is reduced due to shorter collection times, but energy resolution deteriorates due to Compton interactions

Engineering Contradiction:
Improvecharge carrier collection timeVSAvoidenergy resolution
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent introduces depth segmentation, dividing the detector into multiple layers at different depths. This spatial dimension allows separation of photoelectric events (which occur throughout the depth) from Compton events (which have characteristic depth distributions), enabling energy resolution improvement while maintaining fast collection times in each layer

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The detector is segmented into multiple depth layers, with each layer independently measuring photon interactions. This segmentation allows the system to distinguish between photoelectric and Compton events based on their different depth profiles, resolving the energy resolution issue while preserving the fast response of silicon

Inventive Principle:
Principle #1Segmentation

2Object-affected harmful factors

If a high lower threshold is applied to reject electronic noise, then noise counts are reduced, but primary x-ray signals are lost due to Compton interactions depositing low energy

Engineering Contradiction:
Improveelectronic noise rejectionVSAvoidprimary signal detection
Core Design Contradiction:
Object-affected harmful factorsVSReliability

Solution Approach 1:

By segmenting the detector into multiple depth layers and analyzing the spatial distribution of detected events, the system can distinguish between low-energy primary photons and Compton scattered photons. This allows setting a higher noise threshold without losing primary signals, as Compton events can be identified and handled separately through their characteristic depth profiles

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the approach from using a single energy threshold to using multiple depth-based parameters. By measuring the depth distribution of energy deposits across multiple layers, the system can differentiate between primary and Compton events even when total deposited energy is low, enabling better noise rejection while preserving primary signal detection

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If photon counting mode is used to increase contrast-to-noise ratio, then image quality improves at the same patient dose, but signal pile-up occurs at high flux rates

Engineering Contradiction:
Improvecontrast-to-noise ratioVSAvoidcount rate handling
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

Dividing the detector into multiple depth layers reduces the count rate burden on each individual layer. By distributing the high flux across multiple segments and using their combined depth-profiled measurements, the system maintains the high contrast-to-noise ratio of photon counting while avoiding signal pile-up in each layer

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent employs periodic or sequential readout of multiple depth layers, allowing each layer to clear its signal before the next measurement cycle. This temporal segmentation of the measurement process prevents signal pile-up while maintaining the statistical advantages of photon counting for improved contrast-to-noise ratio

Inventive Principle:
Principle #19Periodic action

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

This approach improves the extraction of energy information, increases image quality, and reduces patient dose by approximately 20% by effectively handling signal pile-up and noise counts, enhancing the spectral imaging capabilities of photon counting detectors.

Implementation Method 1

resetting the shaper filter output to enhance count efficiency and energy resolution

Methodology Applied
Scientific EffectSignal shaping:

Implementation Method 2

When an x-ray quantum deposits energy by means of the photoelectric effect in a direct conversion detector

Methodology Applied
Scientific EffectPhotoelectric effect: Photoelectric Effect

Implementation Method 3

the Compton effect replaces the photoelectric effect as the dominant type of interaction

Methodology Applied
Scientific EffectCompton scattering: Compton Scattering

Data Source

PatentEP3447539B1Improving image quality in photon counting-mode detector systems
Publication Date: 2025.01.15 GE PRECISION HEALTHCARE LLC
  • EP3447539B1 patent drawingFigure 1A~1B
  • EP3447539B1 patent drawingFigure 2
  • EP3447539B1 patent drawingFigure 3

AI summary

The current invention applies to photon counting silicon x-ray detectors with energy discriminating capabilities and applications in x-ray imaging systems. The overall image quality produced by such a system is improved by the presented novel methods for optimally using the energy information in Compton events and making selective use of counts induced from charges collected in neighboring pixels. The pile-up problem during high-flux imaging regimes is reduced by a novel method for signal reset, which improves the count efficiency by reducing the risk of losing event due to signal pile-up in the read out electronics chain.