X-ray Tomogram Image Quality Improvement via Radon Transform Filtering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing X-ray tomography techniques suffer from streak artifacts due to objects in the beam path, incomplete projection region coverage, and X-ray source movement, leading to misdiagnosis and masking of low-contrast details.

Innovation Solution

A method involving low-pass and high-pass filtering of X-ray tomograms, followed by Radon transform and inverse Radon transform calculations, to isolate and remove streak artifacts while preserving diagnostic details, using non-linear filtering and value modification techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional X-ray tomography methods are used, then the imaging process is simple and fast, but streak artifacts occur due to objects in the beam path, incomplete projection coverage, and X-ray source movement

Engineering Contradiction:

Improveimage qualityVSAvoidprocessing complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The image processing is divided into distinct stages: low-pass filtering to preserve structural information, high-pass filtering to isolate streak artifacts, Radon transform to identify artifact patterns, and selective value modification. This segmentation allows targeted artifact removal while preserving diagnostic details.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The method extracts streak artifacts from the original image by applying high-pass filtering after low-pass filtering. The high-pass filtered image contains primarily the artifact information, which can then be processed separately through Radon transform and value modification before being combined back with the low-pass filtered image.

Inventive Principle:

Principle #2Taking out (Extraction)

2Reliability

If filtering operations are applied to remove streak artifacts, then artifact reduction is achieved, but spatial resolution and noise spectrum may be impaired

Engineering Contradiction:

Improveartifact reductionVSAvoidspatial resolution

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The low-pass filter is applied selectively only to pixels with image values within a predetermined interval, preserving other regions unchanged. This local application ensures that diagnostic details outside the target intensity range maintain their original spatial resolution and contrast.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The method converts the harmful streak artifacts into a separable component through high-pass filtering. By transforming the artifact problem into a frequency-domain separation task, the artifacts can be identified and removed through Radon transform analysis without affecting the beneficial diagnostic information in other frequency components.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

3Reliability

If complete projection region coverage is achieved, then streak artifacts are reduced, but measurement time increases

Engineering Contradiction:

Improveprojection coverageVSAvoidmeasurement time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The method performs preliminary low-pass filtering on the available projection data before reconstruction. This preprocessing step prepares the data to minimize artifact formation during subsequent reconstruction, allowing acceptable image quality even with incomplete projection coverage, thereby reducing the need for extended measurement times.

Inventive Principle:

Principle #10Preliminary action

4Reliability

If objects are removed from the beam path, then streak artifacts are eliminated, but examination capabilities are restricted

Engineering Contradiction:

Improveartifact eliminationVSAvoidexamination flexibility

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The method introduces computational processing as an intermediary between the physical imaging process and the final image. Instead of physically removing objects from the beam path, the artifacts they cause are identified and removed through digital signal processing techniques including filtering and Radon transform analysis, preserving both the objects and diagnostic capabilities.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11559263B2Method for improving the image quality of x-ray tomograms

Publication Date: 2023.01.24 SIEMENS HEALTHINEERS AG

AI summary

A method for improving an image quality of X-ray tomograms includes generating a low-pass filtered X-ray tomogram by applying a low-pass filter to a two-dimensional X-ray tomogram. The low-pass filter is only applied to pixels with image values lying within a predetermined image value interval. A high-pass filtered X-ray tomogram is generated by subtracting the low-pass filtered X-ray tomogram from the two-dimensional X-ray tomogram. A Radon transform image is generated by calculating a Radon transform of the high-pass filtered X-ray tomogram. A modified Radon transform image is generated by modifying values of the pixels of the Radon transform image with values lying outside a predetermined value interval. A modified high-pass filtered X-ray tomogram is generated by calculating an inverse Radon transform of the modified Radon transform image. A modified X-ray tomogram is generated by the addition of the modified high-pass filtered X-ray tomogram to the low-pass filtered X-ray tomogram.