Rotation-Based Image Reconstruction for Multi-Column Detector Systems

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Multi-column detector systems in nuclear medicine imaging require substantial computational resources and long reconstruction times due to iterative reconstruction algorithms, particularly for forward and back projection operations, which can be cumbersome and impractical for large numbers of projection angles.

Innovation Solution

A radiation detector system with plural detector units and a processor that organizes projections by projection angles, rotates images, convolutes and sums slices using kernels, derives error projections, and performs back projections to efficiently reconstruct images, reducing computational resources and reconstruction time while maintaining image quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If iterative reconstruction algorithms are used for multi-column detector systems, then image quality is improved, but computational resources and reconstruction time increase substantially

Engineering Contradiction:

Improveimage qualityVSAvoidreconstruction time

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent divides the projection data into multiple groups based on projection angles. For each group, the algorithm performs rotation, convolution, and summation operations independently. This segmentation allows the reconstruction process to be broken down into manageable chunks, reducing the overall computational burden while maintaining image quality through systematic processing of all projection groups.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary rotation of the image so that projection groups become parallel to a first axis before convolution and summation. This preliminary action simplifies subsequent processing steps by aligning data in a favorable orientation, reducing computational complexity during the main reconstruction operations while preserving the final image quality.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If iterative reconstruction algorithms are used for multi-column detector systems, then image quality is improved, but computational resources increase substantially

Engineering Contradiction:

Improveimage qualityVSAvoidcomputational resources

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

By segmenting projection data into angle-based groups and processing each group through rotation, convolution, and summation independently, the patent reduces the overall computational resource requirements. This approach breaks down the computationally intensive iterative reconstruction into smaller, more manageable operations that consume fewer resources while still achieving high image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent replaces traditional mechanical forward and back projection operations with an efficient convolution and summation approach using separable kernels. This substitution eliminates the need for computationally expensive ray-tracing mechanics, using instead mathematical convolution operations that require significantly fewer computational resources while maintaining reconstruction accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If traditional forward and back projection operations are used, then reconstruction is performed, but computational complexity and time are excessive for large numbers of projection angles

Engineering Contradiction:

Improvereconstruction speedVSAvoidcomputational complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent replaces traditional forward and back projection mechanics with an efficient convolution-based approach. By using separable kernels for convolution and summation operations on rotated images, the system achieves faster reconstruction speeds while reducing computational complexity. This substitution transforms the mechanical projection process into more efficient mathematical operations suitable for large numbers of projection angles.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent performs preliminary rotation of images to align projection groups parallel to a first axis before convolution operations. This preliminary action simplifies the subsequent computational steps by creating a favorable data orientation, reducing the overall computational complexity while enabling faster reconstruction processing for multi-column detector geometries.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9958552B1Systems and methods for rotation based reconstruction for multi-column detector geometries

Publication Date: 2018.05.01 GE PRECISION HEALTHCARE LLC

AI summary

A radiation detector system is provided that includes plural detector units and at least one processor. The detector units are configured to acquire imaging information at plural corresponding projection angles. The at least one processor is configured to acquire projections at the projection angles; organize the projections into groups based on the projection angles; and, for each group of projections, rotate a corresponding image from an original orientation so that the group of projections are parallel to a first axis of the rotated image, convolute and sum slices from the group of projections using kernels to provide a corresponding coordinate set forward projection; perform a back projection to provide back projections; and rotate the back projections to the original orientation and sum the rotated back projections to provide a back projected transaxial image.