Quarter Detector Offset CT Reconstruction for Resolution

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current nutated slice reconstruction algorithms for CT scanners, such as NSR and AMPR, are limited by the in-plane sampling rate of 1/wdiso near isocenter due to the use of a fixed focal spot, which restricts the resolution of reconstructed images, especially in helical scanning modes.

Innovation Solution

The implementation of a quarter detector offset (QDO) geometry in CT scanners, where detectors are displaced by half a detector width relative to the x-ray source during each 180° rotation, allowing for increased sampling in the x-direction by interleaving projection data before reconstruction, thereby enhancing in-plane resolution without requiring a 'flying focal spot' hardware.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a fixed focal spot is used in helical CT scanning, then the system cost is reduced and hardware complexity is simplified, but the in-plane sampling rate is limited to 1/wdiso near isocenter, which restricts the resolution of reconstructed images

Engineering Contradiction:
Improvehardware complexityVSAvoidimage resolution
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The patent applies quarter detector offset geometry, which introduces a spatial offset in the detector array arrangement. This dimensional change in detector positioning allows projection data from 180° apart views to be interleaved, effectively increasing the in-plane sampling rate from 1/wdiso to 2/wdiso near isocenter, thereby improving image resolution without changing the focal spot hardware

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

Solution Approach 2:

The patent performs preliminary interpolation of projection data into tilted planes before the final reconstruction step. By pre-processing the data in tilted coordinate systems that match the helical trajectory, the method prepares interleaved projection data that can be directly used for high-resolution reconstruction, resolving the sampling rate limitation imposed by fixed focal spot hardware

Inventive Principle:
Principle #10Preliminary action

2Manufacturing precision

If interleaving of rays 180 degrees apart with quarter detector offset is implemented, then the in-plane sampling rate increases and image resolution improves, but the reconstruction algorithm complexity increases

Engineering Contradiction:
Improvein-plane sampling rateVSAvoidreconstruction algorithm complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent employs dynamic tilted plane interpolation where the plane orientation angles (alpha and beta) vary depending on the helical pitch and detector row position. This dynamic adaptation of interpolation parameters allows the algorithm to optimize sampling for each specific scan configuration, achieving high in-plane sampling rates while managing algorithmic complexity through parameterized solutions rather than fixed geometric transformations

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS9095259B2Method and system for high resolution nutated slice reconstruction using quarter detector offset
Publication Date: 2015.08.04 ANALOGIC CORP
  • US9095259B2 patent drawing
  • US9095259B2 patent drawing
  • US9095259B2 patent drawing

AI summary

A method and system are provided for generating high resolution CT images. The NSR# method improves on the AMPR method, by increasing the in-plane image resolution of CT scanners, in the helical scanning mode. The provided method uses the quarter detector offset and interleaving of complementary data to achieve in plane image resolution that is similar to the high resolution axial scanning mode utilizing quarter detector offset and interleaving. The method includes several ways of choosing the data to be interleaved, like NSR# with two planes, NSR# with 3 planes, NSR# with multiple planes. The interleaved data are used to create high resolution tilted slices. The NSR# method optimizes the untilting filter to create a mix of high and low resolution tilted slices to achieve the desired in-plane image resolution-image artifact balance required for the imaging task. In one embodiment in the untilting process one may use only high resolution tilted slices, for maximum resolution benefit. In another embodiment one may mix high resolution tilted slices with standard resolution tilted slices resulting from data that did not go through interleaving. This creates unfilled slices of higher resolution than the standard and with lower artifacts. In another embodiment for scans with pitch lower than ⅔ one may reduce the collimation to reduce the dose to the patient.