Task-Based Trajectory Optimization for Cone-Beam CT Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cone-beam CT interventional systems lack patient-specific and task-based approaches for defining source-detector scan orbits, relying on coarse heuristics that do not account for anatomical context or quantitative imaging task requirements, leading to suboptimal image quality and inefficiencies.
Innovation Solution
A method that identifies task-based performance predictors such as noise, spatial resolution, and detectability index, using numerical observer models and penalized-likelihood iterative reconstruction to determine optimal projection views and scan trajectories for improved image quality, tailored to specific patient and imaging tasks, with the ability to constrain orbits arbitrarily.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional prescriptive scan orbits are used, then the system is simple to operate and mechanically convenient, but image quality is suboptimal and does not account for patient-specific characteristics
Solution Approach 1:
The system performs preliminary actions by pre-calculating optimal scan trajectories using task-based performance predictors and numerical observer models before actual imaging. The method identifies performance predictors that affect image quality, applies them to reconstruction properties, and determines optimal projection views in advance, allowing the system to execute pre-planned optimized orbits rather than relying on prescriptive mechanical convenience
Solution Approach 2:
The system changes parameters by transitioning from fixed prescriptive orbital parameters to dynamically optimized parameters based on patient-specific anatomy and task requirements. The method uses penalized-likelihood iterative reconstruction with task-based performance predictors to determine variable scan trajectories, adjusting source-detector positioning parameters to maximize detectability index while accounting for local noise and spatial resolution properties
2Measurement precision
If coarse heuristics are used for positioning, then the acquisition process is fast and simple, but image quality and detectability are suboptimal
Solution Approach 1:
The system replaces mechanical heuristics with computational models by substituting coarse positioning rules with task-based performance predictors and numerical observer models. The method uses general vectorized forward models and system matrices to computationally determine optimal projection views, replacing manual or rule-based positioning with automated optimization algorithms that calculate detectability index and reconstruction properties
Solution Approach 2:
The system performs self-service by automatically optimizing its own scan trajectories without requiring extensive manual intervention. The method uses iterative reconstruction algorithms that self-adjust parameters based on task-based performance predictors, with the system autonomously identifying optimal projection views and scan orbits based on patient-specific characteristics and imaging tasks
3Productivity
If traditional scan orbits are used, then the system is easy to operate, but radiation dose efficiency is poor and repeat scans are more frequent
Solution Approach 1:
The system implements feedback by using task-based performance predictors to continuously evaluate and optimize scan trajectories. The method incorporates iterative reconstruction with penalized-likelihood estimation that uses detectability index and reconstruction properties as feedback signals to adjust projection views and orbital parameters, ensuring radiation dose is efficiently utilized to maximize image quality and minimize the need for repeat scans
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 results in enhanced spatial resolution and noise characteristics, reducing radiation dose, improving image quality, and streamlining interventional workflows by optimizing scan trajectories based on patient-specific and task-specific data, leading to superior image quality and reduced repeat scans.
Implementation Method 1
Traditional cone-beam CT interventional systems
Implementation Method 2
y=D{b}exp(−Aμ)
Data Source
AI summary
An embodiment in accordance with the present invention provides a method for applying task-based performance predictors (measures of noise, spatial resolution, and detectability index) based on numerical observer models and approximations to the local noise and spatial resolution properties of the CBCT reconstruction process (e.g., penalized-likelihood iterative reconstruction). These predictions are then used to identify projections views (i.e., points that will constitute the scan trajectory) that maximize task performance, beginning with the projection view that maximizes detectability, proceeding to the next-best view, and continuing in an (arbitrarily constrained) orbit that can be physically realized on advanced robotic C-arm platforms. The performance of CBCT reconstructions arising from a task-based trajectory is superior to simple and complex orbits by virtue of improved spatial resolution and noise characteristics (relative to the specified imaging task) associated with the projection views constituting the customized scan orbit.