Method and System for Image Reconstruction for Computer Tomography

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention provides an improved method and system for reconstructing images in computed tomography (CT) scans, specifically aiming to reduce or eliminate beam hardening artifacts. It utilizes a reconstruction algorithm based on the generalized Lambert-Beer law adapted for polychromatic X-ray sources, and applies advanced photon simulation techniques (such as differentiable Monte Carlo simulations) to produce clearer, more accurate images, especially in scans involving multiple materials.

Use CasesContent extracted from patent full text and abstract with AI.

• Medical CT imaging for patients, especially those with metal implants (e.g. orthopedic hardware, dental fillings) to improve diagnostic accuracy by removing artifacts.
• Industrial quality control in manufacturing processes involving objects with mixed materials, such as electronics (circuit boards), lenses, devices with metal-plastic components, etc.
• Non-destructive testing and analysis of composite materials or complex assemblies, where accurate differentiation of multiple materials is required.
• Archaeological or paleontological imaging of fossils or artifacts composed of different materials.
• Research applications needing precise imaging of multi-material samples.

BenefitsContent extracted from patent full text and abstract with AI.

• Significantly reduces or eliminates beam hardening artifacts in CT images, leading to much clearer and more accurate results.
• Enables reliable and precise measurements for both industrial inspection and medical diagnostics, improving defect detection and diagnosis.
• Improves image quality for objects comprised of multiple materials of different densities, such as metal and plastic or bone and tissue.
• Facilitates accurate imaging even when strong artifacts previously hindered the analysis (e.g. scans of patients with implants).
• Supports automated and optimized image reconstruction via iterative, differentiable photon simulation and gradient-based optimization, leading to better convergence and consistency in results.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The present invention pertains to a method and system for image reconstruction for computer tomography, in particular for reducing beam hardening effects, wherein image reconstruction is performed based on the generalized Lambert-Beer law for polychromatic sources.