Non-Contrast CT Functional Imaging with Radiomics and Virtual PET

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current PET-CT imaging techniques face limitations such as patient motion between scans, high costs, radiation exposure, local complications from radiotracers, and allergies, making it difficult to accurately superimpose functional and structural images.

Innovation Solution

A method to identify functional features from non-contrast CT images using radiomic analysis, generative adversarial networks, and classification models, eliminating the need for radiotracers and contrast agents.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If PET-CT imaging is performed with radiotracers, then functional features can be detected, but patient motion between scans and registration accuracy deteriorate

Engineering Contradiction:
Improvefunctional feature detection accuracyVSAvoidimage co-registration accuracy
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent combines functional and structural imaging into a single CT scan procedure. By integrating both imaging modalities into one scan, patient motion between separate scans is eliminated, ensuring accurate co-registration while maintaining functional feature detection capabilities through radiomic analysis

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent creates virtual PET images from CT data using deep learning techniques. Instead of requiring actual PET scanning with radiotracers, the system generates synthetic PET-like images that replicate functional information, thereby eliminating motion-related registration errors while preserving functional detection accuracy

Inventive Principle:
Principle #26Copying

2Measurement precision

If radiotracers are administered to patients, then functional imaging can be performed, but production costs and facility requirements increase

Engineering Contradiction:
Improvefunctional activity detectionVSAvoidimaging system accessibility
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent uses CT scan data to generate virtual PET images through deep learning models. This copying approach replicates the functional imaging capability without requiring actual radiotracers, cyclotrons, or specialized PET facilities, making functional imaging accessible through standard CT scanners

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent extracts functional information from standard CT scan data through radiomic analysis and deep learning. By separating functional feature extraction from the need for radiotracers, the system achieves functional imaging using only widely available CT technology

Inventive Principle:
Principle #2Taking out (Extraction)

3Measurement precision

If radiotracers and contrast agents are injected into patients, then imaging quality improves, but local complications and allergic reactions increase

Engineering Contradiction:
Improveimaging qualityVSAvoidlocal complications and allergies
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent extracts functional imaging capability from the radiotracer injection process. By obtaining functional information from standard CT data without injections, the system eliminates local complications from needle insertions and allergic reactions to radiotracers and contrast agents

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system creates virtual functional images from CT data, copying the functional imaging effect without the harmful side effects of radiotracer and contrast agent administration

Inventive Principle:
Principle #26Copying

Data Source

PatentEP4150569B1Functional imaging features from computed tomography images
Publication Date: 2025.11.05 OXFORD UNIVERSITY INNOVATION LTD
  • EP4150569B1 patent drawingFigure 1A~1B
  • EP4150569B1 patent drawingFigure 2A~2B
  • EP4150569B1 patent drawingFigure 3A~3B

AI summary

Methods, apparatus and computer readable media are provided for identifying functional features from a computed tomography (CT) image.. The CT image may be a contrast-enhanced CT image or a non-contrast CT image. According to some examples, methods, apparatus and computer readable media are also provided for using machine learning to identify functional features from CT images. According to some examples, simulated functional image datasets such as simulated PET images or simulated SUV images are generated from a received CT image.