PCCT-Guided PET Resolution Enhancement for Small Lesion Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Positron emission tomography (PET) scans suffer from inherently lower spatial resolution compared to computed tomography (CT) scans, limiting their diagnostic utility, especially in detecting small lesions or providing detailed anatomical context to metabolic activity.

Innovation Solution

Integrate high-resolution photon counting CT (PCCT) imaging with PET imaging using advanced generative models such as conditional Diffusion Models, CycleGAN, Attention-UNet, Multi-Modal Variational Autoencoders, and Neural Radiance Fields to enhance the resolution of PET images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional PET imaging is used, then functional metabolic information is obtained, but spatial resolution is inherently lower compared to CT imaging

Engineering Contradiction:

Improvespatial resolutionVSAvoidanatomical detail information

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent uses CT images as an intermediary to transfer high-resolution anatomical information to PET images. The CT image serves as a mediator that provides structural details which are then integrated with the functional PET data through machine learning models, allowing PET to benefit from CT's superior spatial resolution without requiring hardware modifications to the PET scanner itself.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical/image fusion methods with machine learning-based image-to-image translation models. Instead of using conventional registration and fusion algorithms, the system employs deep learning architectures (such as GANs and diffusion models) to automatically learn and transfer high-resolution anatomical features from CT to PET images, achieving superior resolution enhancement without complex mechanical integration.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If image resolution enhancement techniques are applied to PET images, then diagnostic utility improves, but artifacts may be introduced

Engineering Contradiction:

Improveimage resolutionVSAvoidimage artifacts

Core Design Contradiction:

Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The patent employs adversarial feedback mechanisms through Generative Adversarial Networks (GANs) where a generator network creates enhanced PET images and a discriminator network evaluates them against ground truth high-resolution images. This feedback loop continuously refines the enhancement process, ensuring that the generated images maintain anatomical accuracy and minimize artifact introduction by learning from real high-resolution examples.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary registration and alignment of CT and PET images before the main enhancement process. By pre-processing the input images to ensure proper spatial correspondence and removing obvious misalignments, the system prepares the data in advance to prevent artifacts from being generated during the resolution enhancement phase.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20260044935A1High-resolution pet-CT imaging using PCCT and advanced generative models

Publication Date: 2026.02.12 SIEMENS MEDICAL SOLUTIONS USA INC

AI summary

Systems and methods for using paired high-resolution photon counting CT (PCCT) and PET images from the same patient to generate high-resolution PET images. A machine learning network is trained on paired images from patients. When the trained model is applied, a new patient's PET and PCCT images may be used to generate a high-resolution PET image for a medical diagnosis or further processing.