Lesion Proxy-Based PET Image Reconstruction Parameter Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional PET reconstruction systems are not adaptive to individual patient conditions, making it difficult to optimize image quality and quantitation, as they lack the necessary information to adjust reconstruction and processing parameters.

Innovation Solution

The introduction of lesion proxies with known characteristics allows for dynamic adjustment of reconstruction and processing parameters, enabling optimal image quality and quantitation by simulating events caused by actual lesions and monitoring image attributes, which can be toggled on or off for analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If fixed reconstruction parameters are used for all patients, then the system operation is simple and consistent, but the image quality and quantitation accuracy cannot be optimized for individual patients

Engineering Contradiction:
Improveadaptability to individual patient conditionsVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by inserting lesion proxies into the patient image before reconstruction, simulating potential lesions with known characteristics. This preliminary setup enables subsequent optimization of reconstruction parameters based on how well these proxies are recovered, allowing adaptive parameter selection without requiring complex real-time adjustments during the actual reconstruction process.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The invention uses copies (lesion proxies) of actual lesions with known characteristics and positions. These synthetic lesion copies are inserted into patient images to serve as reference objects for evaluating and optimizing reconstruction parameters. The proxies replicate lesion features but with ground-truth properties, enabling parameter optimization without needing the actual patient data.

Inventive Principle:
Principle #26Copying

2Measurement precision

If ground truth information is required for parameter optimization, then personalized medicine can be achieved, but the ground truth is not available for actual patient data

Engineering Contradiction:
Improvequantitation accuracyVSAvoidavailability of ground truth information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The lesion proxies serve as intermediary objects between the available patient scan data and the desired ground truth information. By inserting these synthetic lesions with known characteristics into the patient image, the system creates a bridge that allows optimization of reconstruction parameters using the proxies' known properties as surrogate ground truth, without requiring actual ground truth from the patient scan.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system creates synthetic copies of lesions with precisely known characteristics (size, shape, activity concentration, position) and inserts them into patient images. These copies provide the necessary ground truth information that would otherwise be unavailable, enabling accurate measurement and optimization of reconstruction parameters while remaining completely synthetic and removable afterward.

Inventive Principle:
Principle #26Copying

3Adaptability or versatility

If lesion proxies are inserted into patient images, then reconstruction parameters can be optimized, but the presence of proxies may interfere with actual lesion detection

Engineering Contradiction:
Improveparameter optimization capabilityVSAvoidinterference with actual lesion detection
Core Design Contradiction:
Adaptability or versatilityVSObject-generated harmful factors

Solution Approach 1:

The system extracts or removes the lesion proxies from the final reconstructed image after they have served their optimization purpose. By taking out the synthetic proxies after parameter optimization is complete, the system eliminates the harmful interference they might cause to actual lesion detection, while retaining the benefit of optimized reconstruction parameters applied to the actual patient data.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The lesion proxies are inserted only for the preliminary purpose of optimizing reconstruction parameters, not for actual diagnosis. This preliminary, temporary presence allows the system to tune parameters without the proxies becoming permanent features that could interfere with clinical interpretation of actual lesions.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3378041B1Pet image reconstruction and processing using lesion proxies
Publication Date: 2020.01.15 KONINKLIJKE PHILIPS NV
  • EP3378041B1 patent drawingFigure 1
  • EP3378041B1 patent drawingFigure 2
  • EP3378041B1 patent drawingFigure 3

AI summary

When performing nuclear medicine image reconstruction, lesion proxies (208) are introduced by a clinician and merged with real acquired scan data outside or inside the patient in the patient image. By monitoring the image attributes of the lesion proxies during reconstruction and processing, reconstruction and processing parameters can be dynamically adapted or adjusted in order to optimize image quality and quantitation to improve delivery of precise, personalized medical treatment.