Automated Background Region Selection for PET Lesion Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In oncology examinations, particularly when analyzing PET images to determine the progression of a disease, it is challenging to consistently identify the boundaries of lesions over time due to variations in background uptake levels, which complicates the segmentation process and requires labor-intensive standardized approaches.
Innovation Solution
A method and system for automatically detecting and displaying an organ of interest across multiple imaging modalities, such as CT and PET, by using a priori clinical information and algorithms to select and propagate a visual indicator, facilitating consistent background reference region selection and computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a standardized approach to background sampling is used (as recommended by PERCIST), then measurement precision is improved, but device complexity and ease of operation worsen due to labor-intensive manual localization and sampling steps
Solution Approach 1:
The system automatically performs background region localization and sampling without requiring manual operator intervention. The computer automatically identifies the background region based on the lesion location and performs the sampling computation, making the system self-sufficient for this task while maintaining PERCIST compliance
Solution Approach 2:
The system pre-defines background sampling parameters and algorithms before the actual PET image analysis. The background region selection criteria and sampling methodology are established in advance according to PERCIST guidelines, so that when analysis is performed, the system can automatically apply these pre-configured parameters without requiring manual setup
2Measurement precision
If manual background sampling is performed according to standardized protocols, then measurement precision is improved, but productivity deteriorates due to the time-consuming nature of the process
Solution Approach 1:
The manual mechanical process of background sampling (where operators manually locate and measure background regions) is replaced with an automated computer-based system. The computer executes algorithms to automatically identify background regions and compute sampling values, eliminating the need for manual manipulation while maintaining measurement accuracy
Solution Approach 2:
The system introduces an automated computational intermediary between the PET image data and the background sampling result. Instead of direct manual measurement, the computer acts as an intermediary that processes the image data through standardized algorithms to produce the background sampling values, thereby increasing throughput while preserving precision
3Measurement precision
If consistent background sampling approaches are used across multiple examinations, then measurement precision is improved, but device complexity increases due to the need for multi-modality image processing and propagation
Solution Approach 1:
The system merges multiple imaging modalities (CT and PET) into a unified analysis workflow. By combining the anatomical information from CT with the functional information from PET, the system achieves more precise lesion and background region identification while maintaining a streamlined integrated process rather than separate manual procedures for each modality
Data Source
AI summary
A method for automatically displaying an organ of interest includes accessing a series of medical images acquired from a first imaging modality, receiving an input that indicates an organ of interest, automatically detecting the organ of interest within at least one image in the series of medical images, automatically placing a visual indicator in the region of interest, and automatically propagating the visual indicator to at least one image that is acquired using a second different imaging modality. A medical imaging system and a non-transitory computer readable medium are also described herein.