Liver Image Segmentation Workflow for Faster Tumor Quantification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing liver analysis methods in medical imaging are time-consuming and lack the ability to optimally combine automated, interactive, and optional steps, leading to inefficiencies and inter-operator variability in tumor quantification and segmentation.
Innovation Solution
A computer-implemented method and system that integrates automated, interactive, and optional steps for liver analysis, allowing parallel processing and user-defined workflows, including liver volume segmentation, vessel-based segment separation, and advanced analysis such as liver fat quantification and perfusion computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual outlining and segmentation of liver tumors is performed, then measurement precision and diagnostic accuracy are improved, but processing time and operational complexity increase significantly
Solution Approach 1:
The liver analysis process is divided into distinct functional modules: automated tumor detection, segmentation, volume calculation, and burden computation. Each module handles a specific aspect of the analysis, allowing parallel processing and reducing overall processing time while maintaining precision through specialized algorithms for each task.
Solution Approach 2:
A computer-readable program acts as an intermediary between the medical images and the clinician, automatically performing detection, segmentation, and quantification tasks. This software mediator processes images through multiple phases (unenhanced, arterial, portal venous, delayed) and generates standardized measurements, eliminating manual intervention while preserving diagnostic accuracy.
2Productivity
If automated segmentation methods are used, then processing speed and productivity are improved, but measurement precision and reliability decrease due to inter-operator variability
Solution Approach 1:
The segmentation algorithm dynamically adapts to different imaging phases and tumor characteristics by adjusting detection parameters based on contrast enhancement patterns. The system processes multiple phases (unenhanced, arterial, portal venous, delayed) and integrates findings across phases, allowing the method to respond to varying tumor densities and boundaries while maintaining consistent, reproducible measurements across different cases and operators.
Solution Approach 2:
The system incorporates feedback mechanisms where detection results from one phase inform processing of subsequent phases. Tumor regions identified in earlier phases guide the search and segmentation in later phases, with results continuously refined and validated across multiple imaging sequences. This iterative feedback process ensures consistent, reliable measurements while maintaining high processing speed.
3Reliability
If comprehensive multi-phase imaging analysis is performed, then diagnostic reliability and tumor detection accuracy are improved, but device complexity and computational requirements increase
Solution Approach 1:
The complex multi-phase analysis is segmented into distinct processing stages: unenhanced phase for baseline anatomy, arterial phase for hypervascular tumor detection, portal venous phase for general tumor visualization, and delayed phase for characterization. Each phase is processed independently with phase-optimized algorithms, then results are integrated to produce comprehensive tumor burden assessment. This segmentation reduces computational complexity compared to simultaneous processing of all phases.
Solution Approach 2:
A single integrated software platform performs multiple diagnostic functions across all imaging phases: detection, segmentation, volume calculation, burden computation, and reporting. The system universally handles different tumor types, sizes, and locations across all four phases, eliminating the need for multiple separate tools or manual methods while maintaining comprehensive diagnostic capability.
Data Source
Figure 1
Figure 2
AI summary
A method, as well as a system and a computer readable medium based on the method, for liver analysis, in which a basic reviewing (14) is carried out on the set of medical images and the result of its segmentation; and prior to carrying out user selected further segmentations, a liver volume reviewing step (17) is carried out enabling the user to review and manually edit the liver volume, and after a comprehensive reviewing step (22), selectively returning (23) is possible to the basic reviewing step (14) if the comprehensive review indicates that further processing of the set of medical images is necessary.