Image Segmentation Consistency Checking for Ultrasound Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image segmentation methods in medical imaging, particularly ultrasound, are unreliable due to poor image quality, anatomical anomalies, and unexpected imaging fields-of-view, leading to inaccurate delineation of anatomical structures and inconsistent segmentation results.
Innovation Solution
An image analysis method that compares segmentation outcomes of multiple images of the same anatomical region taken at different time points to detect failures by measuring consistency between segmented features, using machine learning algorithms to distinguish between normal and abnormal inconsistencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If model-based segmentation is used to automate anatomical delineation, then processing speed and automation are improved, but segmentation reliability deteriorates due to poor image quality and anatomical anomalies
Solution Approach 1:
The system performs consistency checking by comparing segmentation results across multiple time points and feeds back accuracy assessments to the clinician. The consistency measure calculation and failure detection mechanisms provide continuous feedback loops that verify segmentation reliability before clinical use.
Solution Approach 2:
The system performs preliminary consistency checking across multiple time points before final segmentation is accepted. By proactively comparing segmentations at different times and detecting potential failures beforehand, the system prevents unreliable segmentations from being used for diagnosis.
2Productivity
If segmentation algorithms process images with poor quality or missing signal data, then processing completeness is improved, but measurement precision deteriorates due to inaccurate delineation
Solution Approach 1:
The system calculates consistency measures between segmentations at different time points and provides feedback about accuracy. When inconsistencies exceed thresholds indicating poor quality input, the system notifies the clinician, creating a feedback loop that ensures only accurate segmentations are used.
Solution Approach 2:
The system replaces direct trust in single-time-point segmentation with a statistical consistency checking mechanism. By substituting mechanical segmentation output with a probabilistic consistency assessment across multiple time points, the system can distinguish between true anatomical changes and segmentation errors.
3Reliability
If consistency checking across multiple time points is performed, then segmentation accuracy is improved, but computational complexity increases
Solution Approach 1:
The system performs consistency checking selectively rather than exhaustively. It compares segmentations at multiple time points but only when and where necessary, using thresholds to determine when checking is needed. This partial action approach maintains accuracy while reducing computational burden.
Data Source
AI summary
An image analysis method and device is for detecting failure or error in an image segmentation procedure. The method comprises comparing (14) segmentation outcomes for two or more images, representative of a particular anatomical region at different respective time points, and identifying a degree of consistency or deviation between them. Based on this derived consistency or deviation measure, a measure of accuracy of the segmentation procedure is determined (16).