Medical Image Label Transfer Across Contrast Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The scarcity and time-consuming nature of medical image labeling, particularly for non-contrast computed tomography images, due to the difficulty in manually locating lesions without contrast medium, and the need for converting labels between different medical images captured under varying examination conditions.
Innovation Solution
A medical image processing system and method that calculates a transformation function between medical images captured under different conditions, allowing the conversion of labels from easily labeled images (e.g., with contrast medium) to harder-to-label images (without contrast medium) using alignment and pre-processing techniques, followed by applying this function to adjust label positions and sizes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual labeling is performed on medical images without contrast medium, then labeling accuracy can be maintained, but the time and cost required increases significantly
Solution Approach 1:
The system performs preliminary alignment and transformation function calculation between contrast and non-contrast images before labeling. By pre-processing the images to establish spatial relationships, the system enables efficient label transfer without requiring time-consuming manual labeling of non-contrast images, thus reducing labeling time while maintaining accuracy through the preliminary established transformation relationships
Solution Approach 2:
The system copies labels from contrast images to non-contrast images using calculated transformation functions. Instead of manually creating labels for each non-contrast image, the system generates corresponding labels by applying spatial transformations to proven accurate labels from contrast images, significantly reducing time and cost while preserving labeling accuracy through the mathematical transformation relationships
2Quantity of substance
If manual labeling is performed on all medical images, then complete label coverage is achieved, but the productivity decreases due to the professional expertise required
Solution Approach 1:
The system creates a universal labeling approach that works across different image types (contrast and non-contrast) by establishing transformation functions between them. Once labels are created for contrast images, the same transformation framework enables automatic label generation for non-contrast images, making the labeling process universally applicable to multiple image modalities without requiring separate manual labeling efforts for each type
Solution Approach 2:
The system copies verified labels from contrast images to non-contrast images through transformation functions, ensuring complete label coverage across all image types. This copying mechanism maintains comprehensive label availability while dramatically improving productivity by eliminating the need for expert manual labeling of every individual image
3Adaptability or versatility
If labels are converted between different examination conditions, then label utility is improved, but the device complexity increases due to transformation function calculation
Solution Approach 1:
The system changes spatial parameters (position, orientation, scale) of labels when converting between different examination conditions. By applying transformation functions that adjust these parameters based on the relationship between contrast and non-contrast images, the system maintains label utility across different imaging conditions while managing complexity through mathematical parameter transformations
Data Source
AI summary
A medical image processing method includes the following steps. A first medical image about a first patient under a first examination condition is obtained. A second medical image about the first patient under a second examination condition is obtained. A first label corresponding to the first medical image is collected. The first label marks a lesion within the first medical image. A transformation function between the first medical image and the second medical image is calculated by aligning the first medical image with the second medical image. The transformation function is applied to convert the first label into a second label corresponding to the second medical image.