Augmented Segmented Image Generation for Radiotherapy Replanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In radiotherapy, the anatomy of cancer patients changes over time, necessitating adjustments to treatment plans, but the lack of large-scale labeled segmented image data hinders the performance of automated contouring techniques used for replanning.
Innovation Solution
A system and method for generating augmented segmented image sets by applying transformations such as displacement, rotation, deformation, and appearance/disappearance of anatomical structures to historical segmented image data, using a processor to determine transformation parameters and generate new feature data for training models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated contouring techniques are used for radiotherapy replanning, then the efficiency of treatment plan modification is improved, but the performance is hindered due to lack of large-scale labeled segmented image data
Solution Approach 1:
The system performs preliminary action by generating augmented segmented image data in advance through transformations of historical data. This pre-generated training data is then used to train automated contouring models, enabling them to perform reliably when actual replanning tasks occur. The transformations (displacement, rotation, deformation, appearance/disappearance of structures) are applied beforehand to create diverse training scenarios.
2Reliability
If more labeled segmented image data is collected, then the performance of automated contouring is improved, but the time and resources required for data acquisition and labeling increase
Solution Approach 1:
The system creates copies of existing historical segmented image data and applies various transformations to these copies. Instead of collecting and labeling new data from scratch, the method generates synthetic training data by transforming existing labeled data through displacement, rotation, deformation, and structural appearance/disappearance operations, significantly reducing time and resource requirements.
Solution Approach 2:
The system changes parameters of existing image data through transformations such as displacement vectors, rotation angles, deformation magnitudes, and structural appearance/disappearance probabilities. These parameter changes generate diverse training samples from a single source, expanding the training dataset without requiring additional data collection or labeling time.
Data Source
AI summary
A system and method for generating augmented segmented image set obtain are provided. The method may include: obtaining a first image including a first anatomical structure of a first object; determining first feature data of the first anatomical structure; determining one or more first transformations related to the first anatomical structure, wherein a first transformation includes a transformation type and one or more transformation parameters related to the transformation type; applying the one or more first transformations to the first feature data of the first anatomical structure to generate second feature data of the first anatomical structure; and generating a second image based on the second feature data of the first anatomical structure.