Medical Image Reconstruction Using Trained Models for Quality Index Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging systems require skilled radiologists to manually set complex acquisition parameters and reconstruction algorithms to prioritize specific image quality indexes, which is time-consuming and often leads to trade-offs between different image quality indices, making it difficult to achieve desired image qualities without extensive expertise.
Innovation Solution
A medical imaging system utilizing a trained model that prioritizes and adjusts image quality indices based on learning data, allowing easy specification of preferred feature values tailored to diagnostic needs, using a trained model to process input medical images and output images with desired quality characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a radiologist manually sets acquisition parameters and reconstruction algorithms to prioritize specific image quality indexes, then the desired image quality can be achieved, but the process becomes time-consuming and requires high skill levels
Solution Approach 1:
The system automatically determines optimal acquisition parameters and reconstruction algorithms based on the diagnostic purpose without requiring manual configuration by radiologists. The determination unit autonomously selects parameters prioritizing relevant image quality indexes based on the specified diagnostic objective, enabling the system to serve itself rather than requiring expert intervention.
Solution Approach 2:
The system pre-establishes relationships between diagnostic purposes and optimal image quality parameters through the trained model. By having the determination unit automatically select appropriate parameters based on the diagnostic purpose before image acquisition begins, the system performs the complex decision-making process in advance, eliminating the need for time-consuming manual adjustments during actual imaging.
2Manufacturing precision
If multiple image quality indexes are improved simultaneously, then comprehensive image quality enhancement is achieved, but trade-offs occur between different indexes making it difficult to satisfy all requirements
Solution Approach 1:
The system prioritizes specific image quality indexes based on the diagnostic purpose rather than attempting to optimize all indexes simultaneously. The determination unit identifies which image quality indexes are most relevant to the specific diagnostic objective and prioritizes those, while maintaining other indexes at acceptable levels, thereby avoiding trade-offs between conflicting requirements.
Solution Approach 2:
The system dynamically adjusts acquisition parameters and reconstruction algorithms based on the specified diagnostic purpose. By changing parameters according to the trained model's determination of optimal settings for each diagnostic scenario, the system adapts to different imaging requirements without manual intervention, resolving the conflict between multiple quality indexes.
3Measurement precision
If skilled radiologists perform manual parameter setting, then accurate image quality optimization is achieved, but the process requires extensive time and effort
Solution Approach 1:
The system replaces the manual mechanical process of radiologist parameter setting with an automated determination unit based on a trained model. The determination unit automatically determines optimal acquisition parameters and reconstruction algorithms by referencing the trained model with the specified diagnostic purpose, substituting human expertise with an automated system that provides equivalent or superior accuracy without the time investment.
Solution Approach 2:
The system uses a trained model that has been trained on learning data containing optimal parameter settings for various diagnostic purposes. By copying the knowledge embedded in the trained model rather than requiring radiologists to recreate optimal settings manually each time, the system achieves accurate parameter optimization instantly without time-consuming manual processes.
Data Source
AI summary
A medical imaging described herein includes a model trained using a first medical image in which a first image quality index is prioritized and a second medical image in which a second image quality index is prioritized. Learning data includes a first algorithm and a first condition for processing data acquired to obtain the first medical image, a feature value of the first image quality index, a second algorithm and a second condition for processing data acquired to obtain the second medical image, and a feature value of the second image quality index. The model determines a first model algorithm and a first model parameter for obtaining a first model medical image in which the first image quality index is prioritized, and/or a second model algorithm and a second model parameter for obtaining a second model medical image in which the second image quality index is prioritized.