Magnetic Resonance Image Reconstruction Using Pre-processed Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image reconstruction methods from magnetic resonance measurement data face challenges in flexibility and efficiency, particularly when dealing with varying forms of recorded measurement data, which can lead to suboptimal image quality and increased processing requirements.
Innovation Solution
A method that processes recorded measurement data into a form compatible with existing trained reconstruction functions, allowing for the application of these functions without the need for retraining or reconfiguration, using techniques such as regridding, slice selection, Fourier transforms, and correction methods to ensure data consistency and adaptability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional image reconstruction methods are used with varying forms of magnetic resonance measurement data, then the reconstruction process must be retrained or reconfigured for each data form, but this increases processing time and computational requirements
Solution Approach 1:
The patent applies preliminary action by pre-training the neural network on a standardized data form and pre-processing various input data forms into the standardized format. This eliminates the need for retraining when new data forms are encountered, as the pre-processing step converts them to the expected input format of the already-trained network, thus resolving the contradiction between adaptability and processing time.
2Adaptability or versatility
If traditional image reconstruction methods are used with varying forms of magnetic resonance measurement data, then the reconstruction process must be retrained or reconfigured for each data form, but this increases device complexity
Solution Approach 1:
The patent implements universality by creating a single standardized data form that can accommodate multiple input data forms through pre-processing. The neural network is trained on this universal standardized form, allowing it to handle various input types without requiring separate models or configurations for each data form, thus reducing device complexity while maintaining adaptability.
3Productivity
If data is processed into a standardized form for trained reconstruction functions, then flexibility and efficiency are improved, but additional processing steps are required
Solution Approach 1:
The patent introduces an intermediary pre-processing step that converts various input data forms into a standardized format before feeding them to the neural network. This intermediary layer acts as a mediator that simplifies the overall system by providing a uniform interface to the trained model, improving productivity while managing complexity through modular design.
Data Source
AI summary
A computer-implemented method for creating image data with a trained function from measurement data recorded with a magnetic resonance system may include: providing a trained reconstruction function, which receives magnetic resonance data in a dedicated form as input data, to which the trained reconstruction function is applied and in the process output data comprising image data determines image data, loading recorded measurement data, processing the recorded measurement data into processed magnetic resonance data such that the processed magnetic resonance data is present in a form which corresponds to the dedicated form of the input data, receiving the processed magnetic resonance data as input data, applying the provided trained reconstruction function to the received input data, wherein output data comprising image data is determined, and providing the output data.