Adversarial Learning for MRI Parameter Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic resonance imaging (MRI) apparatuses produce varying image characteristics even with the same imaging parameter settings due to differences in apparatus models and individual characteristics, making it difficult to achieve consistent image quality across different MRI systems.
Innovation Solution
An information processing apparatus employing adversarial learning to generate and discriminate medical images, adjusting imaging parameter groups to converge on settings that produce images with desired characteristics, similar to target images, regardless of the MRI apparatus used.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the same imaging parameter settings are used across different MRI apparatuses, then the operation is simplified, but the image characteristics such as contrast and resolution vary due to apparatus-specific differences
Solution Approach 1:
The system automatically adjusts imaging parameters based on the specific MRI apparatus being used. The parameter adjustment unit modifies at least one imaging parameter from a predetermined set based on apparatus-specific characteristics, ensuring that image characteristics remain consistent across different machines while maintaining ease of operation.
2Manufacturing precision
If imaging parameters are manually adjusted for each MRI apparatus to achieve consistent image characteristics, then image quality is improved, but the workload and time required increase significantly
Solution Approach 1:
The MRI apparatus performs automatic parameter adjustment without requiring manual intervention from operators. The parameter adjustment unit automatically selects and applies appropriate imaging parameters based on the specific apparatus characteristics, enabling the system to self-adjust and eliminate time-consuming manual calibration processes.
Solution Approach 2:
The system automatically modifies imaging parameters based on apparatus-specific characteristics, eliminating the need for manual parameter adjustment while maintaining consistent image characteristics across different MRI machines.
3Manufacturing precision
If multiple trial and error attempts are made to find appropriate imaging parameters for each apparatus, then image quality is improved, but productivity decreases due to repeated adjustments
Solution Approach 1:
The system pre-stores multiple sets of imaging parameters corresponding to different apparatus characteristics. Instead of requiring trial and error during actual operation, the appropriate parameter set is already prepared and can be automatically selected and applied, significantly improving workflow efficiency while maintaining image quality.
Solution Approach 2:
The system automatically selects and applies pre-prepared parameter sets based on apparatus characteristics, eliminating the need for trial and error adjustments during imaging workflows and thereby improving productivity while maintaining consistent image quality.
Data Source
AI summary
An information processing apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to perform an image generating process of generating a plurality of medical images on the basis of a plurality of imaging parameter groups, the plurality of imaging parameter groups being updated by adversarial learning. The processing circuitry is configured to perform a discriminating process of discriminating whether or not each of the plurality of medical images is a target image acquired by a first medical imaging apparatus in the adversarial learning. The processing circuitry is configured to output one of the imaging parameter groups used for generating the medical images in the image generating process when the adversarial learning has converged.