Anatomical Section MRI Reconstruction Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MRI reconstruction methods face challenges in producing high-quality images from undersampled k-space data, leading to noise and artifacts, especially when imaging common anatomical structures like the brain, as they rely heavily on generic reconstruction techniques rather than incorporating specific anatomical knowledge.
Innovation Solution
The method incorporates anatomical knowledge by training reconstruction models such as convolutional neural networks and generative adversarial networks using MRI images and k-space data divided into anatomical sections, allowing for the identification of specific anatomical sections during scans and reconstructing images with reduced noise and artifacts from less data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If k-space samples are undersampled to save scan time, then scan time is reduced, but image quality deteriorates with more noise and artifacts
Solution Approach 1:
The system performs preliminary actions by acquiring scout scan data before the main MRI scan to identify anatomical sections. This preliminary anatomical identification enables the main scan to be reconstructed with higher quality using anatomical-specific models, resolving the contradiction by preparing necessary information in advance that allows quality reconstruction from undersampled data
Solution Approach 2:
The patent applies local quality by dividing the MRI data into different anatomical sections and applying specific reconstruction models tailored to each anatomical region. Instead of using a generic reconstruction method for the entire image, the system optimizes reconstruction quality for each local anatomical area, thereby maintaining high image quality even with undersampled data
2Device complexity
If generic reconstruction methods are used, then the reconstruction process is simple, but image quality deteriorates especially for common anatomical structures
Solution Approach 1:
The system implements local quality by selecting and applying different reconstruction models based on the specific anatomical section being reconstructed. Each anatomical section has its own optimized reconstruction model trained on relevant data, providing superior image quality for common anatomical structures compared to generic methods, while maintaining manageable complexity through automated model selection
Solution Approach 2:
The patent introduces an intermediary component - the anatomical identification module that processes scout scan data to determine which anatomical sections are present. This intermediary enables the system to automatically select appropriate reconstruction models without requiring complex manual configuration, thereby improving image quality while keeping the overall process complexity manageable
3Manufacturing precision
If more MRI data is collected to improve image quality, then image quality improves, but scan time increases
Solution Approach 1:
The system applies partial action by collecting only the necessary scout scan data for anatomical identification rather than acquiring full high-resolution data for all sections. This partial data collection approach enables the system to identify anatomical sections and apply appropriate reconstruction models, achieving good image quality without the need for extensive data collection that would increase scan time
Solution Approach 2:
By performing anatomical identification through scout scans as a preliminary action before the main reconstruction process, the system enables efficient data acquisition strategies. The preliminary anatomical information allows for optimized sampling patterns and reconstruction approaches that achieve high image quality without requiring excessive MRI data collection during the main scan
Data Source
AI summary
Methods and apparatus for MRI reconstruction and data acquisition are provided. The method for MRI reconstruction includes: obtaining MRI images and k-space training data and dividing into anatomical sections; training reconstruction models to predict MRI images from k-space data for individual anatomical sections; while scanning an object, identifying the anatomical sections by scout scans or navigator signals; selecting suitable reconstruction models; reconstructing anatomical sections using the selected models, and merging the images from anatomical sections. Reconstructed images obtained by the above methods and apparatus have better image quality such as lesser noise and artifacts, and less MRI data is needed for the same image quality.