Mobile MRI Denoising Network for Rapid High-Quality Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile magnetic resonance apparatuses produce inferior imaging quality due to lower magnetic field strength, and increasing scanning time to improve quality is not acceptable for patients with strict time requirements, such as those with acute ischemic cerebral stroke.
Innovation Solution
An imaging method using a mobile magnetic resonance apparatus that involves randomly sampling K-space data and inputting it into a pre-trained denoising reconstruction network, generated through fully sampled training data, to produce high-quality magnetic resonance images in a short period.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If scanning time is increased to improve imaging quality of mobile magnetic resonance apparatus, then image quality is improved, but examination time becomes too long for patients with strict time requirements
Solution Approach 1:
A pre-trained denoising reconstruction network is constructed beforehand using fully sampled training data. This network is then applied during actual scanning to rapidly reconstruct high-quality images from randomly sampled data, eliminating the need for prolonged scanning while maintaining image quality
Solution Approach 2:
The traditional mechanical approach of extending scan time to improve quality is replaced with a computational approach using deep learning. The pre-trained network performs signal processing and image reconstruction computationally, achieving high quality images in short scan times
2Loss of time
If random sampling is used to reduce scanning time, then examination time is reduced, but imaging quality deteriorates due to lower magnetic field strength
Solution Approach 1:
A pre-trained denoising reconstruction network serves as an intermediary between the randomly sampled K-space data and the final image. This network processes the incomplete data and reconstructs high-quality images, bridging the gap between fast scanning and high image quality
Solution Approach 2:
The approach changes the sampling strategy from traditional Cartesian sampling to random sampling in the time domain, which corresponds to spiral or radial trajectories in K-space. Combined with the pre-trained network, this enables fast imaging without quality loss
Data Source
AI summary
The present disclosure provides an imaging method using a mobile magnetic resonance apparatus, which includes: randomly sampling an object-to-be-scanned using a mobile magnetic resonance apparatus to acquire encoding of the frequency and phase of the tissue of the object-to-be-scanned, and obtain target K-space data; inputting the target K-space data and pre-generated denoising reconstruction network parameters into a pre-trained denoising reconstruction network; where the pre-trained denoising reconstruction network is generated by training based on fully sampled training data; and outputting a magnetic resonance image corresponding to the object-to-be-scanned. The fully sampled training data is constructed by increasing the scanning time in the present application, the pre-trained denoising reconstruction network is obtained through model training based on the training data, real-time data is obtained through random sampling in a short period of time in practical applications, and high-quality images are output in combination with the pre-trained denoising reconstruction network.