Multi-echo MRI Pulse Sequence with Deep Neural Network Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional quantitative magnetic resonance imaging techniques face challenges in achieving fast acquisition times while maintaining high spatial resolution, and are susceptible to image distortion caused by inhomogeneous magnetic fields.
Innovation Solution
A method and system for fast high-resolution multi-parametric quantitative magnetic resonance imaging using a fast high-resolution multiple overlapping-echo imaging pulse sequence combined with a deep neural network for image reconstruction, which resists image distortion from inhomogeneous magnetic fields.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional quantitative magnetic resonance imaging collects multiple images with different parameter-weighting for fitting, then measurement precision is improved, but acquisition time increases and motion artifacts occur
Solution Approach 1:
The patent segments the acquisition process by using a single multi-echo sequence that simultaneously captures multiple parameter-weighted signals (T2, T2*, PD) in one scan, rather than acquiring separate images for each parameter. This segmentation allows quantitative measurement of multiple parameters while reducing total acquisition time and eliminating motion artifacts between scans.
Solution Approach 2:
The patent merges multiple parameter-weighted image acquisitions into a single multi-echo pulse sequence that collects T2, T2*, and PD-weighted signals simultaneously. By combining these measurements in one scan, the method achieves quantitative precision for multiple parameters without the time penalty and motion artifacts of sequential acquisitions.
2Productivity
If planar echo imaging is used to shorten acquisition time, then productivity is improved, but manufacturing precision deteriorates due to limited spatial resolution and inhomogeneous magnetic field effects
Solution Approach 1:
The patent changes the imaging parameter from planar echo to spin echo, which fundamentally alters the signal acquisition mechanism. Spin echo is less susceptible to magnetic field inhomogeneity, thereby maintaining high spatial resolution and image quality while still achieving fast acquisition through the multi-echo sequence design that captures multiple parameters simultaneously.
3Measurement precision
If multiple overlapping-echo signals are collected, then measurement precision is improved, but device complexity increases due to signal overlap and reconstruction complexity
Solution Approach 1:
The patent introduces a deep neural network as an intermediary between the raw multi-echo signal data and the final quantitative parameter maps. This neural network mediator automatically separates and processes the overlapping echoes, extracting T2, T2*, and PD values without requiring complex traditional signal processing algorithms, thereby reducing computational complexity while maintaining high measurement precision.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The proposed method achieves high-resolution multi-parametric quantitative magnetic resonance images within the same acquisition time as fast spin-echo weighted imaging, while providing resistance to image distortion caused by inhomogeneous magnetic fields.
Implementation Method 1
Magnetic resonance imaging can image biological tissues without invasion and injury
Implementation Method 2
slice selection gradients Gss corresponding to the N RF excitation pulses, and echo shift gradients Gn... phase encoding gradients Gpe,i,m, frequency encoding gradients Gro
Data Source
AI summary
A method for fast high-resolution multi-parametric quantitative magnetic resonance imaging comprises: designing a fast high-resolution multiple overlapping-echo imaging pulse sequence; determining sampling parameters of the pulse sequence; constructing a deep neural network for reconstructing high-resolution multi-parametric quantitative magnetic resonance images; generating training samples of the deep neural network; using the training samples to train the deep neural network to obtain trained deep neural networks; scanning a real imaging object using the pulse sequence under the sampling parameters to obtain k-space data of the real imaging object; pre-processing the k-space data of the real imaging object to obtain image domain data of the real imaging object; and inputting the image domain data of the real imaging object into the trained deep neural networks for the reconstructing to obtain the high-resolution multi-parametric quantitative magnetic resonance images of the real imaging object.