MRI View Ordering for Periodic Signal Modulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional view-ordering techniques for MRI systems with multiple phase encode directions and periodic signal modulation are limited in scanning efficiency, incompatible with non-rectangular k-space grids, and do not support auto-calibrated 2D parallel imaging, leading to artifacts and increased scan time.
Innovation Solution
A method that assigns readout numbers and train numbers to MRI views based on their positions in k-space, allowing for non-rectangular k-space coverage and 2D auto-calibrated parallel imaging, thereby improving sampling efficiency and reducing image artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional view-ordering techniques are used for pulse sequences with multiple phase encode directions and periodic signal modulation, then the acquisition can be performed with standard rectangular k-space grids, but scanning efficiency is limited and scan time is increased
Solution Approach 1:
The patent applies dynamics by making the view ordering adaptive rather than fixed. The system dynamically adjusts the assignment of views to trains based on the periodic signal modulation characteristics, allowing the acquisition strategy to optimize itself for different modulation patterns and k-space trajectories, thereby improving scanning efficiency without increasing scan time
Solution Approach 2:
The patent changes the parameter of view ordering strategy from conventional fixed patterns to optimized assignments based on periodic signal modulation. By varying how views are grouped into trains according to their position in k-space and the modulation characteristics, the system achieves better sampling efficiency and reduced scan time
2Adaptability or versatility
If conventional view-ordering techniques are used, then the implementation is straightforward with regular k-space sampling, but the technique is incompatible with non-rectangular k-space grids and parallel imaging
Solution Approach 1:
The patent achieves universality by creating a view ordering technique that works across multiple acquisition types including rectangular and non-rectangular k-space grids, parallel imaging sequences, and pulse sequences with periodic signal modulation. The unified approach assigns views to trains based on general principles that adapt to different k-space sampling patterns, making the technique compatible with diverse MRI acquisitions without requiring separate specialized methods for each case
Solution Approach 2:
The patent applies segmentation by dividing the k-space acquisition into multiple trains of views, where each train is assigned a specific modulation pattern. This segmentation allows the system to handle complex acquisition requirements by breaking down the overall k-space sampling into manageable segments that can be acquired with different ordering strategies optimized for their specific characteristics
3Object-affected harmful factors
If views are acquired with large jumps in k-space between successive views, then the acquisition can be completed quickly, but image artifacts are produced due to erratic phase behavior
Solution Approach 1:
The patent applies preliminary action by pre-planning the view ordering to minimize k-space jumps before acquisition begins. The system calculates and assigns views to trains in advance, ensuring that successive views within and across trains follow smooth trajectories through k-space. This preliminary optimization prevents erratic phase behavior and image artifacts while maintaining efficient acquisition speed
Data Source
AI summary
A method for acquiring magnetic resonance (MR) data for a pulse sequence with periodic signal modulation and a set of views having at least two phase encode directions includes selecting a direction of modulation. Each view in the set of views is assigned a readout number based on a position of each view along the direction of modulation resulting in a plurality of readout number groups, each readout number group having a plurality of views. Each view within each readout number group is assigned a train number based on a position of each view along a second direction in k-space. MR data for the set of views is acquired based on the assigned readout number and train number for each view.