Cardiac MRI Reconstruction Without Outer Volume Suppression Pulses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current real-time cardiac MRI techniques face limitations in achieving higher acceleration rates due to the aliasing of extra-cardiac tissue into the heart, which is exacerbated by the use of outer volume suppression pulses that disrupt steady-state conditions.
Innovation Solution
A method using deep learning-based outer volume removal techniques to estimate and subtract ghosting artifacts from composite k-space data, followed by reconstructing images using physics-driven deep learning models to achieve higher acceleration rates without temporal regularization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If outer volume suppression pulses are used to reduce aliasing artifacts, then image quality improves, but steady-state conditions are disrupted and acquisition time increases
Solution Approach 1:
The patent extracts and removes outer volume signals from the acquired k-space data using deep learning-based reconstruction. Instead of using suppression pulses during acquisition, the method separates and eliminates aliasing artifacts from extra-cardiac tissue in the reconstruction phase, preserving steady-state conditions during data acquisition.
Solution Approach 2:
The patent introduces deep learning-based outer volume removal as an intermediary step between data acquisition and final image reconstruction. This intermediary process uses trained neural networks to identify and remove aliasing artifacts, enabling higher acceleration rates without compromising image quality or steady-state conditions.
2Productivity
If higher acceleration rates are used to reduce scan time, then productivity improves, but aliasing artifacts from outer volume increase
Solution Approach 1:
The patent converts the harmful aliasing artifacts into useful information by training deep learning models on data with known outer volume characteristics. The model learns to recognize patterns of aliasing and systematically removes them, enabling higher acceleration rates where the artifacts would normally degrade image quality.
Solution Approach 2:
The patent changes the parameter of acceleration rate beyond traditional limits by combining parallel imaging with deep learning-based outer volume removal. This allows acceleration rates higher than what conventional parallel imaging can achieve while maintaining image quality through artifact removal in the reconstruction phase.
3Loss of time
If parallel imaging is used to increase acceleration, then scan time decreases, but spatiotemporal resolution is limited
Solution Approach 1:
The patent combines multiple imaging techniques into a composite approach: parallel imaging for acceleration, deep learning for artifact removal, and outer volume suppression through reconstruction. This composite methodology achieves higher acceleration rates while preserving spatiotemporal resolution that would be lost using any single technique alone.
Data Source
AI summary
A method for magnetic resonance image reconstruction with outer volume removal includes accessing timeframes of k-space data acquired using time-interleaved undersampling patterns in k-space. A composite image is generated from the k-space data by combining timeframes of the k-space data. A machine learning model—trained on training data to extract ghosting artifact signal components from a magnetic resonance image—is used to generate a ghosting artifact image by inputting the composite image data to the machine learning model. Outer volume signals are estimated by subtracting the ghosting artifact image from the composite image. Outer volume removed k-space data are generated by removing the outer volume signals from the k-space data. One or more images are reconstructed from the outer volume removed k-space data.