Related Experiment Videos
A generalized approach to parallel magnetic resonance imaging
1Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA. dsodicks@caregroup.harvard.edu
Medical Physics
|September 11, 2001
Summary
Parallel magnetic resonance (MR) imaging accelerates scans using multiple coils. New generalized methods and conditioning techniques improve image quality and robustness, enabling higher acceleration rates.
Area of Science:
- Medical Imaging
- Physics
- Engineering
Background:
- Parallel magnetic resonance (MR) imaging accelerates data acquisition by using multiple radiofrequency detector coils.
- Existing parallel imaging techniques like SMASH and SENSE have limitations in performance and robustness.
Purpose of the Study:
- To derive a generalized formulation for parallel MR imaging.
- To develop new algorithms and hybrid reconstruction approaches for improved performance.
- To enhance the practical robustness of parallel image reconstructions.
Main Methods:
- A generalized mathematical framework for parallel MR imaging was developed.
- Hybrid reconstruction algorithms combining SMASH-like and SENSE-like methods were constructed.
- Numerical conditioning techniques were incorporated to improve reconstruction robustness and eliminate sensitivity calibration steps.
Main Results:
- The generalized formulation unified existing parallel imaging techniques and suggested novel algorithms.
- Hybrid approaches with numerical conditioning demonstrated improved performance and robustness.
- The methods extended the range of accelerations for obtaining high-quality parallel MR images.
Conclusions:
- The generalized formulation provides a unified view of parallel MR imaging techniques.
- Hybrid reconstruction strategies combined with numerical conditioning significantly enhance parallel imaging capabilities.
- These advancements enable higher acceleration factors for faster, high-quality MR imaging.