Target Audience

MR scientist/engineers and clinicians interested in accelerated imaging.

Outcome/Objectives

Understand the design principles in k-space undersampling schemes and how to tailor them to various imaging sequences and reconstruction schemes, to achieve faster and higher quality imaging.

Principles

Over the years, various image reconstruction approaches (e.g. parallel imaging, CS, prior information, ML) have been developed to reduce the image encoding burden in MRI through k-space under-sampling. Optimal performance for these reconstructions at high undersampling require the use of a tailored sampling strategy. In this course, we will examine these sampling strategies and how to adopt them effectively in a wide variety of imaging sequences under different constraints, including those in structural imaging, fMRI, diffusion, time-series, multi-contrast, and quantitative imaging. Insights on how to flexibly take advantage of the undersampling for either scan time reduction and/or artifact mitigation will be discussed. Similarity and differences in undersampling strategies for 2D, simultaneous multi-slice and 3D volumetric acquisitions will also be examined, along with approaches for optimizing spatio-temporal sampling. Examples will be provided to demonstrate the benefit of a synergistic design approach on sampling, sequence, and reconstruction.

Acknowledgements

No acknowledgement found.

References

No reference found.