Abstract

Parallel imaging refers to a set of techniques used to accelerate magnetic resonance imaging (MRI) acquisitions. Accelerated imaging is beneficial for increasing spatial resolution, temporal resolution, or both. Cartesian MRI is relatively slow due to the sequential acquisition of phase encode lines in k-space. Parallel imaging methods acquire a reduced amount of k-space data to decrease scan time. The acceleration factor, R, describes the amount of undersampling. For example, R=2 corresponds to acquiring every other k-space line, or 50% of the data. The acceleration factor must be less than or equal to the number of coils, but is often much lower in practice due to coil configurations.

Undersampled Cartesian k-space data yields images with aliasing, or fold-over, artifacts. Parallel imaging techniques can generally be divided into two categories. The first category operates in the image domain and aims to unfold aliased images. The second category operates in the k-space domain and aims to synthesize missing k-space data. All techniques require multiple radiofrequency (RF) receiver coils with different spatial sensitivities. This talk covers three common parallel imaging algorithms: SENSE¹, GRAPPA², and SPIRiT³.

Parallel imaging enables accelerated acquisitions at the price of signal-to-noise ratio (SNR). The SNR of reconstructed images is reduced by the square root of the acceleration factor and the coil geometry factor, or g-factor. The acceleration factor affects the SNR across the entire image, but g-factor is spatially varying and related to the coil configuration.

The goal of this presentation is to provide an introduction to parallel imaging methods and an understanding of the tradeoffs between acceleration, aliasing artifacts, and SNR.

Acknowledgements

No acknowledgement found.