Since its introduction in the 1970s, magnetic resonance imaging (MRI) has evolved from an experimental technique to a commonly used imaging modality in both clinical and research settings. Despite its wide adoption, MRI suffers from a major drawback of being slow. With the development of receiver arrays in the 1990s and the introduction of parallel imaging in early 2000s, new doors have been opened towards speeding up MRI. Spatially varying sensitivities of multiple receiving coils in a receiver array provide spatial encoding capabilities for MRI. Parallel imaging techniques speed up MRI exams by acquiring less data. The missing data is recovered using the spatial encoding provided by the coil sensitivities. In late 2000s, a new parallel imaging technique, simultaneous multi-slice (SMS) imaging, has been introduced. SMS imaging accelerates MRI acquisition along two orthogonal spatial dimensions. In the 2010s, this technique has been gaining increasing popularity especially in brain applications because of its high efficiency in accelerating the data acquisition process.
My talk will focus on data reconstruction problems for SMS MRI. In the first part of the talk, I will introduce a three-dimensional signal model for SMS imaging. This signal model serves as a foundation for my work. In the second part, I will present a hybrid-space sensitivity encoding (SENSE) algorithm for data reconstruction in SMS imaging. I will demonstrate that this algorithm can reconstruct SMS data acquired with arbitrary phase encoding patterns. Furthermore, I will describe an analytical method to quantify the signal-to-noise performance of the hybrid-space SENSE algorithm. In the third part of the talk, I will show how to incorporate an artifact correction method into the hybrid-space SENSE reconstruction framework. The artifact, termed Nyquist ghosting artifact, is intrinsic to echo planar imaging (EPI), which is a widely used data acquisition scheme for brain MRI. I will demonstrate that the proposed method can conduct a slice-specific Nyquist ghosting correction for SMS EPI acquisitions. Throughout the talk, brain images will be shown to validate the proposed methods.