Array-Compressed Parallel Transmission for High Resolution Neuroimaging at 7T

项目总结

Project Summary The goal of this project is to develop a framework for high-performance parallel transmission (pTx) that is trans- ferable to a wide range of MRI scanners, and apply it to push the spatial encoding limits of echo planar imaging (EPI) at 7 Tesla. EPI is by far the most widely used pulse sequence for rapid functional, diffusion, and perfusion imaging, and has been the focus of considerable development in recent years to increase its speed and spatial resolution. Now there is a strong desire to push EPI's spatial resolution down to the micro scale. For functional MRI (fMRI), this would enable imaging of fine structures (layers, columns, and nuclei) of cortical and subcortical architecture while better resolving the hemodynamic response. For diffusion MRI (dMRI), micro scale EPI would improve surface and laminar analysis of fibers in the cortex, as well as brain parcelation using fractional anisotropy differences between gray matter regions, while broadly reducing partial volume effects. It would further enable EPI to be broadly applied to accelerate anatomic scans that are geometrically matched to fMRI and dMRI scans. However, increasing the resolution of single-shot EPI requires longer readouts which extend echo times and re- duce functional contrast in fMRI and signal-to-noise in dMRI at 7 Tesla, while increasing geometric distortions and blurring. Segmented or multishot EPI is a classic method to increase spatial resolution without increasing readout durations, but is underutilized, primarily due to its high sensitivity to motion and dynamic phase changes between shots which cause large image artifacts. We propose to develop a new multishot EPI technique called shuttered EPI, which addresses the lim- itations of conventional multishot EPI by imaging a set of spatially disjoint shutters in each shot. The shutters are produced by a multidimensional excitation pulse and are spatially shifted between shots to cover an entire slice. However, with thin slices the length of the excitation pulses are impractical (20-100 ms). Many-coil pTx (> 8 coils) can shorten the length of these pulses to feasible durations, but current 7 Tesla scanners have only 8 transmit channels due to cost, footprint, cabling, and other constraints. In the first project period we pioneered a technique called array-compressed pTx (acpTx) which overcomes this limitation. Using acpTx, 8 transmit chan- nels can control an arbitrarily large number of coils, where the channels and coils are connected via an array compression network that is optimized with RF pulses for specific excitations. In this project, we will develop and apply acpTx methods and hardware (a many-coil head transmit array and an 8 channel-to-many coil array com- pression network) to achieve feasible RF pulse durations when exciting the shutter patterns required for shuttered EPI. These developments will be implemented on two major 7T scanner platforms and evaluated in submillimeter (600 micron) fMRI and dMRI acquisitions. Overall, the project encompasses the synergistic design of RF pulses, hardware, acquisitions and reconstructions to achieve a major advance in spatial encoding.