Parallel imaging acquisition and reconstruction methods for speeding up MRI and motion mitigation


We have developed a number of efficient MRI acquisition methodologies that optimize the inter-play between software algorithm, MR physic/hardware, and the underlying neuroscience. Through these developments we have been able to provide an order of magnitude improvement in MRI acquisition speed and efficiency for a wide range of brain imaging applications including functional MRI, diffusion imaging, susceptibility imaging, structural imaging and quantitative imaging through MRF.

To achieve such high efficiency, our acquisition schemes are specifically tailored to take advantage of the latest imaging hardware that our group has also worked on including ultra-high field MRI, high-channel count receiver and transmitter arrays, and high-strength Connectome gradients. Our acquisitions also utilized volumetric imaging through Simultaneous Multislice or full 3D acquisitions to achieve high SNR efficiency, in conjunction to rapid imaging speed.

A major component in our ability to achieve rapid imaging has been in the way we optimized our acquisition to take full advantage of the spatial encoding information that is available in modern multi-channel receiver arrays (such as 32 and 64 channels). Along this direction, we have also expanded the use of this spatial encoding information to help estimate and mitigate image corruptions that can result from hardware imperfection (such as gradient delay/eddy current) and subject motion during scanning.