Chemical Exchange Saturation Transfer Imaging And Reconstruction Method For Water-fat Separated Imaging Based On Single-shot Spatiotemporal Encoding

Ultrafast imaging is one of the hotspots in the field of magnetic resonance imaging(MRI),whose good temporal resolution can not only improve the working efficiency of magnetic resonance imaging and reduce the motion artifacts,but also promote the emergence of many dynamic imaging techniques.As a widely used ultrafast imaging technique,planar echo imaging(EPI)has excellent performance in both temporal and spatial resolution.However,especially in the high field conditions,EPI images are easily affected by artifacts and distortions because of its susceptibility to inhomogeneous fields and chemical shift effect.The single-shot spatiotemporally encoded MRI technique proposed by the Frydman and co-workers can not only preserve the advantage of EPI in terms of temporal resolution,but also have a strong ability to resist inhomogeneous fields and chemical shift effect.With the aid of the super-resolved reconstruction algorithms,SPEN MRI can reconstruct the image with the spatial resolution comparative to that of EPI.With these advantages,SPEN MRI is quickly applied in many areas of dynamic imaging.In this paper,we focus on a series of studies on imaging methods and applications of single-scan SPEN MRI.The main work includes the following three parts:1.The imaging principle,methods and characteristics of single-shot SPEN MRI are expounded comprehensively,including the generation mechanism of quadratic phase,local spin contribution of SPEN signal,gradient requirements for complete decoding and the mechanism of resisting inhomogeneous field and chemical shift effect.The necessity and principle of super-resolved reconstruction are systematically analyzed and five super-resolved reconstruction algorithms which have been proved to provide good results are introduced,namely the conjugate gradient method with singular value decomposition(SVD)preprocessing,deconvolution,partial Fourier transform Method,hybrid algorithm that combined random sampling with compressed sensing,and super-resolved enhancing and edge deghosting algorithm.2.A chemical exchange saturation transfer(CEST)imaging method based on single-shot SPEN MRI was proposed.CEST imaging and its extended nuclear overhauser enhancement(NOE)effect have been proved to process the ability to detect metabolites and identify tumor.We reconstruct the super-resolved SPEN image with a l1 norm optimization problem that combines total variation(TV)regularization and edge-weighted constraints.The experimental results show that this new CEST-SPEN MRI method can not only maintain the ability of rapidly imaging,but also achieve better CEST/NOE contrast image and more accurate Z spectrum compared to CEST-EPI thanks to its ability to resist inhomogeneous field and chemical shift effect.3.A new method for water-fat separated reconstruction based on single-shot SPEN MRI was proposed.When solving a linear equation system containing water and fat based on one single echo,the spatial resolution of the separated image is halved by the under-determination of the equation system.We exploited the point spread function of SPEN under the condition of water-fat to construct a new linear equation system and combined the prior knowledge related to chemical shift to reconstruct the water/fat separated image.The experimental results show that the new method can provide water/fat separated image with higher spatial resolution,less artifacts and better signal-to-noise ratio compared to the traditional conjugate gradient method.