Research On Motion Correction In Multishot Diffusion Imaging And Magnetic Resonance Fingerprinting

Since its invention,magnetic resonance imaging(MRI)has been widely used in disease diagnosis and research on brain function due to the advantages of high resolution,no radiation,any tomography,multi-contrast imaging and so on,and become more and more improtant.However,the disadvantage of long scan times makes it sensitive to subject’s motion which will lead to image artifacts and further affect doctor’s diagnosis.Though repeated scans and the need for anesthesia can alleviate or eliminate motion-induced image artifacts,this incurs a significant financial cost.Therefore,it is still an important topic how to eliminate the effect of patient motion during scans or postprocessing on image quality.Quantitative MRI with high sensitivity and tissue specificity has very important value of clinical diagnosis and research.Quantitative diffusion imaging utilizes the diffusion of water molecular to obtain image and has been widely used for disease diagnosis in clinic and research on brain.However,the conventional single-shot diffusion imaging suffers from low spatial resolution and geometric distortion.Recently,multishot diffusion imaging has been used in clinic to overcome the litimations of the conventional single-shot DWIL Furthermore,the quantitative magnetic resonance fingerprinting(MRF)permits the simultaneous and fast non-invasive quantification of multiple important properties of tissue.Though a lot of work have been done to improve the image quality for these two quantitative techniques,there are little reports on motion correction.Therefore,we did the following three aspects of work to correct for the potential motion during scans for these two quantitative techniques:1.To correct for the position mismatch during the multishot diffusion acquisition,an iterative motion compensated reconstruction was proposed.In the proposed method,navigator following image data was acquried to estimate motion-induced phase errors and motion information.Then,these estimated phase and motion information were integrated into an iterative imaging reconstruction framwork to eliminate the effect of potetial motion on diffusion images.The proposed method was compared against image reconstruction using image-space sampling functions(IRIS)technique and the experimental results of simulated and in vivo multishot motion data demonstrated that the proposed method can suppress the motion artifacts in diffusion images and provide more accruate diffusion tensor imaging(DTI)estimations.2.To correct for motion-induced diffusion direction changes during the multishot acquisition,a clustering-based motion correction method was proposed.This method grouped shots into different clusters,inside which no larege scale motion occurred,through navigator images.Then IRIS was used to reconstruct each cluster image.Next,the diffusion directions of diffusion images were corrected by using the motion information estimated from these cluster images before DTI analysis.The simulated and in vivo multishot motion data demonstrated that the proposed clustering-based motion correction could reduce the motion-induced blurring and artifacts in DWI and improve the accuracy of DTI estimation.3.To correct for motion-induced inconsistient position of different sampled time points in MRF,the sliding-window-based motion correction was proposed.First,individual image was reconstructed by using sliding window reconstruction and then used to estimate motion information by image registration.Then the estimated motion information was used to correct for the motion-corrupted k-space data.Final,the original MRF reconstuction method was used to obtain parameters mapping.The performance of the proposed method was evaluated by simulation and in vivo data and the experimental results demonstrated that the proposed method can reduce the image blurring and artifacts and improve the accuracy of parameters estimation.