Quantitative Evaluation Of Hepatic Sinusoidal Obstruction Syndrome Based On Diffusion And T1rho Magnetic Resonance Imaging

Hepatic sinusoidal obstruction syndrome(SOS)is a kind of microstructural injuries caused by microcirculation obstruction.It is accompanied by sinusoidal congestion,cellular edema,inflammation,necrosis and the accumulation of macromolecules in extracellular matrix.Diffusion magnetic resonance imaging(dMRI)can detect the speed of the random movement of water molecules in biological tissues.The quantitative parameters of dMRI can be used to characterize the microstructural variation of tissues and the microcirculation in capillaries.The spin-lattice relaxation time in the rotating frame(T1rho)is sensitive to low-frequency motional biological processes,and its quantitative parameter can be used for exploring the composition of macromolecules in tissues and the proton exchange between water molecule and macromolecules.However,the diffusion and Tlrho images are disturbed by severe noise,which may introduce large errors in the parameter fitting process,resulting in inaccurate quantitative parameters for the assessment of liver disease.In this thesis,we focus on the exploration of quantitative diffusion parameters in detecting and staging SOS,the optimization of non-Gaussian intravoxel incoherent motion(nG-IVIM)fitting data and the application of nG-IVIM on evaluating SOS,and the feasibility of the quantitative Tlrho value in monitoring the prognosis of SOS treatment.The research contents in this thesis as follows::(1)Evaluation of SOS induced hepatic heterogeneity by using non-Gaussian diffusion model.The stretched exponential model(SEM)and diffusion kurtosis imaging(DKI)are both used for characterization of dMRI signal decay caused by non-Gaussian diffusion of water molecules and microstructural heterogeneity in biological tissues.By analyzing the changes of quantitative parameters of SEM and DKI during the development of SOS and taking the pathological results as the reference.We confirmed that the severity of sinusoidal congestion,cellular edema,inflammation,necrosis and other microstructural injuries caused by SOS were correlated with the diffusion related parameters(DDC and Dapp)and heterogeneous related parameters(? and Kapp).The DDC and Dapp values in normal liver(1.343±0.111;1.615±0.149×10-3mm2/s)are significantly higher than that of early-stage SOS(1.145±0.073;1.421±0.098×10-3mm2/s;P<0.01)and late-stage SOS(0.978±0.091;1.200±0.100×10-3mm2/s;P<0.01).The results show that DDC and Dapp may be useful for SOS staging.(2)Optimization of nG-IVIM fitting data and the application of this model in SOS assessment.As a hybrid biexponential model,nG-IVIM can more accurately reflect the signal attenuation caused by non-Gaussian diffusion and blood perfusion in biological tissues than DKI or IVIM alone.However,due to the high complexity of the model and the low signal-to-noise ratio(SNR)of the high b-value image,the fitting parameters of the nG-IVIM model are subject to large deviation and fluctuation caused by noise.In order to improve the robustness of fitting parameters,a non-local mean combined with bias correction method is proposed.The non-local mean denoising is used to reduce the fluctuation of the fitting parameters,and the bias correction is used to improve the accuracy of the fitting parameters.The results show that the proposed method can effectively reduce the fitting errors of nG-IVIM parameters,and improve the diagnostic accuracy.(3)Optimization of Tlrho balanced gradient echo sequence and the correlation between pathological features and Tlrho value.The selection of flip angle in Tlrho sequence will affect the accuracy of quantitative parameter.In this research,the correlation between flip angle and the fitting error of quantitative Tlrho parameter is evaluated by simulated data.In addition,multiple regression analysis between T1rho values and pathological scores revealed that Tlrho were mainly affected by cellular inflammation and liver fibrosis during the development of SOS.The relation indicate that Tlrho could be a potential biomarker for monitoring the SOS treatment.