Research On Magnetic Resonance R2* Parameter Based On Morphologic Distribution Of Hepatic Fat Droplets

Hepatic iron overload and nonalcoholic fatty liver disease（NAFLD）are common chronic liver diseases.Research has found that about one-third of NAFLD patients are presented with disorders of iron regulation.Iron overload accelerates the progression of NASH to cancer and contributes to steatosis in hepatocytes.Proton Density Fat Fraction（PDFF）and MRI transverse relaxation rate R2*,measured simultaneously by chemical shift-encoded magnetic resonance imaging,are clinical biomarkers for quantifying liver fat content and iron overload,respectively.However,clinical studies have found a positive correlation between R2*parameters and PDFF in patients without iron overload,which can lead to false positive liver iron diagnoses or poor fat quantification when using magnetic resonance imaging to quantify liver iron or liver fat in patients with both hepatic steatosis and liver iron overload.Therefore,research on the underlying biophysical mechanisms between fat fraction and R2*,and Calibrated for the effect of fat to R2*parameters can improve the quantification accuracy of liver fat and iron deposition in patients with chronic liver disease.Therefore,to address the above issues,this dissertation focuses on the simulation research of MRI R2*parameters.Especially,we investigated the morphological distribution of fat droplets and the correlation between the R2*parameter-PDFF.The research contents in this study include the following three aspects:（1）Morphological analysis of hepatic fat droplets based on stereology and spatial statistics.Fat droplets（i.e.cytoplasmic vacuoles）are generally considered to be the white part of the tissue section in the histological evaluation of hepatic steatosis.This research analyzed 30 liver biopsy samples with different fat fraction（FF）and quantitatively investigated the morphological distribution of fat droplets in two-dimensional（2D）plane and three-dimensional（3D）space（including size,nearest neighbor distance and area anisotropy）using stereology and spatial statistics.Results demonstrated that the morphological distribution of fat droplets can be described by the gamma distribution function（GDF）（R2>0.54）,and that the estimated GDF parameters（i.e.,scale and shape parameters）and FFs were well correlated,with²>0.55.（2）Virtual liver modeling based on morphological distribution of fat droplets.Virtual liver model with hepatic steatosis can be used for liver MR signal simulation,but this model requires accurate information about the underlying tissue properties.This research constructed four kinds virtual liver tissue models by controlling the effect of fat droplet morphological distribution on fat droplets based on the correlation between the GDF fitting parameters and FF.Comparative analysis and quantitative assessment of the model（determined by fat droplet size,nearest neighbor distance and regional anisotropy）and real liver tissues were performed.Results showed that the virtual liver tissue model and the real liver tissue demonstrated high similarity both qualitatively and quantitatively.（3）Research on magnetic resonance R2*based on Monte Carlo simulation.Clinical trials have found a correlation between R2*parameters and PDFF,but the underlying biophysical mechanisms have not been fully explained.The effects of different fat magnetic susceptibility coefficientson the magnetic resonance signals were investigated by the constructed virtual liver tissue model using Monte Carlo simulation at the magnetic field of 1.5 T and 3.0 T.Results indicated that the correlation between the R2*parameter and PDFF was most similar to the clinical trial with magnetic fields of1.5T and 3.0T and fat magnetic sensitivity coefficientsχof 0.5ppm and 0.6ppm,respectively...