Combined Quantitative Susceptibility Mapping And Diffusion Tensor Imaging To Diagnose Parkinson's Disease

Parkinson's disease(PD)is a neurodegenerative disease with early prominent death of dopaminergic neurons in the substantia nigra(SN)pars compacta.With the development of magnetic resonance imaging(MRI)technology,many MRI methods have been developed to diagnose PD,including quantitative susceptibility mapping(QSM)and diffusion tensor imaging(DTI).Specifically,the susceptibility value,fractional anisotropy(FA)and mean diffusivity(MD)may reflect different aspects of PD-related pathological processes.However,there are still some challenges in QSM reconstruction.In addition,sensitivity and specificity of diagnosing PD by using MRI need to be improved.To address the above issues,the following studies were carried out in this thesis.1.Improvement of multi-echo field fitting algorithm in QSMInaccurate field fitting occurs in the pixel whose phase has poor linearity with echo time.In this study,an improved multi-echo field fitting algorithm based on evaluation of phase linearity was developed to improve the accuracy of field map.Our results showed that the proposed algorithm can reduce imaging artifacts and noise in susceptibility map.2.Improvement of phase unwrapping algorithm in QSMThe phase unwrapping algorithms used in QSM include PRELUDE and nonquantitative Laplace methods.PRELUDE can provide accurate unwrapped field map,but it takes too long time.Although non-quantitative Laplace algorithm takes very short time,it underestimates the susceptibility value.Thus,an improved algorithm combining Cusack's method and quantitative 2D Laplace algorithm was developed to unwrap the field map.The results showed that there was no significant difference between the susceptibility values calculated with our improved algorithm and PRELUDE method,and our improved algorithm took less time in unwrapping procedure.3.Research on diagnosing PD with QSM and DTIThis study was designed to evaluate sensitivity and specificity of combining QSM and DTI in diagnosing PD.QSM and DTI images from 16 patients with PD and 15 healthy controls were obtained on a 3T MRI system.Specifically,our improved field fitting algorithm and unwrapping algorithm were used to obtain susceptibility map.Bilateral SNs were delineated manually on the susceptibility map and co-registered to FA and MD maps.Finally,the mean susceptibility,FA and MD values in each SN were calculated and compared between PD and controls.Compared to controls,PD subjects demonstrated increased susceptibility but no significant difference of FA or MD in the SN.