Application Of Tract-based Spatial Statistics Analysis In The Diagnosis And Monitoring Of Parkinson’s Disease

Parkinson’s disease(PD)is a common nervous system degenerative disease in the aged,which mainly manifested as resting tremor,bradykinesia,rigidity and postural balance disorder.According to statistics,it is about 1.7% the prevalence rate of PD in the aged people over 65 years old in our country,which is similar to that of the international prevalence rate.Nowadays,with the growing aged population in China,the incidence of PD is also increasing.The increasing in the morbidity of PD not only seriously affects the work,life and social activities of the patients,but also brings heavy burden to the family and the society.Previous studies have shown that PD is characterized by loss of dopaminergic neurons in the substantia nigra and dysfunction of striatum,the pathological change of that has led to motor symptoms in PD patients.In recent years,PD patients have been found to be complicated with non-motor symptoms(NMS),such as sleep disorders,pain,olfactory disorder and depression etc,which are being paid attention to by doctors and researchers and suggesting that the pathological changes of PD may be more wide.Thus,the Parkinson disease and dyskinesia group of the Chinese Medical Association,the neurology branch of the Chinese Medical Association,took NMS as one of the diagnostic criteria of PD in China in 2016 and emphasized the importance of NMS in the diagnosis of PD,especially in the early diagnosis of PD.It had been found that early PD not only involves the nigrostriatal system,but also involves the white matter system and other brain regions.The abnormal function of these regions may be an objective biological marker for the early diagnosis of PD and the monitoring of the course of disease.Positron emission tomography(PET)and single photon emission computedtomography(SPECT)can reflect the function of dopamine transporter in the brain,which are reliable biomarkers for neuroimaging.However,its application is often restricted by expensive inspection fees,insufficiently widespread use of equipment,and injection of radionuclide tracers.In recent years,with the developing of functional magnetic resonance imaging and molecular biology,a series of PD biomarkers have been found by researchers,which provide the possibility of early diagnosis,course monitoring and individualized management of PD.Some scholars have been using diffusion tensor imaging(DTI)to delineate the region of interest(ROI)for studying the white matter of the brain in PD.Because of the individualized differences among ROI,the repeatability of the research results is poor and the scholars are no longer admired.Meanwhile,some scholars used voxel based analysis(VBA)to show the location and size of different brain regions,but the post-processing step of VBA Gauss smoothing which is used to process the DTI image distorts the original data.Therefore,the research for new and reliable biomarkers of imaging is still under exploration.It has been found that the tract-based spatial statistics(TBSS)method is not necessary to smooth the registration error and greatly optimize the analysis results,because it has extracted the most consistent core white matter skeleton of the population and removed the brain area with larger registration error.Therefore,the accuracy and sensitivity of TBSS are greatly improved and the monitoring of abnormal areas of white matter microstructure is more reliable.In this study,we used TBSS to analyze the microstructure of white matter in PD patients with different courses and explore whether the TBSS method may be used as an imaging biomarker of PD diagnosis and monitoring.Methods:(1)30 PD patients in our hospital from September,2015 to December,2017,were collected,which include 16 cases of early stage and 14 cases of advanced stage,and 15 healthy volunteers with matched sex and age were selected.The H-Y staging and UPDRS score were carried out in the PD patients.(2)The whole brain T1 structure image and DTI function image were obtained by GE 1.5T TWIN SPEED dual gradient MR scanner(maximum gradient field intensity40mT/m,gradient switching rate 150T/m/s)and 8 channel orthogonal head coils.(3)A data processing software PANDA is used to preprocess DTI data.We compared brain white matter in early stage with normal control group,in advanced stage with normal control group and in early stage with advanced stage group by TBSS.The matrix design was two-sample t-test,and the difference between groups was compared.(4)The fractional anisotropy(FA)value of the brain region with statistical significance was obtained by PANDA software.The FA values of early differential brain regions and advanced different brain regions were made correlation analysis with the scores of UPDRS scale respectively.Correlation analysis is carried out between FA value in different brain area of early stage PD group and its H-Y classification,and between FA value in different brain area of advanced PD group and its H-Y classification.Results:(1)Compared with the normal control group,the fiber bundle of FA decreased in early PD group,including the body of corpus callosum,splenium of corpus callosum,middle cerebellar peduncle,left posterior thalamic radiation,genu of corpus callosum,right posterior thalamic radiation,right posterior corona radiata,left anterior corona radiata,right superior corona radiata,right external capsule,right anterior limb of internal capsule,left anterior limb of internal capsule,and right superior longitudinal fasciculus as well(P<0.05,FWE correction,voxel>500).(2)Compared with the normal control group,the fiber bundle of FA decreased in advanced PD group,including genu of corpu callosum,body of corpus callosum,right anterior coron radiata,left anterior coron.radiata,right posterior thalamic radiation,right superior coron radiata,right posterior corona radiata,right anterior limb of internal capsule,right external capsule,left superior corona radiata,left anterior limb of internal capsule,and left posterior thalamic radiation as well(P<0.05,FWE correction,voxel>500).(3)Compared with the early PD group,the fiber bundle of FA decreased in advanced PD group,including right superior corona radiata,right anterior corona radiata,body of corpus callosum,posterior thalamic radiation,left posterior corona radiata,left superior corona radiata,left anterior corona radiata,body of corpus callosum,and genu of corpus callosum(P<0.05,FWE correction).(4)The Pearson correlation analysis showed that the FA value of early PD group was not correlated with the UPDRS-I and UPDRS-Ⅲ scale score(P>0.05).The FA value of advanced PD group was not correlated with the UPDRS-I and UPDRS-Ⅲ scale score(P>0.05).The Spearman correlation analysis showed that the FA value of early PD group was negatively correlated with the H-Y classification(r=0.96,P<0.05).The FA value of advanced PD group was negatively correlated with the H-Y classification(r=0.92,P<0.05).Conclusion:(1)Extensive white matter fiber bundles were damaged in PD patients,and the damage of the white matter fiber bundles was more severe in advanced PD patients,which suggested that the white matter fibers were impaired in the early stage of PD and that increased with the progressing of the disease.(2)The FA value of the damaged white matter fiber bundle in PD patients was negatively correlated with the H-Y classification,which indicated that the damage of white matter increased with the progressing of the disease and suggested that TBSS may be an imaging biomarker of evaluating the progressing of PD.(3)The FA value of impaired white matter fiber bundle in PD patients was not correlated with the score of UPDRS-I and UPDRS-III scale,which suggested that the severity of white matter damage in PD patients can not be assessed by UPDRS scale.