The Study Of Brain White Matter Microstructural Alterations Based On Diffusion Tensor Imaging In Parkinson’s Disease With Depression

Background:Parkinson’s disease(PD)is the second most common neurodegenerative disease after Alzheimer’s disease.Depression is one of the most common mental disease of PD,with an incidence rate of 40%-50%.Depression often has a serious impact on the quality of daily life of PD patients,and some PD patients even have suicidal tendency.In clinical practice,however,the depressive symptoms of PD patients are often ignored or undertreated.The pathogenesis of PD with depression is not fully clear.It has been reported that the occurrence of depressive symptoms may be earlier than the motor symptoms of PD.Diffusion tensor imaging(DTI)technology is based on the characteristics of detecting the movement state of water molecules in vivo.DTI can be used to noninvasively evaluate the microstructure alterations of white matter,such as axonal destruction,demyelination,edema or necrosis of white matter.The main analysis parameters of DTI,such as fractional anisotropy(FA)and mean diffusivity(MD),which are the most commonly used and sensitive diffusion parameters.DTI has been widely used in the pathogenesis of neurodegenerative diseases(including common PD)due to its advantages.Studies have shown that depression is closely related to the abnormal white matter of PD.Purpose:To investigate white matter microstructural alterations in PD patients with depression using diffusion tensor imaging(DTI)by using atlas-based analysis(ABA).In addition,a support vector machine(SVM)was trained with DTI parameters(FA and MD values)of different white matter regions to explore its performance in identifying PD with or without depression.Materials and methods:All subjects underwent DTI examination and routine clinical scales assessment.Based on the MATLAB platform,the FA and MD values were extracted by PANDA software,and the differences of FA and MD values between depressed PD(dPD group)and non-depressed PD(ndPD group)were further compared by the ABA method.Correlation analysis was performed to assess the relations between DTI parameters and clinical severity of depression in dPD patients.Finally,a support vector machine(SVM)was trained to examine the probability of discriminating between dPD and ndPD using DTI parameters(FA and MD values)as classification features.Results:A total of 55 PD patients were included in this study,including 27 dPD patients and 28 ndPD patients.Compared with ndPD patients,the dPD patients exhibited a significantly lower FA in the right cingulum(cingulate gyrus),left cingulum hippocampus,left inferior longitudinal fasciculus and right superior longitudinal fasciculus,and increased MD in the bilateral cingulum(cingulate gyrus).Correlation analysis showed that the FA values of the right cingulum(cingulate gyrus)and left inferior longitudinal fasciculus were negatively correlated with the Hamilton Depression Rating Scale(HAMD)score.Receiver operator characteristic(ROC)curve analysis using white matter fiber diffusion indices significantly associated with clinical features showed that FA values along the right cingulum(cingulate gyrus)(AUC=0.738,p=0.0024)and the left inferior longitudinal fasciculus(AUC=0.766,p=0.0007)showed a moderate potential to differentiate dPD from ndPD.Finally,the SVM classification model exhibited moderate performance in discriminating dPD from ndPD.The accuracy,sensitivity and specificity of SVM classification model were 66.73%,77.72%and 61.07%,respectively.Conclusion:The present study suggested that depressed PD is associated with white matter microstructural integrity impairment.These findings may help to further understand the underlying pathogenesis of depressed PD.The SVM classification model based on DTI parameters has moderate accuracy for distinguishing dPD from ndPD.