The Study Of Quantitative Susceptibility Mapping And Radiomics In Parkinson's Disease

Part ? Detection of iron deposition in Parkinson's disease by quantitative susceptibility mapping Objective:Quantitative susceptibility mapping(QSM)measuring the brain's gray matter nuclei iron deposition in patients with Parkinson's disease(PD)on patients with different subtype movement and the change rule of brain nuclei iron deposit and the correlation between the disease development,speculated that the role of iron deposit in the pathogenesis of Parkinson's disease.Methods :36 patients with Parkinson's disease(PD group)admitted to our hospital from August 2016 to October 2019 were selected,including 19 males and 17 females,aged 65.14±8.22 years.There were 19 normal controls(NC group)matched by age and sex,including 10 males and 9 females,aged 61.42±4.87 years.Clinical data including course of disease,unified Parkinson's disease rating scale(UPDRS),Hoehn-Yahr,Moc A,and MMSE scores were collected for all patients 12 hours after drug withdrawal.According to UPDRS-III,the patients were divided into Postural instability disorder group(Postural flow disorder,PIGD)and Tremor group(Tremor dominant,TD).All PD group and the control group line head MRI scan and conventional QSM imaging,measurement on amplitude figure bilateral caudate head,putamen,globus pallidus,substantia nigra and red nucleus magnetic sensitive value(MSV),compare two sets of gray matter nucleus magnetic sensitive value,and to investigate the correlation between the MSV value of gray matter nuclei and the course of disease,motor impairment score and cognitive score in patients with different subtypes of sports.Results:1.MSV values of all gray matter nuclei in PD group were higher than that of NC group,and there were significant differences between PD group and NC group in bilateral caudate nucleus head and right substantia nigra MSV,with statistical significance(P < 0.05).2.Except for the left red nucleus,the MSV value of the PIGD group was higher than that of the TD group in other gray matter nuclei,and there were statistically significant differences between the PIGD group and the TD group in the head of the bilateral caudate nucleus,the right substantia nigra and the left red nucleus(P < 0.05).3.Correlation analysis was conducted between MSV of each PIGD nucleus and the course of disease,UPDRS-III,Hoehn-Yahr,Moc A and MMSE scores,and the results showed that there was no correlation between each measured nucleus and these scores.4.Correlation analysis was conducted between MSV of each nuclear cluster in TD group and course of disease,UPDRS-III,Hoehn-Yahr,Moc A and MMSE scores.The results showed that there was a correlation between sensitivity value of right substantia nigra and UPDRS-III score(P < 0.05,r=0.580).Conclusion:Quantitative susceptibility mapping can accurately determine the changes of iron content in gray matter nucleus of Parkinson's disease,provide clinical basis for the diagnosis of Parkinson's disease,and can evaluate the disease progression of patients with different motor subtypes.Part ? Diagnostic value of MRI ESWAN sequence " swallow tail sign" of substantia nigra and radiomics in Parkinson's diseaseObjective:To explore and compare diagnostic value between MRI ESWAN sequence " swallow tail sign" of substantia nigra and radiomics in Parkinson's disease.Methods:MRI images and clinical data of 73 patients with Parkinson's disease and 47 healthy controls were collected retrospectively.First,the occurrence and disappearance rates of “swallow tail sign” in the substantia nigra of PD patients and healthy control group were observed,and the sensitivity,specificity and accuracy of “swallow tail sign” in the diagnosis of PD were calculated.Secondly,according to the ratio of training set and testing set of 8:2,96 cases were selected as the training set(58 cases of Parkinson's disease and 38 cases of the control group),and 24 cases were selected as the testing set(15 cases of Parkinson's disease and 9 cases of the control group).The maximum level of melanin displayed in the ESWAN image of each case was selected as the ROI(Region Of Interest,ROI),Applied Darwin intelligence after the platform to sketch a ROI extracts features,first application of maximum minimum normalization of data preprocessing,the second application of the optimal feature selection to select the characteristic parameters of 3% the application logic regression and support vector machine(SVM)iterative filter the data further optimized,finally choose the most diagnostic value of eight characteristic parameter and build image omics SVM model based on linear kernel function,calculated the sensitivity of the SVM model,specificity,accuracy and Area Under Curve(AUC).Result:The sensitivity,specificity and accuracy of observing the disappearance of " swallow tail sign" of substantia nigra in the diagnosis of Parkinson's disease were 0.753,0.423 and 0.683,respectively;based on the linear kernel function,the SVM model,the accuracy and the sensitivity of the training set,specificity,AUC value is 0.737,0.897,0.823,0.891,respectively;the sensitivity of the testing set,specificity,accuracy and AUC value is 0.778,0.933,0.833,0.888,respectively.Conclusion:Radiomics method based on ESWAN sequence is feasible for the diagnosis of Parkinson's disease,and its diagnostic efficiency is better than that of the disappearance rate of "swallow tail sign" of substantia nigra.