Objective To observe the value of diffusion kurtosis imaging(DKI)radiomics for evaluating Parkinson disease(PD).Methods Totally 76 PD patients(PD group)and 80 healthy controls(HC group)were retrospectively analyzed.The subjects were divided into training set(n=125,including 61 PD and 64 HC)and test set(n=31,including 15 PD and 16 HC)at the ratio of 8:2.ROI of bilateral substantia nigra,caudate nucleus,putamen,globus pallidus and thalamus were automatically delineated on mean kurtosis(MK)images of cerebral DKI.The mean MK values(MKmean)of the above ROIs were obtained and compared between groups.Support vector machine(SVM)model was constructed based on 50 selected optimal texture features.Receiver operating characteristic(ROC)curve was drawn,and the area under the curve(AUC)was calculated to evaluate the efficacy of SVM model for evaluating PD.Results MKmean of bilateral substantia nigra,caudate nucleus and thalamus in PD group were all significantly lower than those in HC group(all P＜0.05).No significant difference of MKmean of bilateral putamen nor globus pallidus was found between groups(all P＞0.05).The sensitivity,specificity,accuracy and AUC of SVM model for evaluating PD in training set was 86.89％,93.75％,90.40％and 0.982,respectively,which in test set was 86.67％,93.75％,90.32％and 0.958,respectively.Conclusion DKI radiomics could be used to effectively evaluate PD through description of microstructural changes of cerebral nuclei.