Preoperative Prediction Of Pituitary Macroadenoma Consistency Based On Multiparametric MRI Radiomics

Objective: To investigate the clinical value of preoperative prediction of pituitary macroadenoma consistency based on multiparametric magnetic resonance imaging(mp MRI)radiomics,and construct,analyze and compare the prediction efficacy of different radiomics models on pituitary macroadenoma consistency,to provide guidance for clinical individualized treatment plans.Methods: The clinical data of 137 patients with pituitary macroadenoma who underwent preoperative mp MRI were retrospectively analyzed,and the consistency of pituitary macroadenomas was classified as soft and hard according to the surgical records of neurosurgeons.Patients were randomly divided into the training set(n=108)and internal validation set(n=29),and preoperative baseline MRI of the pituitary gland(T1WI,T2 WI,CE-T1)was collected,and the region of interest(ROI)of the tumor was manually segmented in ITK-SNAP software to generate 2D ROI and 3D ROI for each of the three sequences.The Py Radiomics software was used to extract radiomics features.The variance threshold method,single variable selection method and the least absolute shrinkage and selection operator(LASSO)algorithm were used to screen the predictive features.Single and multifactorial factors were used to analyze clinical high-risk risk factors and establish clinical models.Using the logistic regression(LR)classifier to construct radiomics signature based on 2D and 3D ROI,respectively.Combined with clinical features and radiomics features,a combined clinical-radiomics model was constructed,and the nomogram was drawn.The robustness and accuracy of the prediction model were tested by using multi-center clinical data as an external validation set.The receiver operating characteristic(ROC)curve was used to evaluate the predictive effectiveness of the models,and the area under curve(AUC),accuracy,sensitivity and specificity of each model were analyzed and compared.The calibration curve and decision curve analysis(DCA)was used to evaluate the clinical reliability of the predictive models.Results: 4224 and 5061 radiomics features were extracted based on 2D ROI and 3D ROI,and 28 and 15 predictive features were selected.Among the radiomics signature,the 3Dmulti(T1WI+T2WI+CE-T1)radiomics signature constructed based on 3D ROI has high prediction efficiency.The AUC values in the training and internal validation sets are 0.793(95%CI: 0.711-0.859)and 0.798(95%CI: 0.643-0.942),respectively.Among the combined clinical-radiomics models,the 2D & 3D ROI model have the highest prediction efficiency,with the AUC values of 0.894(95%CI: 0.832-0.942)and 0.813(95%CI:0.667-0.926)in the training and internal validation sets,respectively.Compared with the clinical model,the combined clinical-radiomics model and radiomics signature are more effective in predicting tumor consistency.In addition,the results of the external validation set show that the prediction model has high robustness,and the DCA of the calibration curve shows that the prediction model has good clinical application value.Conclusion: In this study,the mp MRI(T1WI+T2WI+CE-T1)radiomics model could effectively and accurately predict the consistency of pituitary macroadenomas before surgery,and the prediction efficiency of the radiomics model based on 2D ROI and 3D ROI is different.