Predictive value of enhanced MRI radiomics combined with clinical features for the occurrence of post-acute pancreatitis diabetes mellitus
10.3969/j.issn.1002-1671.2024.11.013
- VernacularTitle:增强MRI影像组学及临床特征预测急性胰腺炎后糖尿病发生的价值
- Author:
Yuan WANG
1
;
Xiaohua HUANG
;
Qinglin DU
;
Xiyao WAN
;
Ziyan LIU
;
Ziyi LIU
Author Information
1. 川北医学院附属医院放射科,四川 南充 637000
- Keywords:
pancreatic diabetes mellitus;
acute pancreatitis;
radiomics;
magnetic resonance imaging;
clinical features
- From:
Journal of Practical Radiology
2024;40(11):1810-1813
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the clinical value of models based on clinical features and enhanced MRI radiomics for predicting the occurrence of post-acute pancreatitis diabetes mellitus(PPDM-A).Methods A retrospective selection of 161 acute pancreatitis(AP)patients was conducted,comprising 99 in the non-PPDM-A group and 62 in the PPDM-A group.They were randomly divided into training set and test set in a ratio of 7∶3.Region of interest(ROI)were delineated and radiomics features were extracted on the late arterial phase MRI images.Optimal radiomics features were selected by maximum relevance and minimum redundancy(mRMR)and least absolute shrinkage and selection operator(LASSO).Support vector machine(SVM)was used to develop three predictive models.The efficacy of the models in predicting PPDM-A was evaluated,the receiver operating characteristic(ROC)curve was drawn,and the DeLong test was employed to assess the difference in predictive capability among the models.Results In the training set,the area under the curve(AUC)of the clinical model,radiomics model,and combined model were 0.702,0.810 and 0.901,respectively,and in the test set were 0.678,0.797 and 0.830,respectively.The DeLong test revealed a statistically significant difference in the predictive capability of the combined model compared to the clinical model both in the training and test sets(training set:P＜0.001;test set:P=0.019).Conclusion The combined model based on clinical features and enhanced MRI radiomics features demonstrates good predictive effi-cacy and can provide valuable insights for clinical interventions aimed at preventing PPDM-A.
