Diagnostic and Interventional Radiology
Artificial Intelligence and Informatics - Original Article

Multi-parametric MRI-based peritumoral radiomics on prediction of lymph-vascular space invasion in early-stage cervical cancer

1.

Department of Biomedical Engineering, China Medical University, Shenyang, China

2.

Department of Radiology, Cancer Hospital of China Medical University, Liaoning Cancer Hospital and Institute, Shenyang, China

Diagn Interv Radiol 2020; 1: -
Read: 90 Published: 08 March 2021

PURPOSE: This retrospective study aims to evaluate multi-parametric magnetic resonance imaging (MRI) on predicting lymph-vascular space invasion (LVSI) in early-stage cervical cancer using radiomics methods.

METHODS: A total of 163 patients who underwent contrast-enhanced T1 (CE-T1) and T2 weighted imaging (T2WI) MRI scans at 3.0T were enrolled between Jan. 2014 and Sep. 2019. Radiomics features were extracted and selected from the tumoral and peritumoral regions at different dilation distances outside the tumor. Mann-Whitney U-test, the least absolute shrinkage and selection operator (LASSO) logistic regression and logistic regression was applied to select the predictive features and develop the radiomics signature. Univariate analysis was performed on the clinical characteristics. The radiomics nomogram was constructed incorporating the radiomics signature and the selected important clinical predictor. Prediction performance of the radiomics signature, clinical model and nomogram were evaluated with area under the curve (AUC), specificity, sensitivity, calibration and decision curve analysis (DCA).

RESULTS: A total of 5 features were selected from the peritumoral regions with 3- and 7-mm dilation distances outside tumors in CE-T1 and T2WI MRI, respectively, showed optimal discriminative performance. The radiomics signature comprising the selected features was significantly associated with the LVSI status. The radiomics nomogram integrating the radiomics signature and degree of cellular differentiation exhibited the best predictability with AUCs of 0.771 (SPE=0.831, SEN=0.581) in the training cohort and 0.788 (SPE=0.727, SEN=0.773) in the validation cohort. DCA confirmed the clinical usefulness of our model.

CONCLUSION: Our results illustrate that the radiomics nomogram based on MRI features from peritumoral regions and degree of cellular differentiation can be used as a noninvasive tool for predicting LVSI in cervical cancer.
 

EISSN 1305-3612