MRI-based radiomics analysis improves preoperative diagnostic performance for the depth of stromal invasion in patients with early stage cervical cancer

Abstract

Background: The depth of cervical stromal invasion is one of the important prognostic factors affecting decision-making for early stage cervical cancer (CC). This study aimed to develop and validate a T2-weighted imaging (T2WI)-based radiomics model and explore independent risk factors (factors with statistical significance in both univariate and multivariate analyses) of middle or deep stromal invasion in early stage CC.

Methods: Between March 2017 and March 2021, a total of 234 International Federation of Gynecology and Obstetrics IB1-IIA1 CC patients were enrolled and randomly divided into a training cohort (n = 188) and a validation cohort (n = 46). The radiomics features of each patient were extracted from preoperative sagittal T2WI, and key features were selected. After independent risk factors were identified, a combined model and nomogram incorporating radiomics signature and independent risk factors were developed. Diagnostic accuracy of radiologists was also evaluated.

Results: The maximal tumor diameter (MTD) on magnetic resonance imaging was identified as an independent risk factor. In the validation cohort, the radiomics model, MTD, and combined model showed areas under the curve of 0.879, 0.844, and 0.886. The radiomics model and combined model showed the same sensitivity and specificity of 87.9% and 84.6%, which were better than radiologists (sensitivity, senior = 75.7%, junior = 63.6%; specificity, senior = 69.2%, junior = 53.8%) and MTD (sensitivity = 69.7%, specificity = 76.9%).

Conclusion: MRI-based radiomics analysis outperformed radiologists for the preoperative diagnosis of middle or deep stromal invasion in early stage CC, and the probability can be individually evaluated by a nomogram.

Keywords: Cervical cancer; Magnetic resonance imaging; Radiomics; Risk factor; Stromal invasion.

Conflict of interest statement

Co-Author Chen Xia, Ying Cao, Wen Tang, and Kuan Chen are employees of Beijing Infervision Technology Co. The other authors declare that they have no competing interests. The authors not employed by Beijing Infervision Technology Co. were in control of this study.

© 2022. The Author(s).

Figures

Fig. 1

Flowchart of the study

Fig. 2

Radiomics feature selection using the least absolute shrinkage and selection operator (LASSO) regression method. a The optimal λ was selected as the lowest binomial in the LASSO model using fivefold cross-validation. b LASSO coefficient profiles of the features show vertical lines that are drawn at the value selected using fivefold cross-validation, and the optimal λ results in 5 nonzero coefficients

Fig. 3

ROC curves of combined model, radiomics model, and tumor maximum diameter on MRI for predicting middle or deep stroma invasion in the validation cohort. The senior radiologist’s performance is indicated by the black cross and the junior radiologist’s performance is indicated by the red cross

Fig. 4

Nomogram for individual prediction of the probability of middle or deep stroma invasion in early stage CC. The nomogram was was a visual representation of the combined model in training cohort, which integrated radiomics signature and independent risk factor. The radiomics signature in the nomogram was the linear sum of the selected 5 radiomics features and their corresponding coefficients. (Rsignature: radiomics signature; MTD: maximal tumor diameter on MRI)

Fig. 5

Representative images of middle or deep cervical stroma invasion (a) and superficial cervical stroma invasion (b). The lesions in the frames on sagittal T2WI are cervical tumors. a1 a 35-year-old, 2018 FIGO IB2, SCC patient with MTD on MRI of 28.0 mm. The probability of the middle or deep stroma invasion predicted by the nomogram was 98%. a2 a 49-year-old, 2018 FIGO IB1, SCC patient with MTD on MRI of 14.1 mm. The probability of the middle or deep stroma invasion predicted by the nomogram was 77%. b1 a 34-year-old, 2018 FIGO IB2, SCC patient with MTD on MRI of 20.1 mm. The probability of the middle or deep stroma invasion predicted by the nomogram was 33%. b2 a 52-year-old, 2018 FIGO IB1, AC patient with MTD on MRI of 12.2 mm. The probability of the middle or deep stroma invasion predicted by the nomogram was 13%. (MTD: maximal tumor diameter; FIGO: Federation International of Gynecology and Obstetrics; SCC: squamous cell carcinoma; AC: adenocarcinoma)