MR imaging of epithelial ovarian cancer: a combined model to predict histologic subtypes

Qian, Luodan; Ren, Jialiang; Liu, AiShi; Gao, Yang; Hao, Fene; Zhao, Lei; Wu, Hui; Guang-ming, Niu

doi:10.1007/s00330-020-06993-5

Cited by 44 publications

(30 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among the patients with HGSOC and non-HGSOC, GBM showed the highest accuracy (0.91), with an AUC of 0.93 and an AUCPR of 0.91. These results were better than those of the conventional logistic regression model of mean ADC value ( Figure 4 ) and similar to those of Qian et al, derived from mixed radiomics models for type I/II EOCs (AUC = 0.96, 95% CI = 0.92–1) [ 39 ]. For the five histological types of EOC, the XGBoost ML tool provided the highest accuracy (0.68), with an AUC of 0.83 and AUCPR of 0.64.…”

Section: Discussionsupporting

confidence: 87%

An Application of Machine Learning That Uses the Magnetic Resonance Imaging Metric, Mean Apparent Diffusion Coefficient, to Differentiate between the Histological Types of Ovarian Cancer

Song

Bak

Kim

et al. 2021

JCM

View full text Add to dashboard Cite

This retrospective single-center study included patients diagnosed with epithelial ovarian cancer (EOC) using preoperative pelvic magnetic resonance imaging (MRI). The apparent diffusion coefficient (ADC) of the axial MRI maps that included the largest solid portion of the ovarian mass was analysed. The mean ADC values (ADCmean) were derived from the regions of interest (ROIs) of each largest solid portion. Logistic regression and three types of machine learning (ML) applications were used to analyse the ADCs and clinical factors. Of the 200 patients, 103 had high-grade serous ovarian cancer (HGSOC), and 97 had non-HGSOC (endometrioid carcinoma, clear cell carcinoma, mucinous carcinoma, and low-grade serous ovarian cancer). The median ADCmean of patients with HGSOC was significantly lower than that of patients without HGSOCs. Low ADCmean and CA 19-9 levels were independent predictors for HGSOC over non-HGSOC. Compared to stage I disease, stage III disease was associated with HGSOC. Gradient boosting machine and extreme gradient boosting machine showed the highest accuracy in distinguishing between the histological findings of HGSOC versus non-HGSOC and between the five histological types of EOC. In conclusion, ADCmean, disease stage at diagnosis, and CA 19-9 level were significant factors for differentiating between EOC histological types.

show abstract

Section: Discussionsupporting

confidence: 87%

An Application of Machine Learning That Uses the Magnetic Resonance Imaging Metric, Mean Apparent Diffusion Coefficient, to Differentiate between the Histological Types of Ovarian Cancer

Song

Bak

Kim

et al. 2021

JCM

View full text Add to dashboard Cite

show abstract

“…Additionally, some reports have indicated that the MRI radiomics model can achieve higher accuracy in discriminating benign ovarian lesions from malignancies and between type I and type II ovarian epithelial cancer. ( Zhang et al, 2019 ; Qian et al, 2020 ). Pan et al developed a nomogram model that combined CT radiomics and semantic features, which could be used for imaging biomarkers (radiomic and semantic features) to classify serous and mucinous types of ovarian cystadenomas.…”

Section: Discussionmentioning

confidence: 99%

Diagnosis of Ovarian Neoplasms Using Nomogram in Combination With Ultrasound Image-Based Radiomics Signature and Clinical Factors

Chen

et al. 2021

Front. Genet.

View full text Add to dashboard Cite

Objectives: To establish and validate a nomogram integrating radiomics signatures from ultrasound and clinical factors to discriminate between benign, borderline, and malignant serous ovarian tumors.Materials and methods: In this study, a total of 279 pathology-confirmed serous ovarian tumors collected from 265 patients between March 2013 and December 2016 were used. The training cohort was generated by randomly selecting 70% of each of the three types (benign, borderline, and malignant) of tumors, while the remaining 30% was included in the validation cohort. From the transabdominal ultrasound scanning of ovarian tumors, the radiomics features were extracted, and a score was calculated. The ability of radiomics to differentiate between the grades of ovarian tumors was tested by comparing benign vs borderline and malignant (task 1) and borderline vs malignant (task 2). These results were compared with the diagnostic performance and subjective assessment by junior and senior sonographers. Finally, a clinical-feature alone model and a combined clinical-radiomics (CCR) model were built using predictive nomograms for the two tasks. Receiver operating characteristic (ROC) analysis, calibration curve, and decision curve analysis (DCA) were performed to evaluate the model performance.Results: The US-based radiomics models performed satisfactorily in both the tasks, showing especially higher accuracy in the second task by successfully discriminating borderline and malignant ovarian serous tumors compared to the evaluations by senior sonographers (AUC = 0.789 for seniors and 0.877 for radiomics models in task one; AUC = 0.612 for senior and 0.839 for radiomics model in task 2). We showed that the CCR model, comprising CA125 level, lesion location, ascites, and radiomics signatures, performed the best (AUC = 0.937, 95%CI 0.905–0.969 in task 1, AUC = 0.924, 95%CI 0.876–0.971 in task 2) in the training as well as in the validation cohorts (AUC = 0.914, 95%CI 0.851–0.976 in task 1, AUC = 0.890, 95%CI 0.794–0.987 in task 2). The calibration curve and DCA analysis of the CCR model more accurately predicted the classification of the tumors than the clinical features alone.Conclusion: This study integrates novel radiomics signatures from ultrasound and clinical factors to create a nomogram to provide preoperative diagnostic information for differentiating between benign, borderline, and malignant ovarian serous tumors, thereby reducing unnecessary and risky biopsies and surgeries.

show abstract

“…Before image segmentation, preprocessing was required. First, according to the research of Qian et al (17), the fs-T2WI and DWI sequence images of each patient were selected for registration to the DCE-MRI (late arterial stage only) image. Then, the planar resolution of each mode was uniformly resampled to 1x1x1mm.…”

Section: Imaging Acquisition and Preprocessingmentioning

confidence: 99%

Multiparameter MRI Radiomics Model Predicts Preoperative Peritoneal Carcinomatosis in Ovarian Cancer

Ren²,

Jia

et al. 2021

Front. Oncol.

Self Cite

View full text Add to dashboard Cite

ObjectivesTo evaluate the predictive value of radiomics features based on multiparameter magnetic resonance imaging (MP-MRI) for peritoneal carcinomatosis (PC) in patients with ovarian cancer (OC).MethodsA total of 86 patients with epithelial OC were included in this retrospective study. All patients underwent FS-T2WI, DWI, and DCE-MRI scans, followed by total hysterectomy plus omentectomy. Quantitative imaging features were extracted from preoperative FS-T2WI, DWI, and DCE-MRI images, and feature screening was performed using a minimum redundancy maximum correlation (mRMR) and least absolute shrinkage selection operator (LASSO) methods. Four radiomics models were constructed based on three MRI sequences. Then, combined with radiomics characteristics and clinicopathological risk factors, a multi-factor Logistic regression method was used to construct a radiomics nomogram, and the performance of the radiomics nomogram was evaluated by receiver operating characteristic curve (ROC) curve, calibration curve, and decision curve analysis.ResultsThe radiomics model from the MP-MRI combined sequence showed a higher area under the curve (AUC) than the model from FS-T2WI, DWI, and DCE-MRI alone (0.846 vs. 0.762, 0.830, 0.807, respectively). The radiomics nomogram (AUC=0.902) constructed by combining radiomics characteristics and clinicopathological risk factors showed a better diagnostic effect than the clinical model (AUC=0.858) and the radiomics model (AUC=0.846). The decision curve analysis shows that the radiomics nomogram has good clinical application value, and the calibration curve also proves that it has good stability.ConclusionRadiomics nomogram based on MP-MRI combined sequence showed good predictive accuracy for PC in patients with OC. This tool can be used to identify peritoneal carcinomatosis in OC patients before surgery.

show abstract

MR imaging of epithelial ovarian cancer: a combined model to predict histologic subtypes

Cited by 44 publications

References 30 publications

An Application of Machine Learning That Uses the Magnetic Resonance Imaging Metric, Mean Apparent Diffusion Coefficient, to Differentiate between the Histological Types of Ovarian Cancer

An Application of Machine Learning That Uses the Magnetic Resonance Imaging Metric, Mean Apparent Diffusion Coefficient, to Differentiate between the Histological Types of Ovarian Cancer

Diagnosis of Ovarian Neoplasms Using Nomogram in Combination With Ultrasound Image-Based Radiomics Signature and Clinical Factors

Multiparameter MRI Radiomics Model Predicts Preoperative Peritoneal Carcinomatosis in Ovarian Cancer

Contact Info

Product

Resources

About