Computed Tomographic Radiomics in Differentiating Histologic Subtypes of Epithelial Ovarian Carcinoma

Abstract

ImportanceEpithelial ovarian carcinoma is heterogeneous and classified according to the World Health Organization Tumour Classification, which is based on histologic features and molecular alterations. Preoperative prediction of the histologic subtypes could aid in clinical management and disease prognostication.ObjectiveTo assess the value of radiomics based on contrast-enhanced computed tomography (CT) in differentiating histologic subtypes of epithelial ovarian carcinoma in multicenter data sets.Design, Setting, and ParticipantsIn this diagnostic study, 665 patients with histologically confirmed epithelial ovarian carcinoma were retrospectively recruited from 4 centers (Hong Kong, Guangdong Province of China, and Seoul, South Korea) between January 1, 2012, and February 28, 2022. The patients were randomly divided into a training cohort (n = 532) and a testing cohort (n = 133) with a ratio of 8:2. This process was repeated 100 times. Tumor segmentation was manually delineated on each section of contrast-enhanced CT images to encompass the entire tumor. The Mann-Whitney U test and voted least absolute shrinkage and selection operator were performed for feature reduction and selection. Selected features were used to build the logistic regression model for differentiating high-grade serous carcinoma and non–high-grade serous carcinoma.ExposuresContrast-enhanced CT-based radiomics.Main Outcomes and MeasuresIntraobserver and interobserver reproducibility of tumor segmentation were measured by Dice similarity coefficients. The diagnostic efficiency of the model was assessed by receiver operating characteristic curve and area under the curve.ResultsIn this study, 665 female patients (mean [SD] age, 53.6 [10.9] years) with epithelial ovarian carcinoma were enrolled and analyzed. The Dice similarity coefficients of intraobserver and interobserver were all greater than 0.80. Twenty radiomic features were selected for modeling. The areas under the curve of the logistic regression model in differentiating high-grade serous carcinoma and non–high-grade serous carcinoma were 0.837 (95% CI, 0.835-0.838) for the training cohort and 0.836 (95% CI, 0.833-0.840) for the testing cohort.Conclusions and RelevanceIn this diagnostic study, radiomic features extracted from contrast-enhanced CT were useful in the classification of histologic subtypes in epithelial ovarian carcinoma. Intraobserver and interobserver reproducibility of tumor segmentation was excellent. The proposed logistic regression model offered excellent discriminative ability among histologic subtypes.

Hand-Crafted Quantitative Radiomic Analysis of Computed Tomography Scans Using Machine and Deep Learning Techniques Accurately Predicts Histological Subtypes of Non-Small Cell Lung Cancer

“…29 Radiomic analysis also proved to have a high AUC-ROC in detecting epithelial ovarian carcinoma subtypes on CT with AUCs or 0.836. 30 In contrast to most of such studies, our study has achieved AUC-ROCs of 0.95 in detecting the histological subtype of NSCLC. Moreover, these studies have extracted variable numbers of features ranging from 107 to 1160 radiomic features, using 3D-Slicer and MATLAB to do so.…”

Machine learning-based radiomic models for predicting metachronous liver metastases in colorectal cancer patients: a multimodal study

“…Notably, the inclusion of wavelet features in our modeling approach aligns with the ndings of previous studies. 13,14,15 Wavelet transform enables the extraction of multi-frequency and multi-scale image information, making it particularly useful for capturing complex clinical characteristics that cannot be easily described by simple visual features of tumor images. The highdimensional abstract nature of wavelet features may serve a crucial role in capturing clinical information that may not be readily apparent to the naked eye.…”

“…The highdimensional abstract nature of wavelet features may serve a crucial role in capturing clinical information that may not be readily apparent to the naked eye. 15 It is worth mentioning that, among the selected features, all but two of them are high-order textures that provide valuable insights into the distribution of pixel points. This characteristic indicates that high-order textures are more effective in re ecting the spatial heterogeneity changes associated with MLM.…”

“…17 Despite the limited clinical relevance of the visual interpretation of the nal transformed features, these features are may be integral to the radiomics and may reveal potential applications. 15 By using the above radiological features, combined with clinical information, we developed ve models to predict MLM in patients with colorectal cancer and compared the performance between different models.…”

Radiomics in Gynaecological Imaging: A State-of-the-Art Review

“…A large study including 1329 patients with ovarian tumours provided an AUC of 0.91 of the machine-learning-based radiomics model for the differentiation between the benign and malignant tumours on contrast-enhanced CT [150]. Furthermore, a multicentric study involving 665 patients from four centres reported an AUC of 0.836 for differentiating high-grade and non-high-grade serous carcinoma [151].…”