Diagnosis of triple negative breast cancer based on radiomics signatures extracted from preoperative contrast-enhanced chest computed tomography

Abstract

Background: To explore the diagnostic value of radiomics features of preoperative computed tomography (CT) for triple negative breast cancer (TNBC) for better treatment of patients with breast cancer. Methods: A total of 890 patients with breast cancer admitted to our hospital from June 2016 to January 2018 were analyzed. They were diagnosed by surgery and pathology to have mass and invasive breast cancer and had contrast-enhanced chest CT examination before operation. 300 patients were randomly selected for the study, including 100 TNBC and 200 non-TNBC (NTNBC) patients. Among them 180 were used in discovery group and 120 were used in validation group. The molecular subtypes of breast cancer in the patients were determined immunohistochemistrially. Radiomics features were extracted from three dimensional CT-images. The LASSO logistic method was used to select image features and calculate radiomics scores. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic value of radiomics scores for TNBC. Results: Five image features were found to be related to TNBC subtype (P < 0.001). These image features basedradiomic signatures had good predictive values for TNBC with the areas under ROC curve (AUC) of 0.881 (95% CI: 0.781-0.921) in the discovery group and 0.851 (95% CI: 0.761-0.961) in the validation group, respectively. The sensitivities and specificities were 0.767, and 0.873 in the discovery group and 0.785 and 0.915 in the validation group. Conclusions: Radiomic signature based on preoperative CT is capable of distinguishing patients with TNBC and NTNBC. It adds additional value for conventional chest contrast-enhanced CT and helps plan the strategy for clinical treatment of the patients.

Potential of the Non-Contrast-Enhanced Chest CT Radiomics to Distinguish Molecular Subtypes of Breast Cancer: A Retrospective Study

“…In another radiomics study based on chest CT images, the LASSO logical method was used to construct a prediction model. The AUC values to distinguish TNBC from non-TNBC were 0.881 and 0.851 in the discovery and validation groups, respectively ( 28 ). In our study, we tried 42 different feature screening and machine learning classifiers for modeling to predict luminal breast cancer.…”

Radiomics in Triple Negative Breast Cancer: New Horizons in an Aggressive Subtype of the Disease

“…Using a radiomic signature based on 15 quantitative features extracted from DCE-MRI images of breast cancer, Ma and collaborators were able to make a differential diagnosis of TNBC/non-TNBC. In this study, ROIs were automatically segmented by a deep learning algorithm and subsequently validated by two radiologists [66,67].…”

“…Radiomics for TNBC: the table includes the imaging method chosen for radiomic analysis, the type, the number of features used in creating the model including the radiomic score and nomograms, study objectives, authors and year of publication[58,[62][63][64][65][66][67][68][69][70][71][72][73][74].…”

FDG PET/CT radiomics as a tool to differentiate between reactive axillary lymphadenopathy following COVID-19 vaccination and metastatic breast cancer axillary lymphadenopathy: a pilot study

“…This technique uses intensity distribution (texture analysis) of pixel or voxel gray levels and pixel/voxel inter-connections within a region or volume of interest (e.g., tumor, lymph node) to extract these variables. In patients with cancer, first-order histogram variables (e.g., tumor shape, heterogeneity, uniformity) and second-order texture variables (e.g., Gray Level Co-occurrence Matrix [GLCM], Gray Level Dependence Matrix [GLDM]) can be used to characterize tumors [ 10 – 12 ] and have been correlated with tumor aggressiveness [ 13 ] and prognosis [ 14 ]. Machine learning models study pre-input samples with known labels (known as training data) and identify patterns from which they learn a general rule that maps inputs to outputs [ 15 ].…”