ABSTRACT
OBJECTIVES: To preoperatively predict the high expression of Ki67 and positive pituitary transcription factor 1 (PIT-1) simultaneously in pituitary adenoma (PA) using three different radiomics models. METHODS: A total of 247 patients with PA (training set: n = 198; test set: n = 49) were included in this retrospective study. The imaging features were extracted from preoperative contrast-enhanced T1WI (T1CE), T1-weighted imaging (T1WI), and T2-weighted imaging (T2WI). Feature selection was performed using Spearman's rank correlation coefficient and least absolute shrinkage and selection operator (LASSO). The classic machine learning (CML), deep learning (DL), and deep learning radiomics (DLR) models were constructed using logistic regression (LR), support vector machine (SVM), and multi-layer perceptron (MLP) algorithms. The area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, specificity, accuracy, negative predictive value (NPV) and positive predictive value (PPV) were calculated for the training and test sets. In addition, combined with clinical characteristics, the best CML and the best DL models (SVM classifier), the DL radiomics nomogram (DLRN) was constructed to aid clinical decision-making. RESULTS: Seven CML features, 96 DL features, and 107 DLR features were selected to construct CML, DL and DLR models. Compared to CML and DL model, the DLR model had the best performance. The AUC, sensitivity, specificity, accuracy, NPV and PPV were 0.827, 0.792, 0.800, 0.796, 0.800 and 0.792 in the test set, respectively. CONCLUSIONS: Compared with CML and DL models, the DLR model shows the best performance in predicting the Ki67 and PIT-1 expression in PAs simultaneously.
ABSTRACT
Using radiomics to predict O6-methylguanine-DNA methyltransferase promoter methylation status in patients with newly diagnosed glioblastoma and compare the performances of different MRI sequences. Preoperative MRI scans from 215 patients were included in this retrospective study. After image preprocessing and feature extraction, two kinds of machine-learning models were established and compared for their performances. One kind was established using all MRI sequences (T1-weighted image, T2-weighted image, contrast enhancement, fluid-attenuated inversion recovery, DWI_b_high, DWI_b_low and apparent diffusion coefficient), and the other kind was based on single MRI sequence as listed above. For the machine-learning model based on all sequences, a total of seven radiomic features were selected with the Maximum Relevance and Minimum Redundancy algorithm. The predictive accuracy was 0.993 and 0.750 in the training and validation sets, respectively, and the area under curves were 1.000 and 0.754 in the two sets, respectively. For the machine-learning model based on single sequence, the numbers of selected features were 8, 10, 10, 13, 9, 7 and 6 for T1-weighted image, T2-weighted image, contrast enhancement, fluid-attenuated inversion recovery, DWI_b_high, DWI_b_low and apparent diffusion coefficient, respectively, with predictive accuracies of 0.797-1.000 and 0.583-0.694 in the training and validation sets, respectively, and the area under curves of 0.874-1.000 and 0.538-0.697 in the two sets, respectively. Specifically, T1-weighted image-based model performed best, while contrast enhancement-based model performed worst in the independent validation set. The machine-learning models based on seven different single MRI sequences performed differently in predicting O6-methylguanine-DNA methyltransferase status in glioblastoma, while the machine-learning model based on the combination of all sequences performed best.
Subject(s)
Brain Neoplasms , DNA Modification Methylases , DNA Repair Enzymes , Glioblastoma , Magnetic Resonance Imaging , Tumor Suppressor Proteins , Humans , Glioblastoma/diagnostic imaging , Glioblastoma/genetics , Magnetic Resonance Imaging/methods , Female , Male , DNA Modification Methylases/genetics , DNA Modification Methylases/metabolism , Middle Aged , Brain Neoplasms/diagnostic imaging , Brain Neoplasms/genetics , DNA Repair Enzymes/genetics , DNA Repair Enzymes/metabolism , Adult , Aged , Tumor Suppressor Proteins/genetics , Tumor Suppressor Proteins/metabolism , Machine Learning , DNA Methylation , Retrospective Studies , Young Adult , RadiomicsABSTRACT
OBJECTIVES: This article aims to predict the presence of vascular endothelial growth factor (VEGF) expression and to predict the expression level of VEGF by machine learning based on preoperative magnetic resonance imaging (MRI) of glioblastoma (GBM). METHODS: We analyzed the axial T2-weighted images (T2WI) and T1-weighted contrast-enhancement images of preoperative MRI in 217 patients with pathologically diagnosed GBM. Patients were divided into negative and positive VEGF groups, with the latter group further subdivided into low and high expression. The machine learning models were established with the maximum relevance and minimum redundancy algorithm and the extreme gradient boosting classifier. The area under the receiver operating curve (AUC) and accuracy were calculated for the training and validation sets. RESULTS: Positive VEGF in GBM was 63.1% (137/217), with a high expression ratio of 53.3% (73/137). To predict the positive and negative VEGF expression, 7 radiomic features were selected, with 3 features from T1CE and 4 from T2WI. The accuracy and AUC were 0.83 and 0.81, respectively, in the training set and were 0.73 and 0.74, respectively, in the validation set. To predict high and low levels, 7 radiomic features were selected, with 2 from T1CE, 1 from T2WI, and 4 from the data combinations of T1CE and T2WI. The accuracy and AUC were 0.88 and 0.88, respectively, in the training set and were 0.72 and 0.72, respectively, in the validation set. CONCLUSION: The VEGF expression status in GBM can be predicted using a machine learning model. Radiomic features resulting from data combinations of different MRI sequences could be helpful.
Subject(s)
Brain Neoplasms , Glioblastoma , Humans , Glioblastoma/diagnostic imaging , Glioblastoma/genetics , Glioblastoma/pathology , Vascular Endothelial Growth Factor A , Brain Neoplasms/diagnostic imaging , Brain Neoplasms/genetics , Brain Neoplasms/pathology , ROC Curve , Magnetic Resonance Imaging/methodsABSTRACT
[This corrects the article DOI: 10.3389/fonc.2021.699789.].
ABSTRACT
BACKGROUND: Differentiation between cerebral glioblastoma multiforme (GBM) and solitary brain metastasis (MET) is important. The existing radiomic differentiation method ignores the clinical and routine magnetic resonance imaging (MRI) features. PURPOSE: To differentiate between GBM and MET and between METs from the lungs (MET-lung) and other sites (MET-other) through clinical and routine MRI, and radiomics analyses. METHODS AND MATERIALS: A total of 350 patients were collected from two institutions, including 182 patients with GBM and 168 patients with MET, which were all proven by pathology. The ROI of the tumor was obtained on axial postcontrast MRI which was performed before operation. Seven radiomic feature selection methods and four classification algorithms constituted 28 classifiers in two classification strategies, with the best classifier serving as the final radiomics model. The clinical and combination models were constructed using the nomograms developed. The performance of the nomograms was evaluated in terms of calibration, discrimination, and clinical usefulness. Student's t-test or the chi-square test was used to assess the differences in the clinical and radiological characteristics between the training and internal validation cohorts. Receiver operating characteristic curve analysis was performed to assess the performance of developed models with the area under the curve (AUC). RESULTS: The classifier fisher_decision tree (fisher_DT) showed the best performance (AUC: 0.696, 95% CI:0.608-0.783) for distinguishing between GBM and MET in internal validation cohorts; the classifier reliefF_random forest (reliefF_RF) showed the best performance (AUC: 0.759, 95% CI: 0.613-0.904) for distinguishing between MET-lung and MET-other in internal validation cohorts. The combination models incorporating the radiomics signature and clinical-radiological characteristics were superior to the clinical-radiological models in the two classification strategies (AUC: 0.764 for differentiation between GBM in internal validation cohorts and MET and 0.759 or differentiation between MET-lung and MET-other in internal validation cohorts). The nomograms showed satisfactory performance and calibration and were considered clinically useful, as revealed in the decision curve analysis. DATA CONCLUSION: The combination of radiomic and non-radiomic features is helpful for the differentiation among GBM, MET-lung, and MET-other.
ABSTRACT
OBJECTIVE: To identify optimal machine-learning methods for the radiomics-based differentiation of gliosarcoma (GSM) from glioblastoma (GBM). MATERIALS AND METHODS: This retrospective study analyzed cerebral magnetic resonance imaging (MRI) data of 83 patients with pathologically diagnosed GSM (58 men, 25 women; mean age, 50.5 ± 12.9 years; range, 16-77 years) and 100 patients with GBM (58 men, 42 women; mean age, 53.4 ± 14.1 years; range, 12-77 years) and divided them into a training and validation set randomly. Radiomics features were extracted from the tumor mass and peritumoral edema. Three feature selection and classification methods were evaluated in terms of their performance in distinguishing GSM and GBM: the least absolute shrinkage and selection operator (LASSO), Relief, and Random Forest (RF); and adaboost classifier (Ada), support vector machine (SVM), and RF; respectively. The area under the receiver operating characteristic curve (AUC) and accuracy (ACC) of each method were analyzed. RESULTS: Based on tumor mass features, the selection method LASSO + classifier SVM was found to feature the highest AUC (0.85) and ACC (0.77) in the validation set, followed by Relief + RF (AUC = 0.84, ACC = 0.72) and LASSO + RF (AUC = 0.82, ACC = 0.75). Based on peritumoral edema features, Relief + SVM was found to have the highest AUC (0.78) and ACC (0.73) in the validation set. Regardless of the method, tumor mass features significantly outperformed peritumoral edema features in the differentiation of GSM from GBM (P < 0.05). Furthermore, the sensitivity, specificity, and accuracy of the best radiomics model were superior to those obtained by the neuroradiologists. CONCLUSION: Our radiomics study identified the selection method LASSO combined with the classifier SVM as the optimal method for differentiating GSM from GBM based on tumor mass features.