Prediction of thyroid function based on CT radiomics: a two-center study

ObjectivesTo evaluate the value of quantitative parameters and radiomic features based on dual-layer spectral detector CT (DLCT) in predicting spread through air spaces (STAS) of lung adenocarcinoma (LUAD).MethodsThis study analyzed 266 patients with pathologically confirmed LUAD from two medical centers. Patients from center 1 were divided into training (n = 131) and internal validation (n = 57) sets, while center 2 (n = 78) formed the external validation set. Clinical data, conventional imaging features, and DLCT quantitative parameters were analyzed to develop a clinical-radiological model. Radiomic features were extracted from venous-phase images, including conventional images, virtual monoenergetic images (VMI) at 40 keV, 65 keV, and 100 keV, along with iodine density maps, effective atomic number (Zeff) maps, and electron density (ED) maps. The best-performing radiomics model was combined with clinical-radiological predictors to create a nomogram. Model performance was evaluated through ROC analysis, calibration curves, and decision curve analysis.ResultsMultivariate analysis revealed that tumor-lung interface and ED values were independent predictive factors in the clinical-radiological model. The optimal radiomics model was constructed based on VMI 40 keV, demonstrating AUCs of 0.899, 0.835, and 0.828 in the training, internal validation, and external validation sets, respectively. The nomogram, which incorporated the VMI 40 keV radiomics signature along with tumor-lung interface and ED values, outperformed the clinical-radiological model in the training set (AUC = 0.910 vs. 0.870; P = 0.018) and the internal validation set (AUC = 0.868 vs. 0.798; P = 0.046). While the improvement in the external validation set was not statistically significant (AUC = 0.848 vs. 0.819; P = 0.184).ConclusionThe nomogram, which integrates conventional imaging features, DLCT quantitative parameters and VMI 40 keV radiomic features, demonstrates promising performance and represents a potential valuable non-invasive tool for the preoperative assessment of STAS in LUAD.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12885-025-15436-7.

To construct a nomogram based on sonogram features and radiomics features to differentiate granulomatous lobular mastitis (GLM) from invasive breast cancer (IBC). A retrospective collection of 213 GLMs and 472 IBCs from three centers was divided into a training set, an internal validation set, and an external validation set. A radiomics model was built based on radiomics features, and the RAD score of the lesion was calculated. The sonogram radiomics model was constructed using ultrasound features and RAD scores. Finally, the diagnostic efficacy of the three sonographers with different levels of experience before and after combining the RAD score was assessed in the external validation set. The RAD score, lesion diameter, orientation, echogenicity, and tubular extension showed significant differences in GLM and IBC (p < 0.05). The sonogram radiomics model based on these factors achieved optimal performance, and its area under the curve (AUC) was 0.907, 0.872, and 0.888 in the training, internal, and external validation sets, respectively. The AUCs before and after combining the RAD scores were 0.714, 0.750, and 0.830 and 0.834, 0.853, and 0.878, respectively, for sonographers with different levels of experience. The diagnostic efficacy was comparable for all sonographers when combined with the RAD score (p > 0.05). Radiomics features effectively enhance the ability of sonographers to discriminate between GLM and IBC and reduce interobserver variation. The nomogram combining ultrasound features and radiomics features show promising diagnostic efficacy and can be used to identify GLM and IBC in a noninvasive approach.

Aim: To develop and validate a radiomics-based combined model (ModelRC) to predict the pathological grade of endometrial cancer. Methods: A total of 403 endometrial cancer patients from two independent centers were enrolled as training, internal validation and external validation sets. Radiomic features were extracted from T2-weighted images, apparent diffusion coefficient map and contrast-enhanced 3D volumetric interpolated breath-hold examination images. Results: Compared with the clinical model and radiomics model, ModelRC showed superior performance; the areas under the receiver operating characteristic curves were 0.920 (95% CI: 0.864-0.962), 0.882 (95% CI: 0.779-0.955) and 0.881 (95% CI: 0.815-0.939) for the training, internal validation and external validation sets, respectively. Conclusion: ModelRC, which incorporated clinical and radiomic features, exhibited excellent performance in the prediction of high-grade endometrial cancer.

Accurate early diagnosis of ovarian cancer is crucial. The objective of this research is to create a comprehensive model that merges clinical variables, O-RADS, and deep learning radiomics to support preoperative diagnosis and assess its efficacy for sonographers. Data from two centers were used: Center 1 for training and internal validation, and Center 2 for external validation. DL and radiomics features were extracted from transvaginal ultrasound images to create a DL radiomics model using the LASSO method. A machine learning model ensemble was created by merging clinical variables, O-RADS scores, and DL radiomics model predictions. The model's effectiveness was evaluated by measuring the area under the receiver operating characteristic curve (AUC) and analyzing its impact on improving the diagnostic skills of sonographers. Moreover, the model's additional usefulness was assessed through integrated discrimination improvement (IDI), net reclassification improvement (NRI), and subgroup analysis. The ensemble model demonstrated superior diagnostic performance for ovarian cancer compared to standalone clinical models and clinical O-RADS models. Notably, there were significant improvements in the NRI and IDI across all three datasets, with p-values < 0.05. The ensemble model exhibited exceptional diagnostic performance, achieving AUCs of 0.97 in both the internal and external validation sets. Moreover, the implementation of this ensemble model substantially improved the diagnostic precision and reliability of sonographers. The sonographers' average AUC improved by 11 % in the internal validation set and by 7.7 % in the external validation set. The ensemble model significantly enhances preoperative ovarian cancer diagnosis accuracy and improves sonographers' diagnostic capabilities and consistency.

PurposeTo assess the added value of radiomics models from preoperative chest CT in predicting the presence of spread through air spaces (STAS) in the early stage of surgically resected lung adenocarcinomas using multiple validation datasets.Materials and MethodsThis retrospective study included 550 early-stage surgically resected lung adenocarcinomas in 521 patients, classified into training, test, internal validation, and temporal validation sets (n=211, 90, 91, and 158, respectively). Radiomics features were extracted from the segmented tumors on preoperative chest CT, and a radiomics score (Rad-score) was calculated to predict the presence of STAS. Diagnostic performance of the conventional model and the combined model, based on a combination of conventional and radiomics features, for the diagnosis of the presence of STAS were compared using the area under the curve (AUC) of the receiver operating characteristic curve.ResultsRad-score was significantly higher in the STAS-positive group compared to the STAS-negative group in the training, test, internal, and temporal validation sets. The performance of the combined model was significantly higher than that of the conventional model in the training set {AUC: 0.784 [95% confidence interval (CI): 0.722–0.846] vs. AUC: 0.815 (95% CI: 0.759–0.872), p=0.042}. In the temporal validation set, the combined model showed a significantly higher AUC than that of the conventional model (p=0.001). The combined model showed a higher AUC than the conventional model in the test and internal validation sets, albeit with no statistical significance.ConclusionA quantitative CT radiomics model can assist in the non-invasive prediction of the presence of STAS in the early stage of lung adenocarcinomas.

Objective: To investigate the value of CT-based tumor and peritumoral radiomics in predicting the World Health Organization/International Society of Urological Pathology (WHO/ISUP) grading for non-metastatic clear cell renal cell carcinoma (ccRCC). Methods: A total of 503 patients with non-metastatic ccRCC were retrospectively enrolled from 7 tertiary hospitals between February 2017 and December 2023. Patients from 4 hospitals within Shandong Province were divided into a training set and an internal validation set in a 7∶3 ratio, while patients from 3 hospitals outside Shandong Province constituted the external validation set. Regions of interest (ROI) were manually delineated slice-by-slice along the tumor margin on contrast-enhanced CT images. Peritumoral regions were obtained by expanding 10 mm outward from the tumor boundary. Key radiomics features were selected using Least Absolute Shrinkage and Selection Operator (LASSO) regression with five-fold cross-validation to build tumor models and peritumoral models for predicting the WHO/ISUP grading. Univariate logistic regression was used to screen clinical factors associated with the WHO/ISUP grading, which was included to construct a combined predictive model together with tumor and peritumoral features. The predictive performance of the models for WHO/ISUP grading was evaluated using receiver operating characteristic (ROC) curves, and the Youden index and optimal cutoff value were calculated for risk stratification. Calibration curves were used to validate model performance, and decision curve analysis (DCA) was employed to evaluate the clinical value of the models. Results: The median age [M(Q1,Q3)] of the 503 patients was 59.0 (52.0, 66.0) years, with 335 males and 168 females. Univariate logistic regression analysis showed there was a statistically significance in age between low and high WHO/ISUP grades (P=0.043). The AUC of the tumor model and peritumoral model in the internal validation set were 0.744 (95%CI: 0.700-0.788) and 0.742 (95%CI: 0.709-0.774), respectively. In the external validation set, the AUC of the tumor model and peritumoral model were 0.685 (95%CI: 0.651-0.720) and 0.655 (95%CI: 0.628-0.683), respectively. The combined model demonstrated the best predictive performance in both internal and external validation sets, with AUC of 0.760 (95%CI: 0.721-0.800) and 0.706 (95%CI: 0.675-0.737), respectively. Using the Youden index calculated from the ROC curve from the combined model, the optimal cutoff value was 0.504 for risk stratification, where 190 cases were classified as low-grade and 313 as high-grade, achieving a concordance rate of 0.718 (361/503) with the WHO/ISUP grading. Calibration curve tests indicated good fit for the combined model (internal validation set: P=0.932; external validation set: P=0.404). DCA showed that the combined model provided favorable clinical net benefit within the threshold probability range of 0.2-0.8. Conclusion: The combined model incorporating age, CT-based tumor features, and peritumoral features demonstrates good performance in predicting the WHO/ISUP grading for patients with non-metastatic ccRCC.

Rationale and ObjectivesThe human epidermal growth factor receptor 2 (HER2) gene status is crucial for determining treatment efficacy. This study assessed preoperative HER2 classification in breast cancer using machine learning based on clinicopathological and MRI characteristics.Materials and MethodsThis retrospective study involved 1,015 patients (1,030 lesions) across two centers. Patients were divided into training, internal validation, and external validation sets. Nomograms were developed using clinicopathological and MRI features. Predictive models were constructed using decision trees (DT), support vector machines (SVM), k-nearest neighbors (k-NN), artificial neural networks (ANN), and multivariable logistic regression (LR). Model performance was evaluated using receiver operating characteristic curves, decision curve analysis, and calibration curves. Model interpretability was achieved by developing nomograms and employing SHAP (SHapley Additive exPlanations) analysis.ResultsKey variables for distinguishing HER2-positive from HER2-negative cases included regional N category, estrogen receptor, PR (progesterone receptor) status, Ki-67 status, lesion number, distribution quadrant, and accompanying signs. The SVM model achieved the highest AUC of 0.86 (95% confidence interval (CI): 0.81–0.90) in the training set, while the ANN model had an AUC of 0.77 (95% CI: 0.67–0.86) in the internal validation set. In the external validation set, the LR model achieved the highest AUC of 0.66 (95% CI: 0.56–0.76), although the overall performance was modest. For HER2-low versus HER2-zero differentiation, Ki-67 status, lesion number, distribution quadrant, mass shape, early enhancement rate, and ADC (apparent diffusion coefficient) were significant. The SVM model attained the highest AUC of 0.87 (95% CI: 0.83–0.91) in the training set, while the LR model demonstrated superior generalizability, yielding the highest AUCs in both the internal and external validation sets (internal: 0.67, 95% CI: 0.58–0.76; external: 0.74, 95% CI: 0.65–0.83). Radiologists benefited from the nomogram for improved diagnostic accuracy, especially junior radiologists. SHAP analysis revealed that PR status was paramount for HER2-positive classification, whereas mass shape and ADC values were dominant for identifying HER2-low status.ConclusionIntegrating machine learning with clinicopathological and MRI characteristics improves the accuracy of HER2 status classification in breast cancer and enhances diagnostic capabilities for radiologists in clinical practice.

BackgroundThere remains a demand for a practical method of identifying lipid-poor adrenal lesions.PurposeTo explore the predictive value of computed tomography (CT) features combined with demographic characteristics for lipid-poor adrenal adenomas and nonadenomas.Materials and MethodsWe retrospectively recruited patients with lipid-poor adrenal lesions between January 2015 and August 2021 from two independent institutions as follows: Institution 1 for the training set and the internal validation set and Institution 2 for the external validation set. Two radiologists reviewed CT images for the three sets. We performed a least absolute shrinkage and selection operator (LASSO) algorithm to select variables; subsequently, multivariate analysis was used to develop a generalized linear model. The probability threshold of the model was set to 0.5 in the external validation set. We calculated the sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) for the model and radiologists. The model was validated and tested in the internal validation and external validation sets; moreover, the accuracy between the model and both radiologists were compared using the McNemar test in the external validation set.ResultsIn total, 253 patients (median age, 55 years [interquartile range, 47–64 years]; 135 men) with 121 lipid-poor adrenal adenomas and 132 nonadenomas were included in Institution 1, whereas another 55 patients were included in Institution 2. The multivariable analysis showed that age, male, lesion size, necrosis, unenhanced attenuation, and portal venous phase attenuation were independently associated with adrenal adenomas. The clinical-image model showed AUCs of 0.96 (95% confidence interval [CI]: 0.91, 0.98), 0.93 (95% CI: 0.84, 0.97), and 0.86 (95% CI: 0.74, 0.94) in the training set, internal validation set, and external validation set, respectively. In the external validation set, the model showed a significantly and non-significantly higher accuracy than reader 1 (84% vs. 65%, P = 0.031) and reader 2 (84% vs. 69%, P = 0.057), respectively.ConclusionsOur clinical-image model displayed good utility in differentiating lipid-poor adrenal adenomas. Further, it showed better diagnostic ability than experienced radiologists in the external validation set.

Background To investigate the clinical value of preoperative prediction of pituitary neuroendocrine tumors (PitNETs) consistency based on multiparametric magnetic resonance imaging (mpMRI) radiomics and to use a multicenter validation set to evaluate the effectiveness of the prediction model. Methods The clinical data of 137 patients with PitNETs who underwent preoperative mpMRI were retrospectively analyzed, and tumor consistency was classified as soft or hard according to the surgical records. The patients were randomly divided into a training set (n = 108) and an internal validation set (n = 29). Preoperative baseline T1- and T2-weighted (T1/T2WI) and contrast-enhanced (CE) T1 images of the pituitary gland were collected, and each tumor was manually segmented to generate two-dimensional (2D) and three-dimensional (3D) regions of interest (ROIs). Radiomics features were extracted, and predictive features were screened using the variance threshold, single variable selection, and least absolute shrinkage and selection operator methods. Single and multifactorial factors were used to analyze the high-risk clinical risk factors and establish clinical models. A logistic regression classifier was used to construct a radiomics signature based on the 2D and 3D ROIs. A combined model of the clinical characteristics and radiomics features was constructed, and a nomogram was drawn. The robustness and accuracy of the prediction model were tested using multicenter clinical data as an external validation set. A receiver operating characteristic (ROC) curve was used to evaluate the predictive effectiveness of the models, and the area under the curve (AUC), accuracy, sensitivity, and specificity of each model were analyzed and compared. Calibration curves and decision curve analysis (DCA) were used to evaluate the clinical reliability of the predictive models. Results In total, 4224 and 5061 radiomics features were extracted and 28 and 15 predictive features were selected based on the 2D and 3D ROIs, respectively. The 3D-multi (T1WI + T2WI + CE-T1) radiomics signature had the highest prediction efficiency. AUCs of the training and the internal validation sets were 0.793 (95% confidence interval(CI): 0.711–0.859) and 0.798 (95% CI: 0.643–0.942), respectively. The 2D and 3D ROI combined clinical-radiomics models had the highest prediction efficiency, with AUCs of 0.894 (95% CI: 0.832–0.942) and 0.813 (95% CI: 0.667–0.926) in the training and internal validation sets, respectively. Compared with the clinical model, the combined clinical-radiomics model and radiomics signature were more effective in predicting tumor consistency. In addition, the results of the external validation set showed that the prediction model was highly robust, and the DCA of the calibration curve showed that the prediction model had good clinical application value. Conclusions The mpMRI (T1WI + T2WI + CE-T1) radiomics model effectively and accurately predicted PitNET consistency before surgery, and the prediction efficiencies of the radiomics models based on 2D and 3D ROIs were different.

A CT radiomics-based machine learning model for predicting high-grade pathological components in stage IA lung adenocarcinoma: a two-center study.

To explore and compare the potential value of radiomics models based on contrast-enhanced computed tomography (CT) for noninvasive preoperative prediction of lymphovascular invasion (LVI) in laryngeal squamous cell carcinoma (LSCC). This multicenter diagnostic study retrospectively enrolled patients with LSCC from three tertiary hospitals who underwent surgical treatment. Standardized preprocessing was performed on the CT images, followed by region-of-interest segmentation and extraction of traditional radiomics features and deep learning (DL) features. Features were selected using least absolute shrinkage and selection operator (LASSO) regression. Traditional radiomics models and deep learning radiomics (DLR) models were established using logistic regression, random forest, and multilayer perceptron algorithms, respectively. A transformer-based hybrid model was developed by integrating radiomics and DL features. The predictive performance of the three types of models was evaluated and compared using the area under the curve (AUC), decision curve analysis (DCA), sample probability distribution histograms, confusion matrices, calibration curves, net reclassification index (NRI), and integrated discrimination improvement (IDI). A total of 1024 patients were allocated to the training set (center1, n =291), internal validation set ( n =126), and external test sets (Center 2, n =437; Center 3, n =170). Three radiomics models and three DLR models were constructed, and the optimal performance was observed in the DLR_ Random Forest model (AUC: 0.812-0.867). The transformer hybrid model demonstrated superior predictive performance, with AUC values of 0.881, 0.843, 0.833, and 0.836 in the training, internal validation, and external test sets, respectively. DCA indicated a higher net benefit for the Transformer model, along with an improved NRI and IDI. Radiomics models based on CT images exhibit potential for noninvasive prediction of LVI in LSCC, with the transformer hybrid model achieving the highest diagnostic performance. This approach may provide clinicians with a preoperative decision support tool to optimize treatment strategies for patients with LSCC.

To develop and validate a radiomic prediction model using initial noncontrast computed tomography (CT) at admission to predict in-hospital mortality in patients with traumatic brain injury (TBI). A total of 379 TBI patients from three cohorts were categorized into training, internal validation, and external validation sets. After filtering the unstable features with the minimum redundancy maximum relevance approach, the CT-based radiomics signature was selected by using the least absolute shrinkage and selection operator (LASSO) approach. A personalized predictive nomogram incorporating the radiomic signature and clinical features was developed using a multivariate logistic model to predict in-hospital mortality in patients with TBI. The calibration, discrimination, and clinical usefulness of the radiomics signature and nomogram were evaluated. The radiomic signature consisting of 12 features had areas under the curve (AUCs) of 0.734, 0.716, and 0.706 in the prediction of in-hospital mortality in the internal and two external validation cohorts. The personalized predictive nomogram integrating the radiomic and clinical features demonstrated significant calibration and discrimination with AUCs of 0.843, 0.811, and 0.834 in the internal and two external validation cohorts. Based on decision curve analysis (DCA), both the radiomic features and nomogram were found to be clinically significant and useful. This predictive nomogram incorporating the CT-based radiomic signature and clinical features had maximum accuracy and played an optimized role in the early prediction of in-hospital mortality. The results of this study provide vital insights for the early warning of death in TBI patients.

Ki-67 labeling index (Ki-67 LI) is a proliferation marker that is correlated with aggressive behavior and prognosis of pituitary adenomas (PAs). Dynamic contrast-enhanced MRI (DCE-MRI) may potentially contribute to the preoperative assessment of Ki-67 LI. To investigate the feasibility of assessing Ki-67 LI of PAs preoperatively using delta-radiomics based on DCE-MRI. Retrospective. 605 PA patients (female = 47.1%, average age = 52.2) from two centers (high Ki-67 LI (≥ 3%) = 229; low Ki-67 LI (< 3%) = 376), divided into a training set (n = 313), an internal validation set (n = 196), and an external validation set (n = 96). 1.5-T and 3-T, DCE-MRI. This study developed a non-delta-radiomics model based on the non-delta-radiomic features directly extracted from four phases, a delta-radiomics model based on the delta-radiomic features, and a combined model integrating clinical parameters (Knosp grade and tumor diameter) with delta-radiomic features. U test, recursive feature elimination (RFE), and least absolute shrinkage and selection operator (LASSO) regression were utilized to select important radiomic features. Support vector machine (SVM), XGBoost (XGB), logistic regression (LR), and Gaussian naive Bayes (GNB) were utilized to develop the models. Receiver operating characteristic (ROC) curve. Calibration curve. Decision curve analysis (DCA). Intraclass correlation coefficients (ICC). DeLong test for ROC curves. U test or t test for numerical variables. Fisher's test or Chi-squared test for categorical variables. A p-value < 0.05 was considered statistically significant. The combined model demonstrated the best performance in preoperatively assessing the Ki-67 LI of PAs, achieving AUCs of 0.937 and 0.897 in the internal and external validation sets, respectively. The models based on delta-radiomic features outperformed the non-delta-radiomic model. A delta-radiomics-based model using DCE-MRI may show high diagnostic performance for preoperatively assessing the Ki-67 LI status of PAs. 3 Technical Efficacy: Stage 2.

To construct a novel immune-inflammatory-nutritional (IIN) score based on peripheral blood biomarkers related to inflammation, immunity, and nutrition, and to predict the efficacy of neoadjuvant chemotherapy (NAC) in patients with triple-negative breast cancer (TNBC). We retrospectively selected 431 patients with TNBC from Xijing Hospital, and then randomly divided the patients into a training set and an internal validation set in a ratio of 7:3. An external validation set was included with 154 patients selected from West China Hospital of Sichuan University. In the training set, patients were divided into the pathological complete response (pCR) group and the non-pathological complete response group. Univariate logistic regression analysis and LASSO regression analysis were used to select biomarkers that affect the efficacy of NAC in TNBC patients and to construct the IIN score. A nomogram model was constructed based on the IIN score and clinical pathological characteristics to predict whether TNBC patients could achieve pCR after NAC before treatment. The predictive performance and clinical application value of the nomogram model were assessed using the receiver operating characteristic (ROC) curve, calibration curve, decision curve analysis, and confusion matrix. Through LASSO regression analysis, 6 biomarkers were ultimately identified to construct the scoring system. A nomogram model was constructed based on the IIN score and clinical pathological characteristics, and the ROC curve showed the areas under the curve to be 0.827, 0.786, and 0.754 in the training set, internal validation set, and external validation set, respectively. Calibration curves, decision curves, and confusion matrices all demonstrated that the nomogram model exhibited robust predictive performance and holds certain clinical application value. The nomogram model based on the IIN score offers high predictive performance and can accurately predict the efficacy of NAC in TNBC patients before treatment, highlighting its clinical application potential.

PurposeThe aim of this study was to evaluate the value of different multiparametric MRI-based radiomics models in differentiating stage IA endometrial cancer (EC) from benign endometrial lesions.MethodsThe data of patients with endometrial lesions from two centers were collected. The radiomics features were extracted from T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC) map, and late contrast-enhanced T1-weighted imaging (LCE-T1WI). After data dimension reduction and feature selection, nine machine learning algorithms were conducted to determine which was the optimal radiomics model for differential diagnosis. The univariate analyses and logistic regression (LR) were performed to reduce valueless clinical parameters and to develop the clinical model. A nomogram using the radscores combined with clinical parameters was developed. Two integrated models were obtained respectively by the ensemble strategy and stacking algorithm based on the clinical model and optimal radiomics model. The area under the curve (AUC), clinical decisive curve (CDC), net reclassification index (NRI), and integrated discrimination index (IDI) were used to evaluate the performance and clinical benefits of the models.ResultsA total of 371 patients were incorporated. The LR model was the optimal radiomics model with the highest average AUC (0.854) and accuracy (0.802) in the internal and external validation groups (AUC = 0.910 and 0.798, respectively), and outperformed the clinical model (AUC = 0.739 and 0.592, respectively) or the radiologist (AUC = 0.768 and 0.628, respectively). The nomogram (AUC = 0.917 and 0.802, respectively) achieved better discrimination performance than the optimal radiomics model in two validation groups. The stacking model (AUC = 0.915) and ensemble model (AUC = 0.918) had a similar performance compared with the nomogram in the internal validation group, whereas the AUCs of the stacking model (AUC = 0.792) and ensemble model (AUC = 0.794) were lower than those of the nomogram and radiomics model in the external validation group. According to the CDC, NRI, and IDI, the optimal radiomics model, nomogram, stacking model, and ensemble model achieved good net benefits.ConclusionsMultiparametric MRI-based radiomics models can non-invasively differentiate stage IA EC from benign endometrial lesions, and LR is the best machine learning algorithm. The nomogram presents excellent and stable diagnostic efficiency.