An Interpretable Radiomics Model Integrating Ultrasound and Clinical Features for Multiclass Classification of Axillary Lymph Nodes.

Non-invasive Prediction of Lymph Node Metastasis in NSCLC Using Clinical, Radiomics, and Deep Learning Features From 18F-FDG PET/CT Based on Interpretable Machine Learning

BackgroundBreast cancer threatens women’s health, and predicting its prognosis facilitates early therapeutic intervention. This study aims to develop radiomics models and combined models based on digital breast tomosynthesis (DBT) for predicting breast cancer prognosis and conducting interpretability analysis.MethodsPatients pathologically diagnosed with invasive breast cancer at Fudan University Shanghai Cancer Center from January 2019 to August 2020 were retrospectively included and randomly divided into a training set and a testing set at a 7:3 ratio. An independent external validation set was constructed using invasive breast cancer patients who visited Ruijin Hospital and The Affiliated Hospital of Qingdao University from December 2021 to August 2022. Disease-free survival (DFS) served as the endpoint. Univariate and multivariate Cox regression analyses were performed to identify prognosis-associated conventional imaging features on DBT. Radiomics features were extracted from the maximum layer of lesions in the craniocaudal (CC) and mediolateral oblique (MLO) views of DBT images. Selected radiomics features were incorporated into the Cox proportional hazards model to predict prognosis and a combined model in conjunction with conventional imaging features was constructed. Stratified assessment was conducted for evaluating the model performance by comparing the C-index value, the area under the receiver operating characteristic curve (AUC), decision curve analysis (DCA), and calibration curves. Nomograms and Kaplan-Meier curves were plotted to stratify the disease risks. Additionally, SHapley Additive exPlanations (SHAP) were employed to carry out the interpretability analysis.ResultsA total of 395 patients were enrolled in the training and testing cohorts, whereas the validation cohort had 140 patients. High-density masses (P=0.01) and axillary adenopathy (P<0.001) were identified as independent factors associated with DFS. Eight radiomics features were ultimately incorporated into the model. In the validation set, the radiomics model exhibited the C-index value of 0.71, while that of the combined model was 0.76. Based on the combined model for stratified prediction, the AUC values for predicting 1-, 2-, and 5-year DFS in the testing set were 0.73, 0.74, and 0.76. In the validation set, the AUC values for predicting 1- and 2-year DFS were 0.74 and 0.76. Both DCA curves and calibration curves confirmed the clinical utility of the combined model. Kaplan-Meier curves showed that the combined model stratified patients into high-risk and low-risk groups (P values were <0.001 in the training set, 0.03 in the testing set, and 0.03 in the external validation set). SHAP analysis revealed that radiomics features derived from wavelet transformation and those from the CC view contributed more substantially and carried higher weights among the selected features.ConclusionsRadiomics based on DBT have potential to predict breast cancer prognosis in terms of short-term DFS, with the combined model exhibiting superior efficacy. SHAP analysis is conducive to mining imaging biomarkers related to prognosis.

ObjectiveThis study aims to explore the impact of different ROI delineation strategies on the axillary lymph nodes metastasis (ALNM) prediction model by analyzing two-dimensional ultrasound images of lymph nodes. In addition, we integrated clinical and pathological information to construct a comprehensive model, and based on this model, developed a nomogram for individualized assessment of the probability of ALNM.MethodsA total of 146 axillary lymph nodes were randomly divided into a training set and a testing set at a ratio of 8:2. Clinical and pathological features were selected using univariate and multivariate logistic regression analyses, followed by the construction of a clinical prediction model. Radiomic features were extracted from both the internal and surrounding regions of the two-dimensional ultrasound images of the axillary lymph nodes. The least absolute shrinkage and selection operator (LASSO) algorithm was then used to select and retain the optimal features, followed by the construction of a radiomic prediction model. A combined prediction model was developed by integrating the clinical and radiomic models, and a nomogram was created for the combined prediction model.ResultsThe clinical status of axillary lymph nodes was an independent predictor for metastasis. The clinical prediction model based on the status of axillary lymph nodes achieved an AUC of 0.728 in the testing set. The radiomic prediction model based on the LASSO logistic regression algorithm with a 1 mm extended region had the highest AUC of 0.856 in the testing set. The combined prediction model integrating the clinical and optimal radiomic models achieved an AUC of 0.841 in the testing set, with a sensitivity of 77.4% and an accuracy of 79.5%. This combined model outperformed the individual clinical and radiomic models and was more effective in predicting axillary lymph node metastasis.ConclusionThis study developed a predictive model for ALNM based on ultrasound images of ALNs and their peripheral extended regions. The results demonstrated that the combined model incorporating both the lymph node and a 1-mm peripheral extension yielded the best predictive performance. Furthermore, a comprehensive integrated model was established by incorporating clinical and pathological characteristics, which effectively enhanced the prediction of ALNM.

This study aimed to identify radiomic features of primary tumor and develop a model for indicating extrahepatic metastasis of hepatocellular carcinoma (HCC). Contrast-enhanced computed tomographic (CT) images of 177 HCC cases, including 26 metastatic (MET) and 151 non-metastatic (non-MET), were retrospectively collected and analyzed. For each case, 851 radiomic features, which quantify shape, intensity, texture, and heterogeneity within the segmented volume of the largest HCC tumor in arterial phase, were extracted using Pyradiomics. The dataset was randomly split into training and test sets. Synthetic Minority Oversampling Technique (SMOTE) was performed to augment the training set to 145 MET and 145 non-MET cases. The test set consists of six MET and six non-MET cases. The external validation set is comprised of 20 MET and 25 non-MET cases collected from an independent clinical unit. Logistic regression and support vector machine (SVM) models were identified based on the features selected using the stepwise forward method while the deep convolution neural network, visual geometry group 16 (VGG16), was trained using CT images directly. Grey-level size zone matrix (GLSZM) features constitute four of eight selected predictors of metastasis due to their perceptiveness to the tumor heterogeneity. The radiomic logistic regression model yielded an area under receiver operating characteristic curve (AUROC) of 0.944 on the test set and an AUROC of 0.744 on the external validation set. Logistic regression revealed no significant difference with SVM in the performance and outperformed VGG16 significantly. As extrahepatic metastasis workups, such as chest CT and bone scintigraphy, are standard but exhaustive, radiomic model facilitates a cost-effective method for stratifying HCC patients into eligibility groups of these workups.

At present, it is still challenging to predict the clinical outcome of acute ischemic stroke (AIS). In this retrospective study, we explored whether radiomics features extracted from fluid-attenuated inversion recovery (FLAIR) and apparent diffusion coefficient (ADC) images can predict clinical outcome of patients with AIS. Patients with AIS were divided into a training (n = 110) and an external validation (n = 80) sets. A total of 753 radiomics features were extracted from each FLAIR and ADC image of the 190 patients. Interquartile range (IQR), Wilcoxon rank sum test, and least absolute shrinkage and selection operator (LASSO) were used to reduce the feature dimension. The six strongest radiomics features were related to an unfavorable outcome of AIS. A logistic regression analysis was employed for selection of potential predominating clinical and conventional magnetic resonance imaging (MRI) factors. Subsequently, we developed several models based on clinical and conventional MRI factors and radiomics features to predict the outcome of AIS patients. For predicting unfavorable outcome [modified Rankin scale (mRS) > 2] in the training set, the area under the receiver operating characteristic curve (AUC) of ADC radiomics model was 0.772, FLAIR radiomics model 0.731, ADC and FLAIR radiomics model 0.815, clinical model 0.791, and clinical and conventional MRI model 0.782. In the external validation set, the AUCs for the prediction with ADC radiomics model was 0.792, FLAIR radiomics model 0.707, ADC and FLAIR radiomics model 0.825, clinical model 0.763, and clinical and conventional MRI model 0.751. When adding radiomics features to the combined model, the AUCs for predicting unfavorable outcome in the training and external validation sets were 0.926 and 0.864, respectively. Our results indicate that the radiomics features extracted from FLAIR and ADC can be instrumental biomarkers to predict unfavorable clinical outcome of AIS and would additionally improve predictive performance when adding to combined model.

BackgroundIt is difficult for radiologists to differentiate adrenal lipid-poor adenomas from non-adenomas; nevertheless, this differentiation is important as the clinical interventions required are different for adrenal lipid-poor adenomas and non-adenomas.PurposeTo develop an unenhanced computed tomography (CT)-based radiomics model for identifying adrenal lipid-poor adenomas to assist in clinical decision-making.Materials and methodsPatients with adrenal lesions who underwent CT between January 2015 and August 2021 were retrospectively recruited from two independent institutions. Patients from institution 1 were randomly divided into training and test sets, while those from institution 2 were used as the external validation set. The unenhanced attenuation and tumor diameter were measured to build a conventional model. Radiomics features were extracted from unenhanced CT images, and selected features were used to build a radiomics model. A nomogram model combining the conventional and radiomic features was also constructed. All the models were developed in the training set and validated in the test and external validation sets. The diagnostic performance of the models for identifying adrenal lipid-poor adenomas was compared.ResultsA total of 292 patients with 141 adrenal lipid-poor adenomas and 151 non-adenomas were analyzed. Patients with adrenal lipid-poor adenomas tend to have lower unenhanced attenuation and smoother image textures. In the training set, the areas under the curve of the conventional, radiomic, and nomogram models were 0.94, 0.93, and 0.96, respectively. There was no difference in diagnostic performance between the conventional and nomogram models in all datasets (all p < 0.05).ConclusionsOur unenhanced CT-based nomogram model could effectively distinguish adrenal lipid-poor adenomas. The diagnostic power of conventional unenhanced CT imaging features may be underestimated, and further exploration is worthy.

Background The tumor burden within the axillary lymph nodes (ALNs) constitutes a pivotal factor in breast cancer, serving as the primary determinant for treatment decisions and exhibiting a close correlation with prognosis. Objective This study aimed to investigate the potential of ultrasound-based radiomics and clinical characteristics in non-invasively distinguishing between low tumor burden (1-2 positive nodes) and high tumor burden (more than 2 positive nodes) in patients with node-positive breast cancer. Methods A total of 215 patients with node-positive breast cancer, who underwent preoperative ultrasound examinations, were enrolled in this study. Among these patients, 144 cases were allocated to the training set, 37 cases to the validation set, and 34 cases to the testing set. Postoperative histopathology was used to determine the status of ALN tumor burden. The region of interest for breast cancer was delineated on the ultrasound image. Nine models were developed to predict high ALN tumor burden, employing a combination of three feature screening methods and three machine learning classifiers. Ultimately, the optimal model was selected and tested on both the validation and testing sets. In addition, clinical characteristics were screened to develop a clinical model. Furthermore, Shapley additive explanations (SHAP) values were utilized to provide explanations for the machine learning model. Results During the validation and testing sets, the models demonstrated area under the curve (AUC) values ranging from 0.577 to 0.733 and 0.583 to 0.719, and accuracies ranging from 64.9% to 75.7% and 64.7% to 70.6%, respectively. Ultimately, the Boruta_XGB model, comprising five radiomics features, was selected as the final model. The AUC values of this model for distinguishing low from high tumor burden were 0.828, 0.715, and 0.719 in the training, validation, and testing sets, respectively, demonstrating its superiority over the clinical model. Conclusions The developed radiomics models exhibited a significant level of predictive performance. The Boruta_XGB radiomics model outperformed other radiomics models in this study.

Purpose: Predicting colorectal cancer liver metastasis (CRLM) is essential for prognostic assessment. This study aims to develop and validate an interpretable multimodal machine learning framework based on multiparametric MRI for predicting CRLM, and to enhance the clinical interpretability of the model through SHapley Additive exPlanations (SHAP) analysis and deep learning visualization. Methods: This multicenter retrospective study included 463 patients with pathologically confirmed colorectal cancer from two institutions, divided into training (n = 256), internal testing (n = 111), and external validation (n = 96) sets. Radiomics features were extracted from manually segmented regions on axial T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI). Deep learning features were obtained from a pretrained ResNet101 network using the same MRI inputs. A least absolute shrinkage and selection operator (LASSO) logistic regression classifier was developed for clinical, radiomics, deep learning, and combined models. Model performance was evaluated by AUC, sensitivity, specificity, and F1-score. SHAP was used to assess feature contributions, and Grad-CAM was applied to visualize deep feature attention. Results: The combined model integrating features across the three modalities achieved the highest performance across all datasets, with AUCs of 0.889 (training), 0.838 (internal test), and 0.822 (external validation), outperforming single-modality models. Decision curve analysis (DCA) revealed enhanced clinical net benefit from the integrated model, while calibration curves confirmed its good predictive consistency. SHAP analysis revealed that radiomic features related to T2WI texture (e.g., LargeDependenceLowGrayLevelEmphasis) and clinical biomarkers (e.g., CA19-9) were among the most predictive for CRLM. Grad-CAM visualizations confirmed that the deep learning model focused on tumor regions consistent with radiological interpretation. Conclusions: This study presents a robust and interpretable multiparametric MRI-based model for noninvasively predicting liver metastasis in colorectal cancer patients. By integrating handcrafted radiomics and deep learning features, and enhancing transparency through SHAP and Grad-CAM, the model provides both high predictive performance and clinically meaningful explanations. These findings highlight its potential value as a decision-support tool for individualized risk assessment and treatment planning in the management of colorectal cancer.

To develop a radiomics model based on ultrasound images for predicting risk of recurrence in breast cancer patients. In this retrospective study, 420 patients with pathologically confirmed breast cancer were included, randomly divided into training (70%) and test (30%) sets, with an independent external validation cohort of 90 patients. According to St. Gallen recurrence risk criteria, patients were categorized into two groups, low-medium-risk and high-risk. Radiomics features were extracted from a radiomics analysis set using Pyradiomics. The informative radiomics features were screened using the minimum redundancy maximum relevance (mRMR) and the least absolute shrinkage and selection operator (LASSO) algorithms. Subsequently, radiomics models were constructed with eight machine learning algorithms. Three distinct nomogram models were created using the features selected through multivariate logistic regression, including the Clinic-Ultrasound (Clin-US), Clinic-Radiomics (Clin-Rad), and Clinic-Ultrasound-Radiomics (Clin-US-Rad) models. The receiver operating characteristic (ROC), calibration, and decision curve analysis (DCA) curves were used to evaluate the model's clinical applicability and predictive performance. A total of 12 ultrasound radiomics features were screened, of which wavelet.LHL first order Mean features weighed more and tended to have a high risk of recurrence. The higher the risk of recurrence, the higher the radiomics score (Rad-score) in all three sets (training, test, and external validation set, all p < 0.05). Rad-score is equally applicable in four different subtypes of breast cancer. In the test set and external validation set, the Clin-US-Rad model achieved the highest AUC values (AUC = 0.817 and 0.851, respectively). The calibration and DCA curves also demonstrated the good clinical utility of the combined model. The machine learning-based ultrasound radiomics model were useful for predicting the risk of recurrence in breast cancer. The nomograms show promising potential in assessing the recurrence risk of breast cancer. This non-invasive approach offers crucial guidance for the diagnosis and treatment of the condition.

Accurate identification of bone marrow invasion (BMI) is critical for determining the prognosis of and treatment strategies for lymphoma. Although bone marrow biopsy (BMB) is the current gold standard, its invasive nature and sampling errors highlight the necessity for noninvasive alternatives. We aimed to develop and validate an interpretable machine learning model that integrates clinical data, 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) parameters, radiomic features, and deep learning features to predict BMI in lymphoma patients. We included 159 newly diagnosed lymphoma patients (118 from Center I and 41 from Center II), excluding those with prior treatments, incomplete data, or under 18 years of age. Data from Center I were randomly allocated to training (n = 94) and internal test (n = 24) sets; Center II served as an external validation set (n = 41). Clinical parameters, PET/CT features, radiomic characteristics, and deep learning features were comprehensively analyzed and integrated into machine learning models. Model interpretability was elucidated via Shapley Additive exPlanations (SHAPs). Additionally, a comparative diagnostic study evaluated reader performance with and without model assistance. BMI was confirmed in 70 (44%) patients. The key clinical predictors included B symptoms and platelet count. Among the tested models, the ExtraTrees classifier achieved the best performance. For external validation, the combined model (clinical + PET/CT + radiomics + deep learning) achieved an area under the receiver operating characteristic curve (AUC) of 0.886, outperforming models that use only clinical (AUC 0.798), radiomic (AUC 0.708), or deep learning features (AUC 0.662). SHAP analysis revealed that PET radiomic features (especially PET_lbp_3D_m1_glcm_DependenceEntropy), platelet count, and B symptoms were significant predictors of BMI. Model assistance significantly enhanced junior reader performance (AUC improved from 0.663 to 0.818, p = 0.03) and improved senior reader accuracy, although not significantly (AUC 0.768 to 0.867, p = 0.10). Our interpretable machine learning model, which integrates clinical, imaging, radiomic, and deep learning features, demonstrated robust BMI prediction performance and notably enhanced physician diagnostic accuracy. These findings underscore the clinical potential of interpretable AI to complement medical expertise and potentially reduce the reliance on invasive BMB for lymphoma staging.

This study aimed to develop an artificial intelligence-assisted model for the preoperative prediction of lateral cervical lymph node metastasis (LCLNM) in papillary thyroid carcinoma (PTC) using computed tomography (CT) radiomics, providing a new noninvasive and accurate diagnostic tool for PTC patients with LCLNM. This retrospective study included 389 confirmed PTC patients, randomly divided into a training set ( n = 272) and an internal validation set ( n = 117), with an additional 40 patients from another hospital as an external validation set. Patient demographics were evaluated to establish a clinical model. Radiomic features were extracted from preoperative contrast-enhanced CT images (venous phase) for each patient. Feature selection was performed using analysis of variance and the least absolute shrinkage and selection operator algorithm. We employed support vector machine, random forest (RF), logistic regression, and XGBoost algorithms to build CT radiomic models for predicting LCLNM. A radiomics score (Rad-score) was calculated using a radiomic signature-based formula. A combined clinical-radiomic model was then developed. The performance of the combined model was evaluated using the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). A total of 1724 radiomic features were extracted from each patient's CT images, with 13 features selected based on nonzero coefficients related to LCLNM. Four clinically relevant factors (age, tumor location, thyroid capsule invasion, and central cervical lymph node metastasis) were significantly associated with LCLNM. Among the algorithms tested, the RF algorithm outperformed the others with five-fold cross-validation on the training set. After integrating the best algorithm with clinical factors, the areas under the ROC curves for the training, internal validation, and external validation sets were 0.910 (95% confidence interval [CI]: 0.729-0.851), 0.876 (95% CI: 0.747-0.911), and 0.821 (95% CI: 0.555-0.802), respectively, with DCA demonstrating the clinical utility of the combined radiomic model. This study successfully established a clinical-CT radiomic combined model for predicting LCLNM, which may significantly enhance surgical decision-making for lateral cervical lymph node dissection in patients with PTC.

Thymoma, a tumor originating from thymic epithelial cells, can have its prognosis significantly improved through early risk assessment. We proposed a fuzzy comprehensive evaluation fusion model (FCE-FM) to assess tumor risk. In this retrospective study, we enrolled 286 thymoma patients from two centers between 2018 and 2024 and partitioned the study cohorts into a training set (n = 196), an internal test set (n = 50), and an external test set (n = 40). We developed a fuzzy comprehensive evaluation-based fusion model to predict tumor risk using demographics, radiomic and multi-planar deep features. The FCE-FM integrates five base classification models(Logistic Regression, Support Vector Machine, XGBoost, LightGBM, and GBDT) via fuzzy comprehensive evaluation(FCE), analytic hierarchy process (AHP), and triangular membership function techniques. Feature selection was performed sequentially using Spearman rank correlation followed by LASSO regression. A total of 26 deep learning features (5 transverse, 13 sagittal, and 8 coronal planar features) and 4 radiomic features, along with gender, were identified to construct the models. Model performance was evaluated using the area under the curve (AUC) and accuracy metrics.The SHapley Additive exPlanations (SHAP) methodology was utilized to rank feature importance. The FCE-FM model exhibited superior predictive performance, achieving AUC values of 0.982 (95% CI 0.964-0.996), 0.927 (95% CI 0.847-0.990), and 0.895 (95% CI 0.771-0.992) on the training, internal test, and external test sets, respectively. Corresponding accuracies were 0.949, 0.860, and 0.800 across these datasets. Notably, the model consistently outperformed five baseline classifiers in terms of AUC performance on both internal and external validation sets. The FCE-FM model exhibited high stability and accuracy in multi-center validation, demonstrating its robustness. This interpretable framework offers clinicians a reliable early warning tool for tumor risk assessment, enabling timely intervention to significantly improve patient prognosis.

ObjectivesThis research was designed to compare the ability of different machine learning (ML) models and nomogram to predict distant metastasis in male breast cancer (MBC) patients and to interpret the optimal ML model by SHapley Additive exPlanations (SHAP) framework.MethodsFour powerful ML models were developed using data from male breast cancer (MBC) patients in the SEER database between 2010 and 2015 and MBC patients from our hospital between 2010 and 2020. The area under curve (AUC) and Brier score were used to assess the capacity of different models. The Delong test was applied to compare the performance of the models. Univariable and multivariable analysis were conducted using logistic regression.ResultsOf 2351 patients were analyzed; 168 (7.1%) had distant metastasis (M1); 117 (5.0%) had bone metastasis, and 71 (3.0%) had lung metastasis. The median age at diagnosis is 68.0 years old. Most patients did not receive radiotherapy (1723, 73.3%) or chemotherapy (1447, 61.5%). The XGB model was the best ML model for predicting M1 in MBC patients. It showed the largest AUC value in the tenfold cross validation (AUC:0.884; SD:0.02), training (AUC:0.907; 95% CI: 0.899—0.917), testing (AUC:0.827; 95% CI: 0.802—0.857) and external validation (AUC:0.754; 95% CI: 0.739—0.771) sets. It also showed powerful ability in the prediction of bone metastasis (AUC: 0.880, 95% CI: 0.856—0.903 in the training set; AUC: 0.823, 95% CI:0.790—0.848 in the test set; AUC: 0.747, 95% CI: 0.727—0.764 in the external validation set) and lung metastasis (AUC: 0.906, 95% CI: 0.877—0.928 in training set; AUC: 0.859, 95% CI: 0.816—0.891 in the test set; AUC: 0.756, 95% CI: 0.732—0.777 in the external validation set). The AUC value of the XGB model was larger than that of nomogram in the training (0.907 vs 0.802) and external validation (0.754 vs 0.706) sets.ConclusionsThe XGB model is a better predictor of distant metastasis among MBC patients than other ML models and nomogram; furthermore, the XGB model is a powerful model for predicting bone and lung metastasis. Combining with SHAP values, it could help doctors intuitively understand the impact of each variable on outcome.

BackgroundTo develop a predictive nomogram for breast cancer lympho-vascular invasion (LVI), based on digital breast tomography (DBT) data obtained from intra- and peri-tumoral regions.MethodsOne hundred ninety-two breast cancer patients were enrolled in this retrospective study from 2 institutions, in which Institution 1 served as the basis for training (n = 113) and testing (n = 49) sets, while Institution 2 served as the external validation set (n = 30). Tumor regions of interest (ROI) were manually-delineated on DBT images, in which peri-tumoral ROI was defined as 1 mm around intra-tumoral ROI. Radiomics features were extracted, and logistic regression was used to construct intra-, peri-, and intra- + peri-tumoral radiomics models. Patient clinical data was analyzed by both uni- and multi-variable logistic regression analyses to identify independent risk factors for the non-radiomics clinical imaging model, and the combination of both the most optimal radiomics and clinical imaging models comprised the comprehensive model. The best-performing model out of the 3 types (radiomics, clinical imaging, comprehensive) was identified using receiver operating characteristic (ROC) curve analysis, and used to construct the predictive nomogram.ResultsThe most optimal radiomics model was the intra- + peri-tumoral model, and 3 independent risk factors for LVI, maximum tumor diameter (odds ratio [OR] = 1.486, 95% confidence interval [CI] = 1.082–2.041, P = 0.014), suspicious malignant calcification (OR = 2.898, 95% CI = 1.232 ~ 6.815, P = 0.015), and axillary lymph node (ALN) metastasis (OR = 3.615, 95% CI = 1.642–7.962, P < 0.001) were identified by the clinical imaging model. Furthermore, the comprehensive model was the most accurate in predicting LVI occurrence, with areas under the curve (AUCs) of 0.889, 0.916, and 0.862, for, respectively, the training, testing and external validation sets, compared to radiomics (0.858, 0.849, 0.844) and clinical imaging (0.743, 0.759, 0.732). The resulting nomogram, incorporating radiomics from the intra- + peri-tumoral model, as well as maximum tumor diameter, suspicious malignant calcification, and ALN metastasis, had great correspondence with actual LVI diagnoses under the calibration curve, and was of high clinical utility under decision curve analysis.ConclusionsThe predictive nomogram, derived from both radiomics and clinical imaging features, was highly accurate in identifying future LVI occurrence in breast cancer, demonstrating its potential as an assistive tool for clinicians to devise individualized treatment regimes.

To develop a machine learning-based predictive model for local recurrence after radiotherapy in patients with gliomas, with interpretability enhanced through SHapley Additive exPlanations (SHAP). We retrospectively enrolled 145 patients with pathologically confirmed gliomas who underwent brain radiotherapy (training: validation = 102:43). Physiological and structural magnetic resonance imaging (MRI) were used to define habitat regions. A total of 2153 radiomic features were extracted from each MRI sequence in each habitat region, respectively. Relief and Recursive Feature Elimination were used for radiomic feature selection. Support vector machine (SVM) and random forest models incorporating clinical and radiomic features were constructed for each habitat region. The SHAP method was used to explain the predictive model. In the training cohort and validation cohort, the Physiological_Habitat1 (e-THRIVE)_radiomic SVM model demonstrated the best AUC of 0.703 (95% CI 0.569-0.836) and 0.670 (95% CI 0.623-0.717) compared to the other radiomic models. The SHAP summary plot and SHAP force plot were used to interpret the best-performing Physiological_Habitat1 (e-THRIVE)_radiomic SVM model. Radiomic features derived from the Physiological_Habitat1 (e-THRIVE) were predictive of local recurrence in glioma patients following radiotherapy. The SHAP method provided insights into how the tumor microenvironment might influence the effectiveness of radiotherapy in postoperative gliomas.