Developing a predictive model for septic shock risk in acute pancreatitis patients using interpretable machine learning algorithms.

BackgroundThis study aimed to construct predictive models for the risk of sepsis in patients with Acute pancreatitis (AP) using machine learning methods and compared optimal one with the logistic regression (LR) model and scoring systems.MethodsIn this retrospective cohort study, data were collected from the Medical Information Mart for Intensive Care III (MIMIC III) database between 2001 and 2012 and the MIMIC IV database between 2008 and 2019. Patients were randomly divided into training and test sets (8:2). The least absolute shrinkage and selection operator (LASSO) regression plus 5-fold cross-validation were used to screen and confirm the predictive factors. Based on the selected predictive factors, 6 machine learning models were constructed, including support vector machine (SVM), K-nearest neighbour (KNN), multi-layer perceptron (MLP), LR, gradient boosting decision tree (GBDT) and adaptive enhancement algorithm (AdaBoost). The models and scoring systems were evaluated and compared using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and the area under the curve (AUC).ResultsA total of 1, 672 patients were eligible for participation. In the training set, 261 AP patients (19.51%) were diagnosed with sepsis. The predictive factors for the risk of sepsis in AP patients included age, insurance, vasopressors, mechanical ventilation, Glasgow Coma Scale (GCS), heart rate, respiratory rate, temperature, SpO2, platelet, red blood cell distribution width (RDW), International Normalized Ratio (INR), and blood urea nitrogen (BUN). The AUC of the GBDT model for sepsis prediction in the AP patients in the testing set was 0.985. The GBDT model showed better performance in sepsis prediction than the LR, systemic inflammatory response syndrome (SIRS) score, bedside index for severity in acute pancreatitis (BISAP) score, sequential organ failure assessment (SOFA) score, quick-SOFA (qSOFA), and simplified acute physiology score II (SAPS II).ConclusionThe present findings suggest that compared to the classical LR model and SOFA, qSOFA, SAPS II, SIRS, and BISAP scores, the machine learning model-GBDT model had a better performance in predicting sepsis in the AP patients, which is a useful tool for early identification of high-risk patients and timely clinical interventions.

BackgroundAcute Pancreatitis-Associated Lung Injury (APALI) is one of the most severe and life-threatening systemic complications in acute pancreatitis patients, with high rates of morbidity and mortality. This study aims to develop a prediction model for the diagnosis of APALI based on machine learning algorithms.MethodsThis study included data from the First Affiliated Hospital of Bengbu Medical College (July 2012 to June 2022), which were randomly categorized into the training and testing set. And data from the Second Affiliated Hospital of Zhejiang University (January 2018 to April 2023) served as the external validation set. LASSO regression was applied to eliminate irrelevant or highly collinear independent variables. Six machine learning models were constructed, with evaluation metrics including Area Under Curve (AUC), accuracy, sensitivity, specificity, F1 score, and recall. The impact of model features was analyzed using SHapley Additive exPlanations (SHAP).ResultsA total of 1,975 patients with acute pancreatitis were randomly assigned to a training set (1,480 patients) and a testing set (495 patients). In the training set, 480 cases (32.43%) were diagnosed with APALI. The eXtreme Gradient Boosting (XGBoost) and Random Forest (RF) models demonstrated the best predictive performance, achieving the highest AUC (0.92 and 0.914, respectively), along with higher accuracy, F1 score, and recall in the testing set. Six particularly influential factors were identified and ranked as follows: CRP, BMI, neutrophil, calcium, lactate, and neutrophil-to-albumin ratio (NAR). The global interpretability of the XGBoost and RF models, along with these six features, is shown in the SHAP summary plot. These two models were selected as the optimal models for the development of an online calculator for clinical applications and risk stratification.ConclusionWe developed and internally validated a machine learning model to predict APALI, showing strong performance in our study population. To support further research and clinical use, we created an open-access web-based risk calculator. Prospective multicenter validation is needed to confirm generalizability. If successful, the tool may support early risk identification and guide interventions to prevent APALI.

Background Predicting the prognosis of patients with acute myocardial infarction (AMI) combined with diabetes mellitus (DM) is crucial due to high in-hospital mortality rates. This study aims to develop and validate a mortality risk prediction model for these patients by interpretable machine learning (ML) methods. Methods Data were sourced from the Medical Information Mart for Intensive Care IV (MIMIC-IV, version 2.2). Predictors were selected by Least absolute shrinkage and selection operator (LASSO) regression and checked for multicollinearity with Spearman’s correlation. Patients were randomly assigned to training and validation sets in an 8:2 ratio. Seven ML algorithms were used to construct models in the training set. Model performance was evaluated in the validation set using metrics such as area under the curve (AUC) with 95% confidence interval (CI), calibration curves, precision, recall, F1 score, accuracy, negative predictive value (NPV), and positive predictive value (PPV). The significance of differences in predictive performance among models was assessed utilising the permutation test, and 10-fold cross-validation further validated the model’s performance. SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) were applied to interpret the models. Results The study included 2,828 patients with AMI combined with DM. Nineteen predictors were identified through LASSO regression and Spearman’s correlation. The Random Forest (RF) model was demonstrated the best performance, with an AUC of 0.823 (95% CI: 0.774–0.872), high precision (0.867), accuracy (0.873), and PPV (0.867). The RF model showed significant differences (p < 0.05) compared to the K-Nearest Neighbours and Decision Tree models. Calibration curves indicated that the RF model’s predicted risk aligned well with actual outcomes. 10-fold cross-validation confirmed the superior performance of RF model, with an average AUC of 0.828 (95% CI: 0.800–0.842). Significant Variables in RF model indicated that the top eight significant predictors were urine output, maximum anion gap, maximum urea nitrogen, age, minimum pH, maximum international normalised ratio (INR), mean respiratory rate, and mean systolic blood pressure. Conclusion This study demonstrates the potential of ML methods, particularly the RF model, in predicting in-hospital mortality risk for AMI patients with DM. The SHAP and LIME methods enhance the interpretability of ML models.

To construct and validate the best predictive model for 28-day death risk in patients with septic shock based on different supervised machine learning algorithms. The patients with septic shock meeting the Sepsis-3 criteria were selected from Medical Information Mart for Intensive Care-IV v2.0 (MIMIC-IV v2.0). According to the principle of random allocation, 70% of these patients were used as the training set, and 30% as the validation set. Relevant predictive variables were extracted from three aspects: demographic characteristics and basic vital signs, serum indicators within 24 hours of intensive care unit (ICU) admission and complications possibly affecting indicators, functional scoring and advanced life support. The predictive efficacy of models constructed using five mainstream machine learning algorithms including decision tree classification and regression tree (CART), random forest (RF), support vector machine (SVM), linear regression (LR), and super learner [SL; combined CART, RF and extreme gradient boosting (XGBoost)] for 28-day death in patients with septic shock was compared, and the best algorithm model was selected. The optimal predictive variables were determined by intersecting the results from LASSO regression, RF, and XGBoost algorithms, and a predictive model was constructed. The predictive efficacy of the model was validated by drawing receiver operator characteristic curve (ROC curve), the accuracy of the model was assessed using calibration curves, and the practicality of the model was verified through decision curve analysis (DCA). A total of 3 295 patients with septic shock were included, with 2 164 surviving and 1 131 dying within 28 days, resulting in a mortality of 34.32%. Of these, 2 307 were in the training set (with 792 deaths within 28 days, a mortality of 34.33%), and 988 in the validation set (with 339 deaths within 28 days, a mortality of 34.31%). Five machine learning models were established based on the training set data. After including variables at three aspects, the area under the ROC curve (AUC) of RF, SVM, and LR machine learning algorithm models for predicting 28-day death in septic shock patients in the validation set was 0.823 [95% confidence interval (95%CI) was 0.795-0.849], 0.823 (95%CI was 0.796-0.849), and 0.810 (95%CI was 0.782-0.838), respectively, which were higher than that of the CART algorithm model (AUC = 0.750, 95%CI was 0.717-0.782) and SL algorithm model (AUC = 0.756, 95%CI was 0.724-0.789). Thus above three algorithm models were determined to be the best algorithm models. After integrating variables from three aspects, 16 optimal predictive variables were identified through intersection by LASSO regression, RF, and XGBoost algorithms, including the highest pH value, the highest albumin (Alb), the highest body temperature, the lowest lactic acid (Lac), the highest Lac, the highest serum creatinine (SCr), the highest Ca2+, the lowest hemoglobin (Hb), the lowest white blood cell count (WBC), age, simplified acute physiology score III (SAPS III), the highest WBC, acute physiology score III (APS III), the lowest Na+, body mass index (BMI), and the shortest activated partial thromboplastin time (APTT) within 24 hours of ICU admission. ROC curve analysis showed that the Logistic regression model constructed with above 16 optimal predictive variables was the best predictive model, with an AUC of 0.806 (95%CI was 0.778-0.835) in the validation set. The calibration curve and DCA curve showed that this model had high accuracy and the highest net benefit could reach 0.3, which was significantly outperforming traditional models based on single functional score [APS III score, SAPS III score, and sequential organ failure assessment (SOFA) score] with AUC (95%CI) of 0.746 (0.715-0.778), 0.765 (0.734-0.796), and 0.625 (0.589-0.661), respectively. The Logistic regression model, constructed using 16 optimal predictive variables including pH value, Alb, body temperature, Lac, SCr, Ca2+, Hb, WBC, SAPS III score, APS III score, Na+, BMI, and APTT, is identified as the best predictive model for the 28-day death risk in patients with septic shock. Its performance is stable, with high discriminative ability and accuracy.

Accurate prediction of in-hospital mortality for patients with severe community-acquired pneumonia (SCAP) complicated by respiratory failure admitted to the intensive care unit (ICU) remains a critical challenge. This study aimed to develop and validate a machine learning (ML) model to predict this risk and compare its performance with conventional scoring systems. In this retrospective study, data from 164 patients with SCAP and respiratory failure admitted to the ICU between January 2017 and January 2024 were analyzed. Patients were randomly divided into a training set (n = 114) and a validation (test) set (n = 50). Forty-five clinical features collected at admission were used as candidate predictors. The Least Absolute Shrinkage and Selection Operator (LASSO) regression was employed for feature selection. Six ML models, including Gradient Boosting Decision Tree (GBDT), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Decision Tree(DT),Support Vector Machine (SVM), and Logistic Regression(LR), were constructed and evaluated.Use SHAP analysis to assess the contribution of each feature in a machine learning model. Construct a nomogram using the top six most influential features. The GBDT model demonstrated the best predictive performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.83 (95% CI: 0.757-0.927) in the internal validation set, significantly outperforming the Acute Physiology and Chronic Health Evaluation II (APACHE-II, AUC = 0.70). Calibration curves demonstrated good agreement between predicted and observed mortality risks, particularly across the mid-probability range. Decision curve analysis indicated that the model provided a higher net benefit than "treat-all" and "treat-none" strategies across a broad range of threshold probabilities. SHapley Additive exPlanations (SHAP) analysis identified lactate, D-dimer, temperature, albumin, Prothrombin Time and Fraction of Inspired Oxygen as the six most influential predictors of in-hospital mortality. Based on these key predictors, we further developed a simplified nomogram to facilitate bedside risk estimation. The GBDT ML model, developed from routinely available clinical data, provides a highly accurate and clinically interpretable tool for predicting in-hospital mortality in SCAP patients with respiratory failure. It outperforms traditional severity scores and holds promise for assisting clinicians in risk stratification and early intervention.

ObjectiveTo develop an interpretable stacking ensemble model for predicting in-hospital mortality in intensive care unit (ICU) patients with CKD and sepsis and to deploy it as a web-based tool for bedside clinical use.MethodsData were extracted from the MIMIC-IV 3.0 database and split into training and test sets at a 7:3 ratio. Feature selection was performed by combining the least absolute shrinkage and selection operator (LASSO) regression with the Boruta algorithm. Eight machine learning (ML) models were trained and optimized via ten-fold cross-validation and grid search. The two models with the highest area under the curve (AUC) in the training set were combined using a stacking ensemble strategy. SHapley Additive exPlanations (SHAP) were applied to improve interpretability. Model performance was compared with the SOFA score.ResultsA total of 5344 ICU patients with CKD and sepsis were included, with an in-hospital mortality rate of 19.1%. After feature selection, 16 variables were retained. In the training set, XGBoost and LightGBM performed best. The stacking model achieved an AUC of 0.757 on the test set, outperforming SOFA (AUC = 0.668). SHAP analysis identified age, Acute Physiology Score III, Simplified Acute Physiology Score II, and respiratory rate as the top predictors. The model was also deployed as a publicly accessible web application.ConclusionThe stacking ensemble model demonstrated good discriminatory performance and interpretability for predicting in-hospital mortality in ICU patients with CKD and sepsis. Its web-based deployment provides a convenient platform for early risk assessment, although external validation is needed to confirm its broader applicability.

Chronic pain around the temporomandibular joint (TMJ) and masticatory muscles is a primary symptom of temporomandibular disorders (TMD). However, the clinical significance of magnetic resonance imaging (MRI) features in predicting TMD-related pain remains unclear. This study aimed to develop and interpret machine learning (ML) models based on MRI characteristics for predicting pain severity in patients with TMD. The present retrospective study included 584 patients with TMD between January 2022 and December 2024, yielding a total of 755 TMJ MRI data sets. Pain severity was classified using the visual analogue scale (VAS). Demographic variables (age, sex) and MRI features-including lesion side, disc position, disc morphology, disc signal, disc perforation, bilaminar zone tear, joint space, joint effusion, condylar movement, bony changes and morphology/signal of the lateral pterygoid muscle-were collected. Eleven ML models based on demographic and MRI features were developed: logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), adaptive boosting (AdaBoost), gradient boosting classifier (GBC), bagging classifier (BC), extremely randomised trees (ETC), decision tree classifier (DTC) and multilayer perceptron (MLP). Model performance was evaluated using multiple metrics, including the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity and F1 score. Precision-recall (PR) curves and calibration curves were plotted to assess discrimination and model calibration. Decision curve analysis (DCA) was conducted to evaluate the clinical net benefit across a range of threshold probabilities. Model interpretability was enhanced using Shapley Additive Explanations (SHAP), which quantified the contribution of each feature to individual predictions. Feature selection was conducted based on mean SHAP values, and separate LightGBM models were constructed using the Top 3, 5, and 9 most important features, as well as the full-feature set, for performance comparison. The data set was randomly divided into a training set (n = 604) and a test set (n = 151). Among the 11 ML models, the LightGBM model demonstrated the best predictive performance, with an AUC of 0.899, and was therefore identified as the optimal model. SHAP analysis identified age, disc position and condylar movement as the top three contributing features. Feature selection analysis indicated that selecting the top nine SHAP-ranked variables led to the highest diagnostic performance, with an AUC of 0.829. This study developed an interpretable, high-performing MRI-based ML model incorporating SHAP analysis to integrate imaging and clinical features for objective pain assessment, which may help identify high-risk TMD patients and guide personalised treatment strategies.

ObjectiveAnalyze the psychological and clinical factors of clinically significant tinnitus (THI score ≥38) in patients with hearing loss, construct predictive models based on four machine learning (ML) algorithms, and compare the predictive performance of different models.MethodsPatients with hearing loss who visited the Department of Otolaryngology at Qinghai University between August 2024 and May 2025 were enrolled in this study. Clinical data were retrieved from the hospital’s electronic medical record system. The study outcome was the occurrence of clinically significant tinnitus. Predictive variables were screened using univariate analysis, the least absolute shrinkage and selection operator (LASSO) regression, and the Boruta algorithm. Four ML algorithms—logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), and support vector machine (SVM)—were applied to construct and validate predictive models. The area under the receiver operating characteristic curve (AUC) of each model in the validation set was compared using the DeLong test. Additionally, model performance metrics in the validation set were compared to identify the optimal model. Finally, the Shapley additive explanations (SHAP) algorithm was employed to interpret the best-performing model.ResultsNine key variables—age, hypertension, sleep disorder, anxiety, hearing loss severity, depression, noise exposure history, hearing side, and ototoxic drug use—were retained after LASSO and Boruta feature selection. Among the four ML models, the RF algorithm achieved the best predictive performance, with an AUC of 0.973 in the training set and 0.977 in the validation set, followed by XGBoost (AUC = 0.962 and 0.961, respectively). DeLong tests confirmed that RF significantly outperformed LR and SVM models (p < 0.001), while its difference from XGBoost was not significant. In the validation set, the RF model yielded the highest accuracy (0.923), sensitivity (0.929), specificity (0.914), precision (0.945), and F1-score (0.937). SHAP analysis indicated that hearing loss severity, age, and sleep disorder were the most influential predictors, suggesting that both auditory and non-auditory factors contribute substantially to the risk of clinically significant tinnitus.ConclusionThe RF model showed the best performance in predicting clinically significant tinnitus, with hearing loss severity, age, and sleep disorder identified as major predictors. Integrating auditory and psychological factors can improve early risk identification in patients with hearing loss.

This study aimed to develop and validate a machine learning-based predictive model for individualized assessment and management of chemotherapy-induced myelosuppression (CIM) in patients with colorectal cancer (CRC). A total of 450 patients with CRC undergoing chemotherapy were retrospectively enrolled in the training cohort, and an additional 150 patients were included for external validation. Feature selection was performed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression. Three machine learning algorithms [Logistic Regression (LR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)] were applied to construct predictive models. Model performance was assessed using multiple metrics, including accuracy, area under the receiver operating characteristic curve (AUC), F1 score, sensitivity, and specificity. The SHapley Additive exPlanations (SHAP) method was employed to rank and interpret the importance of predictive features. In the training cohort, 52.4% of patients developed CIM. The feature selection process identified 19 significant variables that were incorporated into the predictive models. Both LR and RF demonstrated optimal performance, with an AUC of 0.83 and an accuracy of 0.76 in the training set. In the test set, RF continued to outperform other models, achieving an AUC of 0.77 and an accuracy of 0.71. External validation confirmed the robustness of the RF model, which achieved an AUC of 0.93 (95% CI: 0.89-0.97), an accuracy of 0.89, a sensitivity of 0.86, and a specificity of 0.92. SHAP analysis revealed that the most important predictors included hematological parameters, nutritional risk score (NRS2002), and a history of radiotherapy. The RF-based machine learning model demonstrated high accuracy and strong external validation capability for predicting the risk of CIM in CRC patients.

Machine learning in epidemiology: An introduction, comparison with traditional methods, and a case study of predicting extreme longevity.

This study aimed to develop and validate a machine learning (ML)-based model for predicting 6-month all-cause mortality in patients diagnosed with infective endocarditis (IE), using retrospective data from a single tertiary care center. Additionally, key prognostic features were identified through model interpretability analysis. A cohort of 444 patients with IE was retrospectively assessed and randomly divided into a training set (n = 310) and a test set (n = 134). Following feature selection, a Cox proportional hazards model and four ML-based survival models: Random Survival Forest, Extremely Randomized Survival Trees (EST), eXtreme Gradient Boosting, and Support Vector Machine were established. Model performance was assessed using the concordance index (C-index), time-dependent area under the curve (AUC), and Kaplan-Meier survival analysis. The model demonstrating optimal performance underwent interpretability analysis using Shapley additive explanations and an exploratory online tool was developed to demonstrate its functionality. Nine predictors were selected for model construction: hemoglobin concentration, presence of severe valvular regurgitation, occurrence of septic shock, New York Heart Association (NYHA) classification, serum albumin level, intracerebral hemorrhage, neutrophil percentage, history of coronary heart disease, and receipt of surgical intervention. The EST model demonstrated the highest predictive performance in the test set (C-index: 0.852), with the smallest discrepancy in performance between the training and test datasets, indicating favorable generalizability. Time-dependent AUCs at 30, 90, and 180 days were 0.957, 0.925, and 0.855, respectively. Surgical intervention was modeled as a protective factor, associated with lower mortality risk, while intracerebral hemorrhage, higher NYHA class, severe valvular regurgitation, and septic shock were associated with increased risk. The EST model demonstrated a high predictive accuracy for 6-month mortality in patients with IE, outperforming both conventional and other ML models. Its robust performance and interpretability suggest that, following rigorous external and multicenter validation, it may serve as a hypothesis-generating toolfor risk stratification. It is not yet validated for direct clinical application and should currently be considered exploratory.

Mesenteric malperfusion (MMP) represents a severe complication of acute aortic dissection (AAD). Research on risk identification models for MMP is currently limited. Based on a retrospective study of medical records from the Beijing Anzhen Hospital spanning from January 2016 to June 2022, we included 435 patients with AAD and allocated their data to training and testing sets at a ratio of 7:3. Key preoperative predictive variables were identified through the least absolute shrinkage and selection operator (LASSO) regression. Subsequently, six machine learning algorithms were used to develop and validate an MMP risk identification model: logistic regression (LR), support vector classification (SVC), random forest (RF), extreme gradient boosting (XGBoost), naive Bayes (NB), and multilayer perceptron (MLP). To determine the optimal model, the performance of the model was evaluated using various metrics, including the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity, and the Brier score. LASSO regression identified white blood cell count (WBC), neutrophil count (NE), lactate dehydrogenase (LDH), serum lactate levels, and arterial blood pH as key predictive variables. Among these, the WBC (OR 1.169, 95% confidence interval [CI] 1.086, 1.258; p < 0.001) and LDH levels (OR 1.001, 95% CI 1.000, 1.003; p = 0.008) were identified as independent risk factors for MMP. Among the six assessed machine learning algorithms, the RF model exhibited the best predictive capabilities, yielding AUROCs of 0.888 (95% CI 0.887, 0.889) and 0.797 (95% CI 0.794, 0.800) in the training and testing datasets, respectively, as well as sensitivities of 0.864 (95% CI 0.862, 0.867) and 0.811 (95% CI 0.806, 0.816), respectively, in the corresponding datasets. This study employed machine learning algorithms to develop a model capable of identifying MMP risk based on initial preoperative laboratory test results. This model can serve as a basis for making decisions in the treatment and diagnosis of MMP.

Development and validation of an explainable machine learning model for mortality prediction among patients with infected pancreatic necrosis.

This study sought to develop and validate different machine learning (ML) models that leverage non-contrast MRI radiomics to predict the degree of nonperfusion volume ratio (NVPR) of high-intensity focused ultrasound (HIFU) treatment for uterine fibroids, equipping clinicians with an early prediction tool for decision-making. This study conducted a retrospective analysis on 221 patients with uterine fibroids who received HIFU treatment and were divided into a training set (N = 117), internal validation (N = 49), and an external test set (N = 55). The 851 radiomics features were extracted from T2-weighted imaging (T2WI), and the max-relevance and min-redundancy (mRMR) and the least absolute shrinkage and selection operator (LASSO) regression were applied for feature selection. Several ML models were constructed by logistic regression (LR), decision tree (DT), random forest (RF), support vector machine (SVM), extreme gradient boosting (XGBoost), and light gradient boosting machine (LGBM). These models underwent internal and external validation, and the best model's feature significance was assessed via the Shapley additive explanations (SHAP) method. Four significant non-contrast MRI radiomics features were identified, with the SVM model outperforming others in both internal and external validations, and the AUCs of the T2WI models were 0.860, 0.847, and 0.777, respectively. SHAP analysis highlighted five critical predictors of postoperative NVPR degree, encompassing two radiomics features from non-contrast MRI and three clinical data indicators. The SVM model combining radiomics features and clinical parameters effectively predicts NVPR degree post-HIFU, which enables timely and effective interventions of HIFU.

ObjectivesThis study aimed to develop a multimodal predictive model that integrates clinical data, radiomics, and three-dimensional deep learning to forecast acute respiratory distress syndrome in patients with acute pancreatitis.MethodsThis retrospective study analyzed data from 759 patients with acute pancreatitis treated at three hospitals. Radiomics features were extracted from three-dimensional computed tomography images, and a three-dimensional deep learning model was developed using convolutional networks. These components were combined with clinical data using the XGBoost algorithm to construct a multimodal model. The performance of the model was compared with that of single-modal models and traditional scoring systems (Modified Computed Tomography Severity Index, Ranson score, and Bedside Index for Severity in Acute Pancreatitis), using area under the curve as the primary metric. Model interpretability was enhanced using variable importance analysis, SHapley Additive exPlanations, local interpretable model-agnostic explanations, calibration plots, and decision curve analysis.ResultsThe multimodal model achieved area under the curve values of 0.872 (training set) and 0.876 (test set), outperforming traditional scores (Modified Computed Tomography Severity Index: 0.747 and 0.759; Ranson score: 0.575 and 0.568; and Bedside Index for Severity in Acute Pancreatitis: 0.748 and 0.757, respectively) and single-modal models (radiomics: 0.638 and 0.727 and deep learning: 0.756 and 0.727, respectively).ConclusionBy integrating clinical tabular data, radiomics, and deep learning features, the multimodal model can predict the risk of acute respiratory distress syndrome in patients with acute pancreatitis at an early stage.