Introduction and importance: Brucellosis is globally acknowledged as the predominant zoonotic infection afflicting humans. Misdiagnosis and underdiagnosis pose significant challenges in the diagnosis of brucellosis due to the potential risk of overlooking or misinterpreting symptoms. Case presentation: This article reports a case of a 10-year-old girl with brucellosis, presenting with fever, rash on both lower limbs, and pain in the right ankle joint. The patient underwent 10 days of inaccurate treatment at a local hospital without improvement, and then sought treatment at a provincial hospital. Considering that it was the peak season for Mycoplasma pneumonia infection, she received intravenous cefotaxime combined with oral azithromycin for anti-infection treatment after admission. Six days later, blood cultures showed positive for Brucella , and it was learned that the child’s family raised sheep. She was ultimately diagnosed with brucellosis. Clinical discussion: In this case, if medical staff had been fully aware of the relevant issues and promptly conducted corresponding laboratory tests, such as the tube agglutination test (SAT) or the colloidal gold immunochromatographic assay, the patient might have received effective treatment earlier. Conclusion: Brucellosis often has nonspecific symptoms that can lead to misdiagnosis or delays in diagnosis. However, early intervention can reduce complications, shorten treatment time, and improve outcomes. Therefore, developing effective diagnostic tests for brucellosis is crucial for timely detection and management. Additionally, healthcare professionals need a thorough understanding of the disease’s clinical signs and diagnostic challenges to ensure accurate and prompt diagnoses.

Objectives: Acute mesenteric ischemia (AMI) is a rare disease with a relatively high mortality rate. We aimed to investigate the factors influencing in-hospital mortality in AMI patients and develop a nomogram for early risk stratification. Methods: Clinical data of AMI patients hospitalized from January 2013 to October 2025 were retrospectively analyzed. Univariate and multivariate logistic regression identified independent pre-treatment risk factors, and a nomogram was constructed. Discrimination, calibration, and clinical utility were evaluated using receiver operating characteristic curve analysis, Hosmer-Lemeshow test, calibration curves, and decision curve analysis (DCA), with Bootstrap internal validation. Subgroup validation was performed by age, sex, and other subgroups. A further adjusted multivariate model examined associations of early clinical characteristics and treatments with mortality. Results: Among 235 patients, 44 died in hospital (18.7%). Advanced age, heart failure, higher white blood cell (WBC), and creatinine were independent predictors. The model showed good discrimination (Bootstrap area under the curve: 0.849, 95% CI: 0.780–0.909) and fit (Hosmer-Lemeshow: χ ² = 10.111, P = 0.257), with good calibration and net benefit on DCA. Performance remained stable across subgroups. Further analysis that incorporated early clinical characteristics and treatment measures found that age, heart failure, WBC, aspartate aminotransferase, anticoagulation therapy, and vasopressor use were associated with mortality. Anticoagulation therapy was negatively correlated with mortality risk, whereas vasopressor use was positively correlated with mortality risk. Conclusions: This study identified risk factors for in-hospital mortality in AMI patients and developed a predictive nomogram model. This aids in early high-risk patient identification, timely intervention, optimized treatment decisions, and improved patient outcomes.

Background: Older adults are the high-risk group for COVID-19-related death. This study aimed to develop an accurate, efficient, clinically interpretable machine learning (ML) model for predicting mortality risk in this population, using only routine hematological indicators at admission to avoid extra medical costs and radiation exposure. Methods: 2393 COVID-19 patients were enrolled in this retrospective study. Missing values were imputed via Random Forest. RandomOverSampler was utilized during model training to alleviate moderate class imbalance. Feature selection was conducted following the maximum relevance-minimum redundancy principle. Five ML algorithms—Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), XGBoost (XGB), Light Gradient Boosting Machine (LGBM) were optimized via bayesian optimization (BO). We performed 10 rounds of random stratified data splitting; models were fitted on the training set, with intermediate screening and hyperparameter optimization implemented on the validation set. The independent held-out test set was strictly reserved for final performance evaluation. Model performances were assessed using the ROC curve, accuracy, precision, recall, F1-score and brier score. Calibration curve evaluated concordance between predicted probabilities and actual outcomes, and decision curve analysis (DCA) quantified net clinical benefit in clinical practice. Shapley additive exPlanations (SHAP) values and partial dependence plots (PDPs) interpreted feature importance and their associations with mortality risk. A simplified model was further developed using the top 10 key features identified by SHAP analysis and a corresponding risk prediction system was constructed to facilitate clinical application by physicians. Results: The LGBM model achieved the best comprehensive performance: AUC of 0.973, recall of 0.924, accuracy of 0.918, F1-score of 0.918, NPV of 0.923, and Brier score of 0.064. It outperformed other algorithms in computational efficiency and cross-dataset stability. Top 10 key features identified included basophil percentage (BA%), C-reactive protein (CRP), procalcitonin (PCT), D-dimer (D-dimer), AST/ALT ratio, cardiac troponin I (cTnI), standard bicarbonate (SB), age, aspartate transaminase (AST), and oxygen saturation (SaO ₂ ) for predicting mortality risk in older adults. Non-linear associations and threshold effects were observed (e.g., risk surged when CRP > 100 mg/L or D-dimer > 5–10 μg/mL). The simplified model reduced training time by 58.31% without compromising performance, comparable accuracy and interpretability. Conclusion: This study developed a TPE-LGBM model based on routine hematological indicators to predict mortality risk in older adults with COVID-19. The model demonstrated favorable accuracy, efficiency, and interpretability, suggesting the potential value of applying explainable machine learning to address unmet medical needs.

Background: Transfer-order heterogeneity may affect live-birth prediction after freeze-all FET cycles, but existing prediction studies have rarely modeled first- and second-transfer records separately. Methods: We developed and compared logistic regression (LR), support vector machine, random forest, XGBoost, LightGBM, and CatBoost models in three related single-center analytical cohorts: an overall cohort comprising 990 eligible cycles, a first-transfer subgroup comprising 576 cycles, and a second retained transfer-record subgroup comprising 238 cycles. Feature templates denoted as T6, T8, and T10 refer to retained predictor sets containing 6, 8, and 10 variables, respectively. Model performance was evaluated using stratified 10-fold cross-validation, and binary metrics were summarized at receiver operating characteristic (ROC)-derived Youden thresholds. Results: CatBoost + T6 was retained for the overall cohort (area under the curve [AUC]=0.704) and the first-transfer subgroup (AUC = 0.812), whereas CatBoost + T10 was retained for the second retained transfer-record subgroup (AUC = 0.825). Parallel LR-based interpretive nomograms were used to support transparent presentation. Shapley additive explanations (SHAP) analysis showed recurrent contributions from female age and ovarian-reserve-related indicators, while the second retained transfer-record subgroup additionally highlighted basal progesterone and interval days. Conclusions: Transfer-order subgroup modeling improved model-population matching and provided more clinically interpretable structures for pre-transfer live-birth assessment after freeze-all FET cycles.

Ex vivo CAR-T changed cancer care, but its made-to-order manufacturing keeps too many patients waiting. In vivo CAR-T takes a different path: deliver the genetic blueprint to a patient’s own immune cells and let the body build the therapy. In this review, we compare the leading delivery platforms—retargeted lentiviral vectors, engineered AAVs, LNPs (mRNA/circRNA), polymeric nanoparticles, and bioinspired systems—through the lenses of specificity, expression durability, redosing, manufacturability, immunogenicity, safety, and clinical maturity. Early human observations in hematologic malignancies and autoimmune disease show what’s possible; solid-tumor progress is still largely preclinical. Key translational challenges include: off-target transduction and murky biodistribution, liver-skewed exposure for non-viral systems, integration risk for durable vectors, repeat-dose immunity, CMC consistency, and regulatory fit. We outline a practical roadmap—quantitative biodistribution and potency, cargo-tuned persistence, receptor-aware targeting, robust CMC, and indication-specific risk–benefit—to support reproducible and clinically interpretable therapeutic development.