Objective To develop and validate a prediction model for in-hospital recurrent intensive care unit (ICU) admission in critically ill patients with ischemic stroke (IS) based on machine learning (ML) algorithms. Methods Clinical data from 2 929 IS patients were included from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Least absolute shrinkage and selection operator (LASSO) regression was used to identify predictive factors, and the synthetic minority over-sampling technique (SMOTE) was employed to create a derivation cohort comprising 2 583 patients. These patients were randomly divided into a training set (n=2 066) and a test set (n=517) at an 8:2 ratio. Five ML algorithms, including decision tree, random forest, adaptive boosting (AdaBoost), gradient boosting decision tree (GBDT), and support vector machine (SVM), were performed to construct prediction models. Five-fold cross-validation was used to evaluate the performance of the model in the training set. The area under the receiver operating characteristic curve (ROC-AUC) and decision curve analysis (DCA) were used to assess and compare the models in the testing set. The best-performing model was interpreted by shapley additive explanations (SHAP). Results Among the 2 929 patients included, 704 (24.0%) experienced in-hospital recurrent ICU admission. Among the five ML models, the random forest model demonstrated the best predictive performance, with an AUC of 0.839 (95%CI 0.801–0.877). Feature importance analysis identified five most significant features affecting model prediction, including APS Ⅲ score, albumin, age, heart rate, and SOFA score. Conclusions ML-based models can effectively predict the risk of in-hospital recurrent ICU admission in critically ill patients with IS. The random forest model showed superior predictive performance, which may have potential applications in early clinical risk stratification and intervention.

Animal experiments are an essential component of life sciences and medical research. However, the external validity and reliability of individual animal studies are frequently challenged by inherent limitations such as small sample sizes, high design heterogeneity, and poor reproducibility, which impede the effective translation of research findings into clinical practice. Systematic reviews and meta-analysis represent a key methodology for integrating existing evidence and enhancing the robustness of conclusions. Currently, however, the application of systematic reviews and meta-analysis in the field of animal experiments lacks standardized guidelines for their conduct and reporting, resulting in inconsistent quality and, to some extent, diminishing their evidence value. To address this issue, this paper aims to systematically delineate the reporting process for systematic reviews and meta-analysis of animal experiments and to propose a set of standardized recommendations that are both scientific and practical. The article's scope encompasses the entire process, from the preliminary preparatory phase [including formulating the population， intervention， comparison and outcome （PICO） question, assessing feasibility, and protocol pre-registration] to the key writing points for each section of the main report. In the core methods section, the paper elaborates on how to implement literature searches, establish eligibility criteria, perform data extraction, and assess the risk of bias, based on the Preferred Reporting Items for Systematic Reviews and Meta-analysis （PRISMA) statement, in conjunction with relevant guidelines and tools such as Animal Research： Reporting of in Vivo Experiments （ARRIVE) and a risk of bias assessment tool developed by the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE). For the presentation of results, strategies are proposed for clear and transparent display using flow diagrams and tables of characteristics. The discussion section places particular emphasis on how to scientifically interpret pooled effects, thoroughly analyze sources of heterogeneity, evaluate the impact of publication bias, and cautiously discuss the validity and limitations of extrapolating findings from animal studies to clinical settings. Furthermore, this paper recommends adopting the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology to comprehensively grade the quality of evidence. Through a modular analysis of the entire reporting process, this paper aims to provide researchers in the field with a clear and practical guide, thereby promoting the standardized development of systematic reviews and meta-analysis of animal experiments and enhancing their application value in scientific decision-making and translational medicine.

The ripe dried fruit of citron(Citrus medica) is one of the important sources of Chinese herb Citri Fructus. At the same time, it is also grown for edible and ornamental uses. There are many species and abundant genetic variation. To clarify the intraspecific variation and resource distribution of citron, this study investigated the variation in 11 citron fruits, basically covering the main species in China, including Xiaoguo citron(C. medica var. ethrog), Goucheng(C. medica var. yunnanensis), Muli citron(C.medica var. muliensis), Dehong citron(C.medica×Citrus spp.), Fuzhou citron(C.medica×C.grandis?), Mawu(C.medica×C.grandis?), Cangyuan citron, Binchuan citron, Sweet citron, Big citron, and Small citron. The natural communities of citron were proved to be mainly distributed in the southwestern and western Yunnan and southeastern Tibet of China, with Yunnan, Sichuan, Guangxi, Chongqing, Hubei, and Zhejiang identified as the main production areas. Citron has also been widely grown in India, the Mediterranean region, and the Caribbean coast countries. The field investigation revealed the large-scale intraspecific variation of citron fruits. Most of the fruits are oval-like or sphere-like in shape. The fruits are green when raw and yellow when ripe, with oil cell dots on the skin, stripe-likes running from top to bottom, and bulge at the top. Usually, in the smaller citron fruits, the pulp and juice vesicles are better developed and the central columella is tighter. By contrast, the juice vesicles and central columella in larger fruits became more vacant, with carpels visible, and the apex segregation and development of the carpels is one of the reasons for variation. These variations should be given top priority in the future variety selection and breeding, and the quality differences of different citron species and their mechanisms should be further studied. In particular, variety selection and classification management according to their medicinal or edible purposes will provide scientific and technological supports for the orderly, safe, and effective production of citron products consumed as food and medicine.
China ; Citrus ; Fruit ; Taste ; Tibet