Background Ozone (O₃) is a major air pollutant. The existing monitoring system has uneven distribution of sites, insufficient coverage in underdeveloped areas, and low temporal resolution, making it difficult to obtain hourly data. This limits the dynamic identification of pollution and the formulation of prevention and control strategies. Objective To construct an hourly O₃ concentration estimation model based on ensemble machine learning, aiming to improve the accuracy of pollution exposure assessment and explore O₃ health impacts. Methods This study integrated land use regression modeling with modern machine learning techniques, employing random forest and XGBoost algorithms to construct base models, and stacking integration using non-negative least squares. The ensemble model was trained and validated across China using high-resolution, multi-source geographic data (e.g., meteorologicaldata, population density, land cover types, and aerosol optical thickness). It was tested in Taiyuan City, combined with a distributed lag non-linear model to analyze the association between O₃ and emergency admissions. Results The constructed ensemble model performed well in predicting O₃ concentration, with a higher coefficient of determination (R²) and a lower root-mean-square deviation (RMSE) compared to the single models. The R² improved from 0.90 to 0.92, and the RMSE decreased from 11.41 to 10.62, enhancing both prediction accuracy and generalization ability. In the application to Taiyuan City, the model successfully imputed the hourly-level data for the entire year. The distributed lag non-linear model analysis revealed that the relative risk (RR) values for the 6th to 8th days following O₃ exposure were 1.14 (95%CI: 1.01, 1.29), 1.16 (95%CI: 1.02, 1.31), and 1.14 (95%CI: 1.01, 1.29), respectively, which were significantly higher than 1, indicating a significant lagged association (lagged 6-8 d) between O₃ and the number of emergency room visits. Conclusion A high-precision, hourly-level O₃ concentration estimation model is successfully constructed by combining the land use regression model with an ensemble machine learning approach to provide a scientific basis for environmental policy formulation and public health intervention. The application of the model verifies its generalization ability and practical application value, which can provide a new technical framework for subsequent environmental health research.

Under the current upsurge of cross-disciplinary research between artificial intelligence （AI） and traditional Chinese medicine （TCM）， the inheritance of academic thought in TCM is confronted with the deep-seated predicament of cognitive paradigm dislocation. The existing AI model is based on the classical reasoning mode of formal logic and statistical probability， which is difficult to adapt to the nonlinear and holistic cognitive logic of TCM with image-based thinking and syndrome-differentiated treatment as the core， resulting in systematic misreading and loss of the core essence in digital inheritance. Based on this， this paper proposes to construct a new cognitive model oriented towards the context of TCM. Its essence is not to have AI imitate the thinking of TCM， but to establish an auxiliary understanding framework based on semantic mapping and relational reasoning. It is advocated that the role of AI should be reconstructed from a knowledge substitute to a clinical interactor， and its application boundary in the inheritance of TCM wisdom should be clarified. By establishing the theoretical framework and practical path of human-computer collaboration， this paper aims to provide a scientific basis for realizing the deep integration of AI technology and the academic thought in TCM.

Under the current upsurge of cross-disciplinary research between artificial intelligence （AI） and traditional Chinese medicine （TCM）， the inheritance of academic thought in TCM is confronted with the deep-seated predicament of cognitive paradigm dislocation. The existing AI model is based on the classical reasoning mode of formal logic and statistical probability， which is difficult to adapt to the nonlinear and holistic cognitive logic of TCM with image-based thinking and syndrome-differentiated treatment as the core， resulting in systematic misreading and loss of the core essence in digital inheritance. Based on this， this paper proposes to construct a new cognitive model oriented towards the context of TCM. Its essence is not to have AI imitate the thinking of TCM， but to establish an auxiliary understanding framework based on semantic mapping and relational reasoning. It is advocated that the role of AI should be reconstructed from a knowledge substitute to a clinical interactor， and its application boundary in the inheritance of TCM wisdom should be clarified. By establishing the theoretical framework and practical path of human-computer collaboration， this paper aims to provide a scientific basis for realizing the deep integration of AI technology and the academic thought in TCM.

Objective:To explore the role of process management for continuous periheral insulin infusion (CPII) for controlling hyperglycemia during enteral nutrition (EN) for critically ill patients.Methods:A total of 75 patients who received continuous EN treatment in the emergency intensive care unit (EICU) of Qinghai Red Cross Hospital from January 2023 to January 2024 were selected in this historical controlled trial study. Patients who were admitted before the implementation of process management for CPII were included in the historical control group ( n=35), and those who were admitted after the implementation were included in the observation group ( n=40). Both groups were treated with continuous EN infusion combined with micropump-based insulin therapy, with the target blood glucose being<10 mmol/L. The blood glucose values at 4 hours and 8 hours after treatment, the time to reach the target blood glucose and the dosage of insulin, the total amount of insulin at 24 hours, the amount of calories administered when reaching the target blood glucose, the frequency of blood glucose measurement, the incidence of hypoglycemia and hypokalemia, the amount of potassium supplemented for hypokalemia, the length of EICU stay, and the incidence of nosocomial infection were compared between these two groups. Results:The blood glucose levels of the observation group at 4 hours and 8 hours after CPII were significantly lower than those of the control group (both P<0.001), and the time for the observation group to reach the target blood glucose level was significantly shorter than that for the control group ( P<0.001). The total amount of insulin in the observation group when reaching the target blood glucose and the total amount of insulin used within 24 hours were significantly smaller than those in the control group (both P<0.05). The amount of calories administered to the observation group when reaching the target blood glucose level was significantly higher than that of the control group ( P=0.002). The number of blood glucose measurements within 24 hours after insulin initiation in the observation group was significantly larger than that in the control group ( P=0.042), but there was no statistically significant difference in the total number of monitoring during EICU stay ( P=0.561). The incidence rates of hypoglycemia and hypokalemia and the amount of potassium supplemented for hypokalemia in the observation group were significantly lower than those in the control group (all P<0.05). There was no statistically significant difference in the length of EICU stay and the incidence of nosocomial infection between the two groups (both P>0.05). Conclusions:Process management for CPII in critically ill patients promotes rapid glycemic control during enteral nutrition (EN), reduces hypoglycemia, hypokalemia, and nosocomial infections, and improves overall blood glucose stability. It is vital for controlling stress hyperglycemia during EN in critical illness, with excellent safety.

Objective:To explore the clinical characteristics of patients with metastatic gastric tumor and to differentiate them from those with primary malignant gastric tumors.Methods:This study retrospectively analyzed the clinical data of 21 patients with metastatic gastric tumor confirmed by gastroscopy biopsy and pathology.The clinical manifestations,gastroscopy characteristics,pathological results,and the time interval between the diagnosis of primary malignant tumor and the confirmation of gastric metastasis were summarized.Results:The patient primarily exhibited symptoms related to gastric involvement;gastroscopy findings resembled those of primary malignant tumors of the stomach;pathological results were consistent with primary malignant tumors;the time interval between the diagnosis of primary cancer and the confirmation of metastatic gastric tumor varied significantly,ranging from up to 18 years to simultaneous detection.Conclusion:Metastatic gastric tumors are similar to primary malignant tumors of the stomach in terms of clinical manifestations and endoscopic features,making clinical differentiation challenging.The main approach for distinguishing them is based on pathological examination.

Objective To investigate the effect of resveratrol on the tumor microenvironment in osteosarcoma. Methods A C57BL/6 xenograft mouse model was established and treated with resveratrol. Single-cell sequencing was performed to analyze changes in the tumor microenvironment. Immunohistochemistry was used to assess immune cell infiltration, while Western blotting was conducted to examine alterations in cellular signaling pathways. Results Resveratrol significantly inhibited the proliferation of LM8 osteosarcoma cells in C57BL/6 mice compared to the control group. Additionally, CD8⁺ T cell recruitment was enhanced. The Janus kinase 2/signal transducer and activator of transcription 3 (JAK2/STAT3) signaling pathway was notably downregulated in LM8 osteosarcoma cells following resveratrol treatment. Conclusion Resveratrol promotes CD8⁺ T cell infiltration by inhibiting the JAK2/STAT3 signaling pathway, suggesting its potential as a therapeutic agent in osteosarcoma treatment.
Osteosarcoma/genetics* ; STAT3 Transcription Factor/genetics* ; Resveratrol/pharmacology* ; Animals ; Janus Kinase 2/genetics* ; Signal Transduction/drug effects* ; Tumor Microenvironment/immunology* ; Cell Line, Tumor ; Mice, Inbred C57BL ; Mice ; Humans ; Cell Proliferation/drug effects* ; Bone Neoplasms/metabolism* ; CD8-Positive T-Lymphocytes/drug effects* ; Xenograft Model Antitumor Assays

Objective:To develop an artificial intelligence（AI）-based sequential ultrasound-magnetic resonance imaging（US-MRI）diagnostic strategy to optimize the imaging workflow for ovarian masses.Methods:A total of 1 120 patients with pathologically confirmed ovarian masses who underwent both preoperative pelvic ultrasound and MRI between January 2021 and December 2023 at Women's Hospital，Zhejiang University School of Medicine were retrospectively included. Patients were randomly divided into the training（ n=672）and internal test set（ n=448）at a ratio of 6∶4. An external test set（ n=128）was established at the Forth Affiliated Hospital of School of Medicine. Deep learning was used for automated segmentation of MRI lesions，followed by radiomic feature extraction and machine learning classification to construct both a US-MRI multimodal model and sequential US-MRI strategy. Diagnostic performance and potential healthcare cost-saving effects were evaluated across strategies. Results:In the internal test set（ n=448），the AI-based sequential US-MRI strategy achieved a F1 score of 0.863 and a diagnostic accuracy of 82.14%，with no significant difference compared to the US-MRI multi-modal model（ P>0.05）. The sequential strategy identified 82 cases（18.30%，82/448）of patients as low-risk true negatives during initial ultrasound screening，suggesting a potential to reduce the need for MRI examinations in future clinical practice. In the external test set（ n=128），the strategy achieved an F1 score of 0.800 and a confirmed diagnosis rate of 85.94%，with a theoretical reduction of 26.56%（34 cases）in MRI utilization while maintaining a diagnostic accuracy rate higher than that of the multi-modal model（82.18%）. Conclusions:The AI-based US-MRI sequential diagnostic strategy demonstrates favorable diagnostic accuracy while offering the potential to optimize MRI utilization. This approach may enhance the efficiency of imaging resource allocation and reduce healthcare burden in the management of ovarian masses.

Imprinted genes play a key role in regulating mammalian placental and embryonic development. Here, we generated glutaminyl-peptide cyclotransferase-knockout (Qpct^-/-) mice utilizing the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) platform and identified Qpct as a novel anti-angiogenic factor in regulating mouse placentation. Compared with Qpct^+/+ mice, placentae and embryos (Qpct^-/+ and Qpct^-/-) showed significant overgrowth at embryonic Day 12.5 (E12.5), E15.5, and E18.5. Using single-cell transcriptome analysis of 32 309 cells from Qpct^+/+ and Qpct^-/- mouse placentae, we identified 13 cell clusters via single-nucleus RNA sequencing (snRNA-seq) (8880 Qpct^+/+ and 13 577 Qpct^-/- cells) and 20 cell clusters via single-cell RNA sequencing (scRNA-seq) (6567 Qpct^+/+ and 3285 Qpct^-/- cells). Furthermore, we observed a global up-regulation of pro-angiogenic genes in the Qpct^-/- background. Immunohistochemistry assays revealed a notable increase in the number of blood vessels in the decidual and labyrinthine layers of E15.5 Qpct^-/+ and Qpct^-/- mice. Moreover, the elevation of multiple pairs of ligand-receptor interactions was observed in decidual cells, endothelial cells, and macrophages, promoting angiogenesis and inflammatory response. Our findings indicate that loss of maternal Qpct leads to altered phenotypic characteristics of placentae and embryos and promotes angiogenesis in murine placentae.
Animals ; Female ; Pregnancy ; Mice ; Placentation/genetics* ; Single-Cell Analysis ; Gene Expression Profiling ; Mice, Knockout ; Transcriptome ; Placenta/blood supply* ; Neovascularization, Pathologic/genetics* ; Genomic Imprinting ; Single-Cell Gene Expression Analysis ; Angiogenesis

OBJECTIVE: To explore the implementation effect of hour-1 bundle for sepsis patients based on crisis resource management (CRM) system. METHODS: A historical control study was conducted. The hour-1 bundle for sepsis based on CRM was used to train 24 nurses in the emergency department from October 2022 to March 2023. Clinical data of sepsis patients admitted to the emergency department of the First People's Hospital of Zunyi from April 2022 to September 2023 were collected. The patients were divided into three groups based on different stages of CRM system construction: control group (before construction, from April to September in 2022), improvement group (during construction, from October 2022 to March 2023) and observation group (after construction, from April to September in 2023). The baseline data, implementation rate of hour-1 bundle [including blood culture, antibiotic usage, blood lactic acid (Lac) detection, fluid resuscitation, hypertensors usage], identification and diagnosis time, and prognosis parameters [including correction rate of hypoxemia, intensive care unit (ICU) occupancy rate, and 28-day survival rate]. Sepsis cognition survey and non-technical skill (NTS) evaluation of nurses in emergency department were conducted before and after training. RESULTS: Finally 43 cases were enrolled in the control group, improvement group and observation group, respectively. There was no statistically significant difference in baseline data including the gender, age, primary site, heart rate, systolic blood pressure, acute physiology and chronic health evaluation II (APACHE II) score, sequential organ failure assessment (SOFA) score, mechanical ventilation ratio among the three groups with comparability. With the gradual improvement of the CRM system, the implementation rate of 1-hour bundle was gradually increased, and the implementation rate in the control group, improvement group and observation group were 65.12% (28/43), 74.42% (32/43) and 88.37% (38/43), respectively, with statistically significant difference (P < 0.05). It was mainly reflected in the completion rate of blood culture, antibiotic usage rate, Lac detection rate and hypertensors usage rate within 1 hour, which were significantly higher in the observation group than those in the control group [completion rate of blood culture: 90.70% (39/43) vs. 62.79% (27/43), antibiotic usage rate: 88.37% (38/43) vs. 60.47% (26/43), Lac detection rate: 93.02% (40/43) vs. 72.09% (31/43), hypertensors usage rate: 88.37% (38/43) vs. 60.47% (26/43), all P < 0.05]. The fluid resuscitation rates within 1 hour in the three groups were all over 90%, with no statistically significant difference among the three groups. The recognition and diagnosis time in the observation group was significantly shorter than that in the control group and the improvement group (hours: 0.41±0.15 vs. 0.61±0.21, 0.51±0.18, both P < 0.05), the correction rate of hypoxemia and 28-day survival rate were significantly higher than those in the control group [correction rate of hypoxemia: 95.35% (41/43) vs. 74.42% (32/43), 28-day survival rate: 83.72% (36/43) vs. 60.47% (26/43), both P < 0.05], and ICU occupancy rate was significantly lower than that in the control group [72.09% (31/43) vs. 93.02% (40/43), P < 0.05]. After training in the CRM system, the score of the sepsis awareness survey questionnaire for emergency department nurses was significantly increased as compared with before training (60.42±5.29 vs. 44.17±9.21, P < 0.01), and NTS also showed significant improvement. CONCLUSION CRM plays a significant role in promoting the implementation of sepsis hour-1 bundle, which can improve the implementation rate of hour-1 bundle and NTS of medical staff, effectively improve patients' hypoxemia, reduce patients' ICU occupancy rate and 28-day risk of death.
Humans ; Sepsis/therapy* ; Emergency Service, Hospital ; Patient Care Bundles ; Intensive Care Units ; Female ; Male ; Middle Aged

From the perspective of Chinese medicine classics and traditional theories,this paper discusses the pathways of acu-puncture-moxibustion treatment for brain diseases,providing a reference for constructing an acupuncture-moxibustion diagnosis and treatment system for brain diseases that aligns with the characteristics of Chinese medicine.First,it reviews the understanding of the brain and brain diseases by physicians throughout history,classifies the conditions included in modern Chinese medicine's scope of brain diseases,and defines the category of brain diseases in Chinese medicine.Based on Chinese medicine classics and the theories and practices of physicians from ancient to modern times,it elaborates on the meridian pathways,viscus pathways and body structure pathways for acupuncture-moxibustion treatment of brain diseases,which includes treatment based on the Governor Vessel,the Foot-Taiyang Bladder Meridian,the Foot-Yangming Stomach Meridian,and other meridians,treatment based on the heart,kidney,intes-tines,and other viscera related to the brain,and treatment based on the characteristics of the body structure related to the cranium and brain and the brain-orifice connection.Finally,it emphasizes the significance of these pathways in enriching the clinical methods of acupuncture-moxibustion for treating brain diseases and highlights the importance of establishing a syndrome differentiation and treat-ment system for brain diseases that reflects the unique features of acupuncture-moxibustion.