Objective To explore an artificial intelligence (AI)-based method for automated hand hygiene monitoring and to compare the effectiveness of three algorithms (UniFormerV2, TDN, C3D) in recognizing hand hygiene steps in surgical settings, thereby aiding hospital infection control. Methods From April to October 2024, we non-invasively collected 641 video recordings of healthcare staff performing hand hygiene at four-bay scrub sinks in two tertiary hospitals using overhead HD cameras. The dataset was annotated by five trained experts for model training and validation. Results Following training on 385 samples, internal validation (n=119) showed the C3D model achieved 81% accuracy, 87% recall, and an 83% F1-score. The TDN model achieved 93%, 91%, and 92% for the same metrics. The UniFormerV2 model outperformed both, with an accuracy, recall, and F1-score of 93%—an improvement of over 10 percentage points compared to traditional CNNs (TDN, C3D). It also achieved an 84% accuracy in external validation, demonstrating strong generalization. Conclusion The UniFormerV2 model is more accurate than CNN-based models for hand hygiene step recognition and shows robust performance in external validation. It presents a viable tool for healthcare facilities to enhance hand hygiene management, ultimately improving medical quality and patient safety.

Intelligent prescription recommendation has become an important research direction in traditional Chinese medicine (TCM), offering new opportunities to support clinical decision-making and promote the modernization of TCM practice. With the rapid development of artificial intelligence (AI), a variety of computational approaches have been proposed to learn prescription patterns from clinical data and generate personalized treatment recommendations. However, despite increasing research activity, systematic and comprehensive reviews of AI-driven methods for TCM prescription recommendation remain limited. In this study, we present a comprehensive review of computational approaches for herbal prescription recommendation (HPR) in TCM. Existing methods are systematically categorized into several major paradigms, including traditional machine learning methods, topic model methods, sequential generative methods, deep learning and graph-based methods, and large language model–based frameworks. In addition to summarizing methodological developments, we also review commonly used public datasets and evaluation metrics in this field. Furthermore, representative models with publicly available implementations are experimentally evaluated on multiple benchmark datasets to provide a comparative analysis of their performance on the HPR task. Finally, we discuss the key challenges that hinder the practical deployment of intelligent prescription recommendation systems, including data heterogeneity, limited interpretability, and insufficient integration of TCM domain knowledge. Future research directions are outlined to facilitate the development of more reliable, interpretable, and clinically applicable AI-assisted HPR systems for TCM.

Objective To construct large medical model named by "Huaxi HongYi"and explore its application effectiveness in assisting medical record generation. Methods By the way of a full-chain medical large model construction paradigm of "data annotation - model training - scenario incubation", through strategies such as multimodal data fusion, domain adaptation training, and localization of hardware adaptation, "Huaxi HongYi" with 72 billion parameters was constructed. Combined with technologies such as speech recognition, knowledge graphs, and reinforcement learning, an application system for assisting in the generation of medical records was developed. Results Taking the assisted generation of discharge records as an example, in the pilot department, after using the application system, the average completion times of writing a medical records shortened (21 min vs. 5 min) with efficiency increased by 3.2 time, the accuracy rate of the model output reached 92.4%. Conclusion It is feasible for medical institutions to build independently controllable medical large models and incubate various applications based on these models, providing a reference pathway for artificial intelligence development in similar institutions.

Objective::Eosinophils (EOS) are inflammatory innate immune cells that play an important role in arterial thrombogenesis. There is a paucity of data on whether EOS levels have an impact on long-term outcomes following ST-segment elevation myocardial infarction (STEMI). This study aimed to investigate the impact of EOS count on the clinical outcomes of STEMI patients who underwent emergency percutaneous coronary intervention (PCI).Methods::This is a retrospective multicenter cohort study. A total of 754 patients with STEMI who required emergency PCI at 5 centers were screened between October 2015 and November 2016. Patients were divided into 2 groups based on EOS count: EOS count <0.02 × 10 9/L group ( n = 264) and EOS count ≥0.02 × 10 9/L group ( n = 490). Baseline demographic characteristics, clinical information, and medical test data were collected at study entry. The primary endpoint was all-cause death. The secondary endpoint was a major adverse cardiac event. Multivariablete Cox regression analysis was performed to identify the independent predictors of all-cause death, with a follow-up period of 5 years. Results::The incidence of all-cause death (14.8% vs. 7.6%, P = 0.002) and major adverse cardiac event (20.1% vs. 13.1%, P = 0.011) were significantly higher in EOS count <0.02 × 10 9/L group compared with ≥0.02 × 10 9/L group. Multivariablete Cox regression analysis showed that an EOS count <0.02 × 10 9/L, age ≥65 years, previous heart failure, previous stroke, and left ventricular ejection fraction ≤ 40% were independent predictors of all-cause death in patients with STEMI who underwent an emergency PCI. Conclusions::Low EOS counts were associated with all-cause mortality in STEMI patients who underwent emergency PCI.

Objective To observe the expression of GLIS3 in triple negative breast cancer(TNBC)and analyze its relationship with the prognosis of TNBC patients.Methods Bioinformatic analysis was applied to analyze the expression level of GLIS3 in Gene Expression Omnibus(GEO).Kaplan-Meier survival curve was plotted to evaluate the impact of GLIS3 expression on the survival rate of patients based on DNA chip data.A total of 125 patients pathologically diagnosed as TNBC in the Fourth Hospital of Hebei Medical University from January to December 2014 were enrolled by cluster random sampling.Among them,53 patients had complete tissue specimens,medical records and follow-up data.Immunohistochemical assay was used to detect the expression of GLIS3 in TNBC and adjacent tissues to tumors,while the relationships between GLIS3 protein expression in breast cancer tissues and clinicopathological parameters such as age,menstrual status,tumor size,clinical stage,histological grade,pathological type,axillary lymph node metastasis,vascular tumor thrombus,and expression of TP53 and Ki-67 were analyzed.Kaplan-Meier survival curve was plotted to analyze the effect of GLIS3 on the overall and disease-free survival of TNBC patients.Cox regression model was established to identify the risk factors impacting the prognosis of the patients.Results Analysis of GEO data showed that the expression of GLIS3 in TNBC was significantly higher than that in paracancer tissues(P＜0.05).The expression of GLIS3 was notably higher in the TNBC tissue than the adjacent tissue to tumor(P＜0.05).A marked augmentation of GLIS3 expression was observed in both the advanced and larger-sized tumors(P＜0.05).Univariate analysis of Cox regression model revealed that lymph node metastasis,TNM stage and GLIS3 expression were all related to disease-free survival of TNBC patients(P＜0.05).Univariate and multivariate analyses displayed that TNM stage was related to the overall survival of TNBC patients(P＜0.05).The patients with high expression of GLIS3 had significant shorter disease-free survival time than those with low expression(P＜0.05),but had no statistical difference in overall survival(P＞0.05).Conclusion GLIS3 protein is highly expressed in TNBC tissues,and tumor size and TNM stage are correlated with its high expression.The high expression of GLIS3 suggests that the patients have poor prognosis and low disease-free survival rate.

Objective::Eosinophils (EOS) are inflammatory innate immune cells that play an important role in arterial thrombogenesis. There is a paucity of data on whether EOS levels have an impact on long-term outcomes following ST-segment elevation myocardial infarction (STEMI). This study aimed to investigate the impact of EOS count on the clinical outcomes of STEMI patients who underwent emergency percutaneous coronary intervention (PCI).Methods::This is a retrospective multicenter cohort study. A total of 754 patients with STEMI who required emergency PCI at 5 centers were screened between October 2015 and November 2016. Patients were divided into 2 groups based on EOS count: EOS count <0.02 × 10 9/L group ( n = 264) and EOS count ≥0.02 × 10 9/L group ( n = 490). Baseline demographic characteristics, clinical information, and medical test data were collected at study entry. The primary endpoint was all-cause death. The secondary endpoint was a major adverse cardiac event. Multivariablete Cox regression analysis was performed to identify the independent predictors of all-cause death, with a follow-up period of 5 years. Results::The incidence of all-cause death (14.8% vs. 7.6%, P = 0.002) and major adverse cardiac event (20.1% vs. 13.1%, P = 0.011) were significantly higher in EOS count <0.02 × 10 9/L group compared with ≥0.02 × 10 9/L group. Multivariablete Cox regression analysis showed that an EOS count <0.02 × 10 9/L, age ≥65 years, previous heart failure, previous stroke, and left ventricular ejection fraction ≤ 40% were independent predictors of all-cause death in patients with STEMI who underwent an emergency PCI. Conclusions::Low EOS counts were associated with all-cause mortality in STEMI patients who underwent emergency PCI.

BACKGROUND: Limited data are available on the comparison of clinical outcomes of complete vs. incomplete percutaneous coronary intervention (PCI) for patients with chronic total occlusion (CTO) and multi-vessel disease (MVD). The study aimed to compare their clinical outcomes. METHODS: A total of 558 patients with CTO and MVD were divided into the optimal medical treatment (OMT) group ( n = 86), incomplete PCI group ( n = 327), and complete PCI group ( n = 145). Propensity score matching (PSM) was performed between the complete and incomplete PCI groups as sensitivity analysis. The primary outcome was defined as the occurrence of major adverse cardiovascular events (MACEs), and unstable angina was defined as the secondary outcome. RESULTS: At a median follow-up of 21 months, there were statistical differences among the OMT, incomplete PCI, and complete PCI groups in the rates of MACEs (43.0% [37/86] vs. 30.6% [100/327] vs. 20.0% [29/145], respectively, P = 0.016) and unstable angina (24.4% [21/86] vs. 19.3% [63/327] vs. 10.3% [15/145], respectively, P = 0.010). Complete PCI was associated with lower MACE compared with OMT (adjusted hazard ratio [HR] = 2.00; 95% confidence interval [CI] = 1.23-3.27; P = 0.005) or incomplete PCI (adjusted HR = 1.58; 95% CI = 1.04-2.39; P = 0.031). Sensitivity analysis of PSM showed similar results to the above on the rates of MACEs between complete PCI and incomplete PCI groups (20.5% [25/122] vs. 32.6% [62/190], respectively; adjusted HR = 0.55; 95% CI = 0.32-0.96; P = 0.035) and unstable angina (10.7% [13/122] vs. 20.5% [39/190], respectively; adjusted HR = 0.48; 95% CI = 0.24-0.99; P = 0.046). CONCLUSIONS For treatment of CTO and MVD, complete PCI reduced the long-term risk of MACEs and unstable angina, as compared with incomplete PCI and OMT. Complete PCI in both CTO and non-CTO lesions can potentially improve the prognosis of patients with CTO and MVD.
Humans ; Treatment Outcome ; Percutaneous Coronary Intervention/methods* ; Coronary Occlusion/surgery* ; Prognosis ; Angina, Unstable/surgery* ; Chronic Disease ; Risk Factors

Objective:To explore the protective effect of AGK2, a selective inhibitor of sirtuin 2 (SIRT2), on the mitochondria of L02 hepatocytes induced by thioacetamide (TAA) and its related mechanism.Methods:Human-derived hepatocyte line L02 cells were cultured in vitro. Different concentrations of SIRT2 inhibitor AGK2 were used as intervention drugs. Cell counting kit-8 (CCK8) was used to detect the effects of different concentrations of AGK2 on the activity of L02 cells, and the appropriate concentration was selected as the AGK2 intervention group. The normal group was not given any drug intervention. The model group was given 90 mmol/L TAA for modeling. Low, medium and high dose AGK2 groups were added with 1, 2 and 4 μmol/L AGK2, respectively 2 h before modeling. CCK8 was used to detect cell activity in each group. Morphological changes of cells were observed under inverted light microscope. The relative protein expression levels of isocitrate dehydrogenase (IDH1), malate dehydrogenase (MDH1), SIRT2 and fission protein 1 homologue (FIS1) were detected by Western blot. The expression of SIRT2 in cells of each group was observed by confocal laser scanning microscope. The mitochondrial membrane potential of cells in each group was observed under a fluorescence microscope. Results:When AGK2 concentration was 1, 2 and 4 μmol/L, the survival rate of cells were 98.05%, 95.76% and 91.65%, respectively, with no statistical significance compared with normal group (all P>0.05). When AGK2 concentration was 8, 16, 32, 64, 128 μmol/L, the cell survival rate was significantly decreased compared with normal group (all P<0.05). Compared with the model group, the L02 cells in low, medium and high AGK2 groups had better activity and adherence, and the floating cells were significantly reduced. The higher the concentration of AGK2, the better the cell activity and adherence, and the less floating cells. Compared with the model group, the red fluorescence of L02 cells in AGK2 group was enhanced, while the green fluorescence was weakened. The higher the AGK2 concentration was, the stronger the red fluorescence was, and the weaker the green fluorescence was. Compared with the model group, the fluorescence of SIRT2 in L02 cells of low, medium and high AGK2 groups was weakened, and the higher the concentration of AGK2, the weaker the fluorescence of SIRT2. The protein expressions of IDH1 and MDH1 in L02 cells of low, medium and high AGK2 groups were significantly higher than those of model group (all P<0.05), and were positively correlated with the concentration of AGK2 ( r=0.818, P<0.05; r=0.960, P<0.05); the protein expressions of SIRT2 and FIS1 were significantly lower than those of the model group (all P<0.05), and were negatively correlated with the concentration of AGK2 ( r=-0.992, P<0.05; r=-0.998, P<0.05). Conclusions:AGK2 can reduce the mitochondrial membrane potential stimulated by TAA in L02 cells, increase the protein expression of IDH1 and MDH1, and inhibit the protein expression of SIRT2 and FIS1 in L02 cells in a dose-dependent manner.

Objective:To investigate the accuracy, effectiveness and feasibility of MassARRAY genotyping assay in the diagnoses of neonatal genetic metabolic diseases.Methods:This is a retrospective study. From December 2016 to January 2020, newborns were screened by tandem mass spectrometry at the Zhejiang Newborn Screening Center, among which the data of 7 922 suspected positive cases of genetic metabolic diseases were collected. These patients were then tested for the common variants of 27 genetic metabolic diseases by MassARRAY genotyping assay, along with further testing using Sanger or next-generation sequencing used to verify and/or further search for potential variants.Results:A total of 1 408 cases were tested with MassARRAY. Among these, 307 cases were confirmed with certain genetic metabolic diseases. The detection rate of hyperphenylalaninemia was the highest, followed by primary carnitine deficiency, short acyl-coA dehydrogenase deficiency and methylmalonic acidemia. With these cases, the consistency of Sanger sequencing and MassARRAY was 100% (307/307). Another 287 cases were identified as carriers by MassARRAY with a 49.1% (141/287) consistency in reference to Sanger sequencing, mainly involving SLC22A5 and MCCC1 genes. Meanwhile, 50.8% (146/287) of these cases were found to have another variant mainly involving PAH, PTS and ACADS genes. The remaining 814 cases have no variants; 158 cases out of these patients have continuously abnormal amino acids, acyl carnitines, urine organic acid and/or other biochemical indices, and were tested by next-generation sequencing, among which 38% (60/158) were detected with two variants. In this study, a total of 513 patients with genetic metabolic disease were diagnosed, and the detection rate of MassARRAY was 59.8% (307/513). Conclusions:MassARRAY genotyping assay can be used as an early molecular screening method for neonatal genetic metabolic diseases. The detection rate is particularly high in diseases with a high concentration of hotspot variants, such as hyperphenylalaninemia and primary carnitine deficiency. The future application value of MassARRAY should be further improved by continuously optimizing its ability to identify new disease genes and potential variable sites.

This paper reported a case of neonatal radial nerve palsy that was successfully treated by conservative therapy. The patient with 39 weeks of gestational age was born vaginally. On the 2nd day after birth, right wrist drop, decreased muscle strength in the right upper limb, the erythema patch, and the subcutaneous nodule on the upper arm were observed. Electromyography revealed acute denervation of the radial nerve. Based on the electromyography results combined with clinical evaluations, neonatal radial nerve palsy was diagnosed. The patient was treated with self-made simple splint fixation along with comprehensive treatment. At 15 days of age, the family members removed the splint fixation themselves (13 days of fixation). At the age of 20 days, the symptoms of right wrist drop had disappeared, and the grasp reflex and Moro reflex of both hands returned to normal. A follow-up electromyography conducted at six months after discharge showed no obvious abnormalities.