[Clinicopathological features of Sjogren's syndrome complicated with liver injury].

Objective: To study the clinicopathological features of Sjogren's syndrome (SS) with liver injury and to improve the understanding of this disease. Methods: Forty-nine patients with SS complicated with liver injury were collected from Beijing Ditan Hospital, Capital Medical University from October 2008 to January 2022. All patients underwent ultrasound-guided liver biopsy, and all specimens were stained with HE. The histopathologic characteristics were observed and the pathologic indexes were graded. Immunohistochemical stains for CK7, CK19, CD38, MUM1 and CD10 were performed by EnVision method; and special histochemical stains for reticulin, Masson's trichrome, Rhodanine, Prussian blue, periodic acid Schiff (PAS) and D-PAS stains were conducted. Results: The age of patients ranged from 31 to 66 years, including 3 males and 46 females. SS combined with drug-induced liver injury was the most common (22 cases, 44.9%), followed by autoimmune liver disease (13 cases, 26.5%, including primary biliary cholangitis in eight cases, autoimmune hepatitis in 3 cases, and PBC-AIH overlap syndrome in 2 cases), non-alcoholic fatty liver disease (NAFLD, 9 cases, 18.4%) and other lesions (5 cases, 10.2%; including 3 cases of nonspecific liver inflammation, 1 case of liver amyloidosis, and 1 case of porto-sinusoidal vascular disease). Among them, 28 cases (57.1%) were associated with obvious interlobular bile duct injury, mainly in SS combined with PBC group and drug-induced liver injury group. Twenty-three cases (46.9%) were associated with hepatocyte steatosis of varying degrees. In SS with autoimmune liver disease group, ISHAK score, degree of fibrosis bile duct injury, bile duct remodeling, lymphocyte infiltration of portal area, and plasma cell infiltration, MUM1 and CD38 expression; serum ALP and GGT, IgM; elevated globulin; positive AMA, proportion of AMA-M2 positive and IgM positive were all significantly higher than those in other groups(all P<0.05). Serum ALT, direct bilirubin and SSA positive ratio in SS combined with drug liver group were significantly higher than those in other groups(all P<0.05). The serum total cholesterol level in SS combined with PBC group (P=0.006) and NALFD group (P=0.011) were significantly higher than those in other groups (P<0.05). Conclusions: The pathologic manifestations of SS patients with liver injury are varied. The inflammatory lesions of SS patients with autoimmune liver disease are the most serious, and the inflammatory lesions of SS patients with non-alcoholic fatty liver disease and non-specific inflammation are mild. Comprehensive analysis of liver histopathologic changes and laboratory findings is helpful for the diagnosis of SS complicated with different types of liver injury.

Effect of avian influenza virus subtype H9N2 on the expression of complement-associated genes in chicken erythrocytes.

1. The H9N2 subtype avian influenza virus can infect both chickens and humans. Previous studies have reported a role for erythrocytes in immunity. However, the role of H9N2 against chicken erythrocytes and the presence of complement-related genes in erythrocytes has not been studied. This research investigated the effect of H9N2 on complement-associated gene expression in chicken erythrocytes.2. The expression of complement-associated genes (C1s, C1q, C2, C3, C3ar1, C4, C4a, C5, C5ar1, C7, CD93 and CFD) was detected by reverse transcription-polymerase chain reaction (RT-PCR). Quantitative Real-Time PCR (qRT-PCR) was used to analyse the differential expression of complement-associated genes in chicken erythrocytes at 0 h, 2 h, 6 h and 10 h after the interaction between H9N2 virus and chicken erythrocytes in vitro and 3, 7 and 14 d after H9N2 virus nasal infection of chicks.3. Expression levels of C1q, C4, C1s, C2, C3, C5, C7 and CD93 were significantly up-regulated at 2 h and significantly down-regulated at 10 h. Gene expression levels of C1q, C3ar1, C4a, CFD and C5ar1 were seen to be different at each time point. The expression levels of C1q, C4, C1s, C2, C3, C5, C7, CFD, C3ar1, C4a and C5ar1 were significantly up-regulated at 7 d and the gene expression of levels of C3, CD93 and C5ar1 were seen to be different at each time point.4. The results confirmed that all the complement-associated genes were expressed in chicken erythrocytes and showed the H9N2 virus interaction with chicken erythrocytes and subsequent regulation of chicken erythrocyte complement-associated genes expression. This study reported, for the first time, the relationship between H9N2 and complement system of chicken erythrocytes, which will provide a foundation for further research into the prevention and control of H9N2 infection.

Morbidity and risk factors for multiple noncommunicable diseases in Northern China: a cross-sectional survey.

OBJECTIVE: This study aimed to determine the morbidity and comorbidity of glucolipid metabolic multiple noncommunicable diseases in a Chinese natural population and associated risk factors. SUBJECTS AND METHODS: A cross-sectional survey with randomized sampling was conducted on a typical sample of 4,002 residents (aged 26-76 years) in the Pinggu District of Beijing. They were subjected to a questionnaire survey, physical examination, and laboratory examination to collect data. Multivariable analysis was used to establish the association between various risk factors and multiple noncommunicable diseases. RESULTS: The overall prevalence rate of chronic glucolipid metabolic noncommunicable diseases was 84.28%. The most common type of noncommunicable diseases was dyslipidemia, abdominal obesity, hypertension, obesity, and type 2 diabetes. The prevalence rate of multiple noncommunicable diseases was 79.60%. Participants with dyslipidemia were at higher risk for underlying chronic diseases. Younger men and women after menopause were more likely to have multiple noncommunicable diseases compared to their older and younger counterparts, respectively. The results of multivariate logistic regression indicated that age above 50, male sex, high household income, low education level, and harmful alcohol consumption were independent risk factors for multiple noncommunicable diseases. CONCLUSIONS: The prevalence of chronic glucolipid metabolic noncommunicable diseases in Pinggu was higher than at the national level. Men with multiple noncommunicable diseases were younger, while women after menopause were more likely to suffer from multiple noncommunicable diseases and the prevalence rate was higher than in men. Intervention programs that aim to target risk factors by sex and region-specific are urgently needed.

Expert surgeons and deep learning models can predict the outcome of surgical hemorrhage from 1 min of video.

Major vascular injury resulting in uncontrolled bleeding is a catastrophic and often fatal complication of minimally invasive surgery. At the outset of these events, surgeons do not know how much blood will be lost or whether they will successfully control the hemorrhage (achieve hemostasis). We evaluate the ability of a deep learning neural network (DNN) to predict hemostasis control ability using the first minute of surgical video and compare model performance with human experts viewing the same video. The publicly available SOCAL dataset contains 147 videos of attending and resident surgeons managing hemorrhage in a validated, high-fidelity cadaveric simulator. Videos are labeled with outcome and blood loss (mL). The first minute of 20 videos was shown to four, blinded, fellowship trained skull-base neurosurgery instructors, and to SOCALNet (a DNN trained on SOCAL videos). SOCALNet architecture included a convolutional network (ResNet) identifying spatial features and a recurrent network identifying temporal features (LSTM). Experts independently assessed surgeon skill, predicted outcome and blood loss (mL). Outcome and blood loss predictions were compared with SOCALNet. Expert inter-rater reliability was 0.95. Experts correctly predicted 14/20 trials (Sensitivity: 82%, Specificity: 55%, Positive Predictive Value (PPV): 69%, Negative Predictive Value (NPV): 71%). SOCALNet correctly predicted 17/20 trials (Sensitivity 100%, Specificity 66%, PPV 79%, NPV 100%) and correctly identified all successful attempts. Expert predictions of the highest and lowest skill surgeons and expert predictions reported with maximum confidence were more accurate. Experts systematically underestimated blood loss (mean error - 131 mL, RMSE 350 mL, R2 0.70) and fewer than half of expert predictions identified blood loss > 500 mL (47.5%, 19/40). SOCALNet had superior performance (mean error - 57 mL, RMSE 295 mL, R2 0.74) and detected most episodes of blood loss > 500 mL (80%, 8/10). In validation experiments, SOCALNet evaluation of a critical on-screen surgical maneuver and high/low-skill composite videos were concordant with expert evaluation. Using only the first minute of video, experts and SOCALNet can predict outcome and blood loss during surgical hemorrhage. Experts systematically underestimated blood loss, and SOCALNet had no false negatives. DNNs can provide accurate, meaningful assessments of surgical video. We call for the creation of datasets of surgical adverse events for quality improvement research.

Utility of the Simulated Outcomes Following Carotid Artery Laceration Video Data Set for Machine Learning Applications.

Importance: Surgical data scientists lack video data sets that depict adverse events, which may affect model generalizability and introduce bias. Hemorrhage may be particularly challenging for computer vision-based models because blood obscures the scene. Objective: To assess the utility of the Simulated Outcomes Following Carotid Artery Laceration (SOCAL)-a publicly available surgical video data set of hemorrhage complication management with instrument annotations and task outcomes-to provide benchmarks for surgical data science techniques, including computer vision instrument detection, instrument use metrics and outcome associations, and validation of a SOCAL-trained neural network using real operative video. Design, Setting, and Participants: For this quailty improvement study, a total of 75 surgeons with 1 to 30 years' experience (mean, 7 years) were filmed from January 1, 2017, to December 31, 2020, managing catastrophic surgical hemorrhage in a high-fidelity cadaveric training exercise at nationwide training courses. Videos were annotated from January 1 to June 30, 2021. Interventions: Surgeons received expert coaching between 2 trials. Main Outcomes and Measures: Hemostasis within 5 minutes (task success, dichotomous), time to hemostasis (in seconds), and blood loss (in milliliters) were recorded. Deep neural networks (DNNs) were trained to detect surgical instruments in view. Model performance was measured using mean average precision (mAP), sensitivity, and positive predictive value. Results: SOCAL contains 31 443 frames with 65 071 surgical instrument annotations from 147 trials with associated surgeon demographic characteristics, time to hemostasis, and recorded blood loss for each trial. Computer vision-based instrument detection methods using DNNs trained on SOCAL achieved a mAP of 0.67 overall and 0.91 for the most common surgical instrument (suction). Hemorrhage control challenges standard object detectors: detection of some surgical instruments remained poor (mAP, 0.25). On real intraoperative video, the model achieved a sensitivity of 0.77 and a positive predictive value of 0.96. Instrument use metrics derived from the SOCAL video were significantly associated with performance (blood loss). Conclusions and Relevance: Hemorrhage control is a high-stakes adverse event that poses unique challenges for video analysis, but no data sets of hemorrhage control exist. The use of SOCAL, the first data set to depict hemorrhage control, allows the benchmarking of data science applications, including object detection, performance metric development, and identification of metrics associated with outcomes. In the future, SOCAL may be used to build and validate surgical data science models.

[Characteristics of high-risk cardiovascular disease among residents aged 35-75 in 8 districts of Beijing].

Objective: To understand the characteristics and current situation of high risks related to cardiovascular disease among residents aged 35-75 in Beijing and to provide scientific reference for the formulation and improvement of cardiovascular disease prevention and control strategies and measures. Methods: According to the data of the Cardiovascular Disease Screening and Management Program in Beijing, 93 520 participants aged 35-75 in 8 districts of Beijing were selected for analysis. We used the χ2 test to compare the high-risk prevalence of cardiovascular disease in different population characteristics. The multivariate logistic regression model was used to analyze the relationship between population characteristics and the high risks of cardiovascular disease. Results: The prevalence of high-risk cardiovascular disease was 20.82% (19 471/93 520). The prevalence of high-risk population in the 65-75 years-old was significantly higher than those of other age groups [29.05% (5 151/17 733), χ2=3 359.37, P<0.001], and the prevalence increased with age (trend χ2=3 121.75, P<0.05). The prevalence of high risk in males was significantly higher than that of women (31.19%, 10 752/34 476 vs. 14.77%, 8 719/59 044, χ2=3 559.87, P<0.05). The most common clustered risk factors appeared as hypertension and diabetes (29.80%, 5 802/19 471), hypertension with smoking (37.84%, 4 069/10 752) in males, and hypertension with diabetes mellitus in females (49.32%, 4 300/8 719), in urban areas (33.62%, 2 571/7 647) and in suburbs (27.33%, 3 231/11 824). Lower education [high school (OR=1.56,95%CI:1.46-1.66), middle school (OR=1.99,95%CI:1.88-2.12), primary school and below (OR=2.28,95%CI:2.12-2.45)], non-Han ethnicity (OR=1.19, 95%CI: 1.07-1.33), unmarried (OR=1.16, 95%CI: 1.08-1.24), drinking alcohol (OR=3.06, 95%CI: 2.94-3.19), obesity (OR=1.85, 95%CI: 1.77-1.93), overweight (OR=1.41, 95%CI: 1.36-1.47), etc., were positively correlated with the high risk of cardiovascular disease. Conclusions: We noticed that the prevalence of high-risk groups of cardiovascular disease aged 35-75 years was around 20% in Beijing, and the proportion in males was higher than females. Low education, drinking, overweight, and obesity were positively correlated with the risks of cardiovascular disease.

miRNA expression analysis of the sheep follicle during the prerecruitment, dominant, and mature stages of development under FSH stimulation.

Follicle stimulating hormone (FSH) has been widely reported to influence ovarian follicular development, and miRNAs play a significant role in mammalian follicular development by regulating their target genes. Therefore, it is of interest to explore the roles of miRNAs in sheep follicular development during FSH stimulation. In the current study, we constructed miRNA expression profiles of small follicles (SFs, prerecruitment stage), medium follicles (MFs, dominance stage), and large follicles (LFs, maturation stage). Three and 50 significant differentially expressed miRNAs (DEMs) were identified in the MF vs. SF and LF vs. SF comparisons, respectively, and none were identified in the LF vs. MF comparison. Oar-miR-10a was significantly downregulated in MFs compared with SFs. In LFs compared with SFs, miR-212-3p, miR-212-5p and miR-202-5p were significantly upregulated, and miR-27a-3p, miR-181a-5p, miR-204-5p, and miR-182-5p were significantly downregulated. Furthermore, we predicted the target genes of significant DEMs and performed functional enrichment analyses of these target genes. Analyses of KEGG pathways and GO terms showed that the putative target genes were significantly enriched in ovarian steroidogenesis, glutathione metabolism, positive regulation of cell differentiation, positive regulation of cell development, and cellular response to oxygen-containing compounds. Analyses of miRNA-gene regulatory networks suggested that miR-181a-5p-CYP11A1, (miR-27a-3p and miR-129-5p)-LDLR, (miR-212-3p and miR-212-5p)-EFNA5, (miR-181a-5p, miR-182-5p, and miR-27a-3p)-INHBA, and miR-182-5p-SOD2 might be involved in follicular development. The present study provides basic data and suggests research directions for further exploration of the roles of miRNAs in sheep follicular development under FSH stimulation.

[Evaluation on effectiveness of community-based comprehensive intervention and internet-based intervention for vaccination among patients with chronic diseases].

To evaluate the effectiveness of community-based comprehensive intervention and internet-based intervention on influenza and pneumococcal vaccination in patients with diabetes or chronic obstructive pulmonary disease (COPD) in communities. Before the onset of influenza in 2015 and 2016, five demonstration areas for comprehensive prevention and control of chronic diseases were selected in the study. Patients aged 35 years old and above with diabetes and COPD were selected from selected demonstration areas as study participants. Two communities were selected from each demonstration area and randomly divided into Group A with the community-based comprehensive intervention and Group B with the combination of community-based comprehensive intervention and internet-based intervention within 6 months. The differences of vaccination knowledge, vaccination intention and vaccination proportion in study participants between the two groups before and after interventions were analyzed and compared. A total of 15 226 patients were included in the study. After the intervention the increase in the vaccine knowledge score, vaccination intention and vaccination proportion of patients in the two groups were all higher than those before the intervention. After intervention, the increase of the median score of vaccine knowledge in group B (50.00 points) was higher than that in Group A (42.86 points). The increase of influenza vaccination intention in Group B (6.91%) was higher than that in Group A (4.16%). The increase of proportion of influenza vaccination in Group B (7.35%) was lower than that in Group A (16.61%). The increase of pneumococcal vaccination intention in Group B (7.90%) was lower than that in Group A (9.08%). The proportion of pneumococcal vaccination in Group B (3.37%) was lower than that in Group A (4.06%). Community-based comprehensive intervention could improve the level of vaccine knowledge, vaccination intention and vaccination proportion of patients in this study. Combined with community-based intervention, internet-based intervention could have a better effect on improving vaccine knowledge and influenza vaccination intention, but its impact on vaccination proportion needs to be further explored.

Impact of Spectators on a Two-Qubit Gate in a Tunable Coupling Superconducting Circuit.

Cross-resonance (CR) gates have emerged as a promising scheme for fault-tolerant quantum computation with fixed-frequency qubits. We experimentally implement an entangling CR gate by using a microwave-only control in a tunable coupling superconducting circuit, where the tunable coupler provides extra degrees of freedom to verify optimal conditions for constructing a CR gate. By developing a three-qubit Hamiltonian tomography protocol, we systematically investigate the dependency of gate fidelities on spurious qubit interactions and present the first experimental approach to the evaluation of the perturbation impact arising from spectator qubits. Our results reveal that the spectator qubits lead to reductions in CR gate fidelity dependent on ZZ interactions and particular frequency detunings between spectator and gate qubits. The target spectator demonstrates a more serious impact than the control spectator under a standard echo pulse scheme, whereas the degradation of gate fidelity is observed up to 22.5% as both the spectators are present with a modest ZZ coupling to the computational qubits. Our experiments uncover an optimal CR operation regime, and the method we develop here can readily be applied to improving other kinds of two-qubit gates in large-scale quantum circuits.

[Analysis of factors associated with pulmonary embolism during pregnancy].

To investigate risk factors associated with pulmonary embolism in women during pregnancy or peripartum. Information of 7 pregnant women diagnosed pulmonary embolism (PE) were collected from January 2018 to December 2020 in Beijing Friendship Hospital. Four non-PE pregnant women as control group were matched with each patient by age, BMI and pregnancy complications during the same period. Binary logistic regression was used to identify the potential risk factors by the comparison of difference between groups. The laboratory indices, including platelet, hemoglobin, prothrombin time, partially activated prothrombin time, D-dimer, triglyceride, total cholesterol, low density lipoprotein (LDL) and high density lipoproteinage (HDL) showed no significant difference between groups. D-dimer and HDL were found to be associated with PE during pregnancy (P<0.05). D-dimer and HDL might be associated with PE during pregnancy.

NF-кB pathway genes expression in chicken erythrocytes infected with avian influenza virus subtype H9N2.

1. Chicken erythrocytes in blood vessels are the most abundant circulating cells, which participate in the host's immune responses. The transcription factor nuclear factor-kappa B (NF-κB) plays a vital role in the inflammatory response following viral infections. However, the expression of the NF-κB pathway, and other immune-related genes in chicken erythrocytes infected with low pathogenic avian influenza virus (LPAIV H9N2), has not been extensively studied.2. The following study determined the interaction of LPAIV H9N2 with chicken erythrocytes using indirect immunofluorescence microscopy. This was followed by investigating myeloid differentiation primary response 88 (MyD88), C-C motif chemokine ligand 5 (CCL5), melanoma differentiation-associated protein 5 (MDA5), the inhibitor of nuclear factor-kappa B kinase subunit epsilon (IKBKE), NF-κB inhibitor alpha (NFKBIA), NF-κB inhibitor epsilon (NFKBIE), interferon-alpha (IFN-α), colony-stimulating factor 3 (CSF3) and tumour necrosis factor receptor-associated factor 6 (TRAF6) by mRNA expression using quantitative real-time PCR (qRT-PCR) at four different time intervals (0, 2, 6 and 10 h).3. There was a significant interaction between erythrocytes and LPAIV H9N2 virus. Furthermore, the mRNA expression of the NF-κB pathway and other immune-related genes were significantly up-regulated at 2 h post-infection in infected chicken erythrocytes, except for TRAF6, which were significantly downregulated. While at 0 h post-infection, IFN-α and CSF3 were significantly upregulated, whereas NFKBIA was significantly downregulated. Further expression of MDA5, CCL5 and NFKBIA was upregulated, while TRAF6 was downregulated at 6 h post-infection. In infected erythrocytes, expression of MyD88, CCL5 and IKBKE was upregulated. However, IFN-α and TRAF6 were downregulated at 10 h post-infection.4. These results give initial evidence that the NF-κB pathway, and other genes related to immunity, in chicken erythrocytes may contribute to LPAIV subtype H9N2 and induce host immune responses.

[Urinary retinol binding protein and ß2-microglobulin were associated with urinary albumin to creatinine ratio and renal function in hospitalized diabetic patients].

Objective: To explore the associations of urinary retinol binding protein (RBP) and ß2-microglobulin (ß2-MG) with urinary albumin to creatinine ratio (UACR) and renal function in hospitalized patients with type 2 diabetes mellitus (T2DM). Methods: A total of 1 030 Chinese patients with T2DM were included in this study. The subjects were divided into the UACR normal group (<30 mg/g), microalbuminuria group (30-300 mg/g) and macroalbuminuria group (>300 mg/g). Patients with normal UACR were further divided into two groups according to the estimated glomerular filtration rate (eGFR): the eGFR low group (<90 ml·min-1·1.73m-2) and the normal eGFR group (≥90 ml·min-1·1.73m-2). Urine RBP and ß2-MG levels among the groups were compared. Multiple linear regression analyses were applied to evaluate risk factors of urine RBP and ß2-MG. Results: In all patients (n=1 030), urine RBP and ß2-MG increased gradually with the increase of UACR across the three groups, the proportions of abnormal urine RBP (>0.7 mg/L) and ß2-MG (>370 µg/L) in these groups were 3.8%, 8.5%, 39.0% (P<0.001), and 12.9%, 26.7%, 46.8% (P<0.001), respectively. In the UACR normal group (n=788), 12.2% of the patients were with eGFR<90 ml·min-1·1.73m-2. The proportion of abnormal ß2-MG (>370 µg/L) was higher in the eGFR low group than that in the eGFR normal group (29.2% vs. 10.7%, P<0.001). Multivariate linear stepwise regression analyses were performed using natural logarithm of urine RBP or ß2-MG as dependent variable, and showed that urine RBP was independently associated with UACR (ß=0.0005, P<0.001), serum creatinine (ß=0.006, P<0.001) and glycosylated hemoglobin A1c (ß=0.050, P=0.001), and ß2-MG was independently correlated with UACR (ß=0.000 4, P<0.001), serum creatinine (ß=0.011, P<0.001), systolic blood pressure (ß=0.005, P=0.031) and fasting blood-glucose (ß=0.027, P=0.046). Conclusions: Urine RBP and ß2-MG are positively associated with high UACR and impaired renal function in T2DM patients, and these changes could occur before UACR and eGFR turned out to be abnormal. It is recommended that urine RBP and ß2-MG be detected as early as possible to identify diabetic kidney disease in patients with normal UACR and eGFR.

[The effect of peripheral blood cell score on the prognosis of multiple myeloma patients treated with bortezomib].

Objective: This study aims to evaluate the prognostic effect of peripheral blood cells in multiple myeloma (MM) patients treated with bortezomib. Methods: The clinical data of 155 newly diagnosed MM patients in two blood disease treatment centers from January 2014 to December 2016 were retrospectively studied. All patients received bortezomib as the first-line treatment. The results of the peripheral blood cell counts, including absolute neutrophil count, absolute monocyte count (AMC) , hemoglobin level, mean corpuscular volume (MCV) , and platelet count, and other clinical features were analyzed. Results: AMC (>0.6×10(9)/L) , MCV (>99.1 fl) , and platelet count (<150×10(9)/L) significantly affected patients' PFS and OS. The above three factors were assigned 1 point, respectively, to form the blood cell score. The analysis showed that 64 cases (41.3% ) had a score of 0, 57 cases (36.8% ) had 1, 32 cases (20.6% ) had 2, and 2 cases (1.3% ) had 3. The median PFS of the four groups were 42.8 m, 26.5 m, 15.8 m, and 6.4 m, respectively (P<0.001) . The median OS were NR, 48.2 m, 31.1 m, and 31.4 m, respectively (P=0.001) . Multivariate analysis suggested that the blood cell score (2-3 vs 0-1) and the proportion of marrow plasma cells (>30% ) were independent prognostic factors for PFS (HR=1.95 and 1.76, respectively) , while age (>65y vs ≤65y) , R-ISS stage (3 vs 1-2) , and blood cell score (2-3 vs 0-1) were independent prognostic factors for OS (HR=2.08, 2.13 and 2.12, respectively) . Conclusion: As an easy-to-access biomarker, the blood cell score can be used to evaluate the prognosis of newly diagnosed MM patients in the era of new drugs, but it is still necessary to expand the cases and make further confirmation in the prospective study.

Prevalence of neural collapse during the terminal phase of deep learning training.

Modern practice for training classification deepnets involves a terminal phase of training (TPT), which begins at the epoch where training error first vanishes. During TPT, the training error stays effectively zero, while training loss is pushed toward zero. Direct measurements of TPT, for three prototypical deepnet architectures and across seven canonical classification datasets, expose a pervasive inductive bias we call neural collapse (NC), involving four deeply interconnected phenomena. (NC1) Cross-example within-class variability of last-layer training activations collapses to zero, as the individual activations themselves collapse to their class means. (NC2) The class means collapse to the vertices of a simplex equiangular tight frame (ETF). (NC3) Up to rescaling, the last-layer classifiers collapse to the class means or in other words, to the simplex ETF (i.e., to a self-dual configuration). (NC4) For a given activation, the classifier's decision collapses to simply choosing whichever class has the closest train class mean (i.e., the nearest class center [NCC] decision rule). The symmetric and very simple geometry induced by the TPT confers important benefits, including better generalization performance, better robustness, and better interpretability.

[Trends regarding the 30-day readmission rates in patients discharged for acute myocardial infarction in Beijing].

Objective: To examine the characteristics and trends regarding the 30-day coronary heart disease (CHD) readmission rates in patients discharged for acute myocardial infarction (AMI) in Beijing, during 2007-2012. Methods: Patients hospitalized for AMI in Beijing from 1 January 2007 to 31 December 2012 were identified from "The Cardiovascular Disease Surveillance System in Beijing". In total, 64 355 patients aged 25 years and over with permanent Beijing residency survived and discharged for AMI in Beijing during the above-said six years. After excluding duplicate and validation for the completeness and accuracy of the records, clinical features and 30-day CHD readmission rates for those AMI discharged patients were analyzed. Trends regarding the 30-day CHD readmission rates in patients discharged for AMI were analyzed by Poisson regression models. Results: The overall age-standardized average 30-day CHD readmission rate for AMI discharged patients was 7.7% in patients aged 25 years and over in Beijing. During the six years under study, an increasing trend was observed on the 30-day CHD readmission rates for AMI discharged patients after adjusting the age and gender (P<0.001). The age-standardized 30-day CHD readmission rates for AMI discharged patients increased by 21.3% in the past six years, from 7.0% in 2007 to 8.5% in 2012. The increase of 30-day CHD readmission rates was noted in both men and women during the six years, whereas it appeared higher in women (8.4%) than in men (7.4%), after adjusting for age. Among the AMI discharged patients, the 30-day CHD readmission rates were higher in patients with non-ST-segment elevation myocardial infarction (NSTEMI) than those with ST-segment elevation myocardial infarction patients (P<0.01), and higher in discharged patients with multiple comorbidities than those without multiple comorbidities (P<0.01). Conclusions: An increasing trend in the 30-day CHD readmission rates for AMI discharged patients was observed during 2007-2012 among Beijing residents aged 25 years and over. It called for an urgent need in improving the secondary prevention of AMI discharged patients, particularly in women, with NSTEMI and those with multiple comorbidities. Findings from these unselected "real-world" data in Beijing may help to guide the management of AMI in the country as well as in other developing countries.

[Analysis of factors affecting olfactory disfunctions in patients with chronic rhinosinusitis with nasal polyps].

Objective: To analyze the factors affecting olfactory disfunctions in patients with chronic rhinosinusitis with nasal polyps (CRSwNP). Methods: This was a retrospective analysis. Eighty-eight patients with CRSwNP who underwent endoscopic sinus surgery in Beijing Anzhen Hospital from 2014 to 2018 were enrolled, including 22 males and 66 females, with the age of (48.1±11.3) years old(Mean±SD). Sniffin' Sticks olfactory test, Lund-Mackay score and modified sinus CT olfactory cleft score, nasal resistance and acoustic reflex examination, blood routine and blood biochemistry test, serum specific IgE test were performed before surgery and nasal polyps of all patients were collected for eosinophil count during surgery. According to bilateral total TDI score, the patients were divided into normal olfactory function group and olfactory disfunction group. The clinical baseline data were compared between the two groups. According to the results of single factor analysis, factors which were significant different between the two groups and clinically useful indicators were further included in the multivariate Logistic regression model analysis, then a model predicting olfactory disfunction in patients with CRSwNP was initially established. P<0.05 was considered statistically significant. Results: Among 88 patients with CRSwNP, 32 (36.4%) patients were with normal olfaction and 56 (63.6%) patients were with olfactory disfunction, including 40 (45.5%) of hyposmia and 16 (18.2%) of anosmia. Tissue eosinophil count, blood eosinophil percentage and blood urea concentration had significant difference between the two groups (12.7[2.0, 52.3]/HP (M[P(25), P(75)]) vs 38.6[16.2, 87.0]/HP, 2.75[1.60, 4.80]% vs 4.35[2.50, 6.60]%, (5.56±1.15) mmol/L vs (4.98±1.33) mmol/L, all P<0.05). Modified sinus CT olfactory cleft score and Lund-Mackay score except for ostiomeatal complex score were statistically significant between the two groups (all P<0.05). Multivariate Logistic regression analysis showed that the bilateral and total olfactory cleft score and blood urea concentration were statistically significant, in addition, the bilateral and total olfactory cleft score was a risk factor (OR=2.108, 95%CI: 1.407-3.159, P<0.001) and blood urea within a certain concentration was a protective factor (OR=0.461, 95%CI: 0.240-0.884, P=0.020). Further studies found that the area under the ROC curve of the model with tissue eosinophil count, blood eosinophil percentage, bilateral and total olfactory cleft score, total inspiratory volume and blood urea concentration was 0.888 (P<0.01), which had good predictive value for olfactory disorders in CRSwNP. Conclusions: The modified sinus CT olfactory cleft score is closely related to the olfactory disorders in patients with CRSwNP. A certain degree of elevated blood urea concentration may have a protective effect on the olfactory function of patients with CRSwNP.

[Application of the superior thyroid artery peak systolic velocity in differentiating Graves' disease from autoimmune thyroiditis].

Objective: To evaluate the clinical value of the superior thyroid artery peak systolic velocity (STA-PSV) for the differential diagnosis of autoimmune thyrotoxicosis. Methods: A total of 301 patients with newly diagnosed thyrotoxicosis and without any anti-thyroid drug intervention were collected from the Department of Endocrinology and Metabolism, Peking University People's Hospital from Jan. 2015 to Oct. 2018. Among them, 241 patients were with Graves' disease (GD) and 60 patients were with autoimmune thyroiditis (AIT). STA-PSV, thyroid function and thyrotropin receptor antibody (TRAb) were determined. A multiple linear regression was used to identify factors associated with STA-PSV. A receiver operating characteristic (ROC) curve and area under the curve (AUC) were used to evaluate the discriminating ability of STA-PSV to GD. Results: STA-PSV leves in GD group were significantly higher than those in AIT group [61.00 (41.00, 86.50) cm/s vs. 34.50 (25.25, 46.00) cm/s, P<0.001]. The ROC curve analysis showed that the AUC was 0.790 (95%CI 0.734-0.845), and 49.5cm/s was the optimal cutoff point for the diagnosis of GD, in which the sensitivity was 64.3% and the specificity was 83.3%. In all patients with thyrotoxicosis, multiple linear regression analyses showed free thyroxine (FT(4)) (ß=0.371, 95%CI 0.005-0.010, P<0.001) and TRAb (ß=0.138, 95%CI 0.001-0.014, P=0.035) were positively associated with STA-PSV. Conclusions: The STA-PSV is positively associated with FT(4) and TRAb levels, and it is a helpful marker in differential diagnosis between GD and AIT.