ABSTRACT
During the COVID-19 coronavirus epidemic, people usually wear masks to prevent the spread of the virus, which has become a major obstacle when we use face-based computer vision techniques such as face recognition and face detection. So masked face inpainting technique is desired. Actually, the distribution of face features is strongly correlated with each other, but existing inpainting methods typically ignore the relationship between face feature distributions. To address this issue, in this paper, we first show that the face image inpainting task can be seen as a distribution alignment between face features in damaged and valid regions, and style transfer is a distribution alignment process. Based on this theory, we propose a novel face inpainting model considering the probability distribution between face features, namely Face Style Self-Transfer Network (FaST-Net). Through the proposed style self-transfer mechanism, FaST-Net can align the style distribution of features in the inpainting region with the style distribution of features in the valid region of a face. Ablation studies have validated the effectiveness of FaST-Net, and experimental results on two popular human face datasets (CelebA and VGGFace) exhibit its superior performance compared with existing state-of-the-art methods. © 2023 SPIE.
ABSTRACT
Background. ADI is a fully human IgG1 monoclonal antibody engineered to have an extended half-life with high potency and broad neutralization against SARS-CoV-2 and other SARS-like coronaviruses. The goal of our analysis was to develop a QSP model in which ADI concentrations in upper airway (UA) epithelial lining fluid (ELF) were linked to a viral dynamic model to describe the impact of ADI on SARS-CoV-2 viral load relative to placebo. Methods. The QSP model was fit inNONMEMVersion 7.4 using PK data from a Phase 1 study (N=24, IV and IM) and from Phase 2/3 COVID-19 prevention (EVADE;N=659, IM) and treatment (STAMP;N=189, IM) studies. Saliva and NP samples were collected from STAMP study participants (pts) infected with the delta or omicron variants. The viral dynamic model was based on a published model and was modified to include both active (V) and deactivated (DV) virus (Fig). The viral dynamic model was fit to the NP swab viral load data (2 samples/pt) standardized to time since infection based upon recorded symptom onset. Saliva data (7-8 samples/ pt) was fit sequentially using a biophase compartment given the peak viral load was modestly lower and peaked later than Day 1. Viral dynamic model (A) and simulated median (90% PI) NP viral load reduction in ADI-treated or placebo participants for delta (B) and omicron (C) variants Results. The QSP model provided an excellent fit to serum ADI concentrationtime data after estimation of a transit rate to account for IM absorption, plasma volume, and the ADI-neonatal Fc receptor dissociation rate constant. The linked viral dynamic model captured the NP swab viral load data after estimating differences in within-host replication factor (R0) and viral production rate (p) by variant. Maximal ADI-induced effect (Smax) on stimulating viral clearance (c) was fixed to 0.43 based upon prior modeling. ADI concentration in UA ELF resulting in 50% of Smax (SC50) was estimated to be 0.086 for delta and 1.05 mg/L for omicron. Figure B and C show model-based simulated median (90% PI) viral load reduction in ADI-treated or placebo pts for delta and omicron variants. Conclusion. This QSP model, in conjunction with information on new variants available early in outbreaks (IC50, infectivity (R0), viral production rate [each a model parameter]), allows for rapid dose identification in response to emerging variants.
ABSTRACT
Background. ADI is a fully human IgG1 monoclonal antibody engineered to have an extended half-life with high potency and broad neutralization against SARS-CoV-2 and other SARS-like coronaviruses with pandemic potential. Our objective was to develop a PPK model that describes the serum ADI concentration time profile following intravenous (IV) and intramuscular (IM) administration. Methods. The ADI PPK model was developed on PK data from a Phase 1 single ascending dose study (24 adults, IV and IM) and from Phase 2/3 COVID-19 prevention (EVADE;659 adults, IM) and treatment (STAMP;189 adults, IM) studies. 1,486 PK samples were included in the analysis. The impact of covariates (e.g. body weight, age, and baseline viral load) on ADI serum disposition were evaluated. Prediction-corrected visual predictive check (PC-VPC) plots were used to qualify the PPK model. Participant-specific ADI exposure estimates were generated using individual post hoc PK parameters. Results. The PPK model is comprised of 2 systemic compartments, zero-order infusion for IV administration and first-order absorption for IM administration and provided a robust fit to the data based on the PC-VPC plots and goodness-of-fit plots (data not shown). Body weight influenced clearance, inter compartmental clearance, and central and peripheral volume compartments. The relationship between body weight and clearance was not suggestive of the need for dose adjustment over the population weight range studied (38.6 to 178.7 kg). There was no apparent impact of baseline viral load or age on ADI clearance. The median [range] half-lives in days by study;Phase 1 (alpha1.71 [1.18-2.46];beta 125 [117-149]), Phase 2/3 prevention (alpha 1.86 [0.640-3.13];beta 136 [105-209]), and Phase 2/3 treatment (alpha 1.89 [0.631-3.01];beta 136 [108-206]). The population mean IM bioavailability estimate was 90.5%. The figure shows the PPK model median (90% confidence interval) concentration-time profile following a single 300 mg IM ADI dose by study. Conclusion. The PPK model provided a precise and unbiasedfit to the observed ADI concentration-time data and will be useful for future PK-pharmacodynamic analyses. Moreover, ADI demonstrated high IM bioavailability and a median terminal elimination half-life of 125 to 136 days.
ABSTRACT
Background. Adintrevimab is a fully human IgG1 monoclonal antibody engineered to have potent and broad neutralization against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and other SARS-like CoVs with pandemic potential. Adintrevimab is being assessed in two separate phase 2/3 clinical trials: the EVADE trial for prevention of COVID-19 in both post-exposure and preexposure settings and the STAMP trial for treatment of COVID-19. Here we report higher doses being evaluated in a healthy volunteer study given that emerging variants may have varying susceptibilities to adintrevimab. Previous results 300 mg IM, 600 mg IM, and 500 mg IV cohorts have been reported. Methods. This is an ongoing Phase 1, randomized, placebo (PBO)-controlled, single ascending-dose study of adintrevimab administered intramuscularly (IM) or intravenously (IV) to healthy adults aged 18-50 years with no current SARS-CoV-2 infection. Participants were randomized 8:2 in 3 high dose cohorts (N=10/cohort: n=8 adintrevimab, n=2 PBO): adintrevimab 1200 mg IM, 1200 mg IV, and 4500 mg IV. Safety, tolerability, and pharmacokinetics (PK) were assessed up to 21 days post dose. Results. Overall, 30 participants received adintrevimab (n=24) or PBO (n=6). Blinded safety data for all cohorts and PK for 1200 mg IV are reported. Through 21 days post dose all doses were well-tolerated, with no study drug-related adverse events (AEs), serious AEs, injection site reactions, or hypersensitivity reactions reported. The observed PK profile of the 1200 mg IV dose included Cmax of 423+/-105 mug/ml. Comparison of 500 mg and 1200 mg IV doses indicate dose proportionality of Cmax and exposure (AUC Day 21). Conclusion. A single dose of adintrevimab, up to 4500 mg, was well tolerated. These preliminary safety data and PK support potential use of higher doses of adintrevimab as needed to address emerging SARS-CoV-2 variants.
ABSTRACT
Introduction: After the outbreak of the Corona Virus Disease 2019 (COVID-19), there have been reports of impaired cardiac function in patients infected with this coronavirus. The tests are mostly based on myocardial injury markers and routine cardiac ultrasound examinations, which are mostly seen in critically ill patients. In this study, two-dimensional speckle tracking imaging (2DSTI) combined with Tei index and serum N-terminal pro-brain natriuretic peptide (NT-proBNP) were used to more sensitively diagnose cardiac function impairment in COVID-19 patients. Materials and methods: For some COVID-19 patients in our hospital, there were 68 cases of mild disease (including mild and common types) and 11 cases of severe disease (4 cases of severe death), and 10 healthy volunteers were included as the control group. On the basis of conventional echocardiography in all subjects, the left ventricular end-diastolic volume (LV-EDV), left ventricular end-systolic volume (LV-ESV), and left ventricular ejection fraction (LV-EF) were obtained by Simpson method, the left ventricular Tei index by tissue Doppler, and the left ventricular global peak longitudinal strain (GLPS), left ventricular global peak radial strain (GRPS), and left ventricular global peak circumferential strain (GCPS) by 2DSTI offline analysis. The COVID-19 patients were subjected to quantitative detection of serum NT-proBNP for statistical analysis. Results: Left ventricular GLPS, left ventricular GRPS, and left ventricular GCPS in COVID-19 patients were significantly lower than those in the control group (P<0.05): The left ventricular GLPS was more significant (P<0.01), and the severe group (including the death group) < the mild group < the control group. The left ventricular Tei index: The severe group (including the death group) of COVID-19 was significantly higher than the mild group and the control group (P<0.05), and there was no statistical significance between the mild group and the control group. NT-proBNP: The severe group of COVID-19 was significantly higher than the mild group (P<0.05). Although the LV-EF in the COVID-19 patients was significantly lower than that in the control group (P<0.05), except for 2 sever cases less than 50%, the rest were all ≥50%;although there was a significant difference in LV-ESV among multiple groups (P<0.05), but there was no significant difference for the pairwise comparison, and there was no significant difference in LV-EDV. Conclusion: 2DSTI can more sensitively detect latent cardiac function impairment in COVID-19 patients, and the left ventricular GLPS is the most sensitive. Tei index is an effective indicator to reflect the degree of cardiac function impairment. NT-proBNP has significant significance in predicting the severity of cardiac dysfunction. The combined application of the three can significantly increase the predictive performance of cardiac function impairment, provide a diagnostic basis for cardiac function impairment with preserved ejection fraction, and predict the degree of impairment. Our study demonstrated that the cardiac function of COVID-19 patients is impaired to varying degrees.
ABSTRACT
Artificial intelligence constantly puts forward new and sensitive requirements for the knowledge transmission of law discipline, focusing on the development and reform of legal education mode in the era of artificial intelligence, and realizing the perfect integration of artificial intelligence and law. The network teaching caused by novel coronavirus epidemic is changing the teaching methods of teachers and the learning methods of students, especially the influence on the education mode of law major. In the era of artificial intelligence, legal talents should pay close attention to the social reality and actively respond to the great challenges and opportunities given by artificial intelligence and other emerging technologies. With the popularization of network education, its own shortcomings have gradually emerged. This paper investigates the views of law students on the network teaching effect of law major in the era of artificial intelligence through questionnaire survey, and summarizes the survey data to find out the shortcomings of the network teaching model. Aiming at the existing problems, this paper puts forward some innovative strategies, such as establishing a normal crisis coping mechanism, selecting ways to stimulate students' interest in learning and implementing teaching, and establishing a diversified teaching system, in the hope of improving the online teaching of law major. © 2021 IEEE.
ABSTRACT
The study focuses on the emerging events of ozone formation in Chengdu-Chongqing areas, southwestern China in March-April 2019, investigating the sensitivities of ozone (O3) formation from anthropogenic precursor emissions in the Chengdu-Chongqing region. The study used the CAMx-DDM air quality model by changing scenarios of pollutants emissions due to the prevention and control measures of the "COVID-19" pandemic in 2020. The simulation results show the negative sensitivity of O3 to NOx and the positive sensitivity to VOCs in Chengdu-Chongqing area. The higher sensitivity areas include the main urban area of Chongqing, the west of the main urban area, the southern urban agglomeration of Sichuan and the western part of Chengdu plain, due to their intensive distributions of pollution emission sources. For example, the average sensitivity of O3 hourly concentration to NOx and VOCs from March to April in 2019 is -19.14 μg•m-3and 7.25 μg•m-3 respectively in the main urban area of Chongqing. The result shows opposite diurnal variations because of the emissions from local and surrounding areas. The simulation results showed that under the condition of 25% reduction of VOCs emissions in these regions, the monthly mean ozone maximum 8-hour concentration in March and April decreased by 2.62 μg•m-3 and 3.59 μg•m-3, respectively. Sensitivity simulation results show that in March 2020, NOx emissions in Sichuan Province and Chongqing decreased by 8.00% and 22.40%, VOCs decreased by 1. 00% and 7.92%. In April, NOx emissions increased by 5.00% and 9.50% year on year, VOCs in Sichuan province were flat year on year, and VOCs in Chongqing increased by 3.63%, which was very consistent with the actual emissions caused by the prevention and control of the "COVID-19" epidemic and the recovery of production in the same period. © 2021, Science Press. All right reserved.