Predicting the correct dose in children: Role of computational Pediatric Physiological-based pharmacokinetics modeling tools.

The pharmacokinetics (PKs) and safety of medications in particular groups can be predicted using the physiologically-based pharmacokinetic (PBPK) model. Using the PBPK model may enable safe pediatric clinical trials and speed up the process of new drug research and development, especially for children, a population in which it is relatively difficult to conduct clinical trials. This review summarizes the role of pediatric PBPK (P-PBPK) modeling software in dose prediction over the past 6 years and briefly introduces the process of general P-PBPK modeling. We summarized the theories and applications of this software and discussed the application trends and future perspectives in the area. The modeling software's extensive use will undoubtedly make it easier to predict dose prediction for young patients.

Neurologic complications in children with seizures and respiratory illness: A comparison between SARS-CoV-2 and other respiratory viruses

OBJECTIVES/GOALS: To compare rates and types of neurological symptoms in children hospitalized with seizures and respiratory infections, including SARS-CoV-2, influenza, and endemic coronaviruses. METHODS/STUDY POPULATION: Retrospective cohort study of children between 0-21 years of age admitted to a single pediatric free-standing quaternary referral center from January 1, 2014 to June 1, 2021 for seizures who had positive respiratory infection PCR for SARS-CoV-2, other coronaviruses (Coronavirus NL63 and Coronavirus OC34), influenza (A and B), adenovirus, Mycoplasma pneumoniae, and parainfluenza 3 or 4 infections. Patient characteristics including age, race, sex, ethnicity, hospital length of stay, intensive care unit admission, intubation, chest x-ray, and MRI results were included. The primary outcomes were rates of neurological diagnoses and mortality. RESULTS/ANTICIPATED RESULTS: A total of 883 children were included: 68 SARS-CoV-2, 232 influenza, and 187 with other coronaviruses (OC), 214 adenovirus, 20 M. pneumoniae, 121 parainfluenza 3, and 41 parainfluenza 4. Mortality rates were 0% M pneumoniae to 4.9% in parainfluenza 4, with 2.9% in SARS-CoV-2. Encephalopathy was noted in 5-15.6% and strokes were seen in all infections except for coronavirus OC43 and M. pneumoniae, with 4.9% in parainfluenza 4 and 5.9% in SARS-CoV-2. The most common brain MRI abnormality was diffusion restriction. Differences between SARS-CoV-2 and OC were observed in stroke (5.9% vs. 0.5%, p-value=0.019), ICU admission (50% vs. 69%, p-value=0.008), and intubation (19.1% vs. 34.8%, p-value=0.021, respectively). However, the rates of neurological symptoms were similar between SARS-CoV-2 and influenza. DISCUSSION/SIGNIFICANCE: We found higher rates of stroke, but lower rates of ICU admission and intubation in SARS-CoV-2 versus OC. Strokes were observed in many infections. Rates of neurological symptoms were similar in SARS-CoV-2 versus influenza patients. Vigilance should be undertaken in treatment of children presenting with all respiratory illnesses.

Can dimensional reduction technology make better use of the information of uncertainty indices when predicting volatility of Chinese crude oil futures?

In this paper, we try to forecast the volatility of Chinese crude oil futures (COF) using multiple economic policy uncertainty indicators. MIDAS-RV model is combined with the principal component analysis (PCA), scaled PCA (SPCA) and partial least squares (PLS) techniques in this work, construct the newly MIDAS-RV-PCA, MIDAS-RV-PLS and MIDAS-RV-SPCA models, their prediction performance is compared with the common combination forecasting methods. The in-sample estimation analysis on MIDAS-RV-X models show the that it is necessary to consider multiple economic policy uncertainty indices when predicting the Chinese COF volatility and the in-sample analysis on dimensionality reduction model further demonstrate the rationality of using dimensionality reduction techniques. The out-of-sample evaluation results show that the MIDAS-RV-SPCA is a superior model when forecasting the short-term volatility of Chinese COF using multiple economic policy uncertainty indicators, especially during the periods of high volatility and COVID-19 pandemic. The results also indicates that the DMSPE(0.9) method in the combination forecasting method shows its superior forecasting ability in long-term volatility of Chinese COF, especially during the low volatility and pre-pandemic period.

Advancing COVID-19 Diagnosis with Privacy-Preserving Collaboration in Artificial Intelligence (preprint)

Artificial intelligence (AI) provides a promising substitution for streamlining COVID-19 diagnoses. However, concerns surrounding security and trustworthiness impede the collection of large-scale representative medical data, posing a considerable challenge for training a well-generalised model in clinical practices. To address this, we launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the AI model can be distributedly trained and independently executed at each host institution under a federated learning framework (FL) without data sharing. Here we show that our FL model outperformed all the local models by a large yield (test sensitivity /specificity in China: 0.973/0.951, in the UK: 0.730/0.942), achieving comparable performance with a panel of professional radiologists. We further evaluated the model on the hold-out (collected from another two hospitals leaving out the FL) and heterogeneous (acquired with contrast materials) data, provided visual explanations for decisions made by the model, and analysed the trade-offs between the model performance and the communication costs in the federated training process. Our study is based on 9,573 chest computed tomography scans (CTs) from 3,336 patients collected from 23 hospitals located in China and the UK. Collectively, our work advanced the prospects of utilising federated learning for privacy-preserving AI in digital health.

COVID-19 infection in patients with multiple myeloma: a German-Chinese experience from Wurzburg and Wuhan

In total, we identified five Caucasian patients from Wurzburg (Nos. 1-5) and three Asian patients from Wuhan (Nos. 6-8). The majority of the patients were male (n = 5, 63%), and the median age at COVID-19 diagnosis was 57 (range 39-83 years). The three patients from Wuhan were infected by COVID-19 in January or February 2020, while the Wurzburg patients were diagnosed in March or April 2020. Due to COVID-19 infection, anti-MM treatment was discontinued in all the patients. Notably, two patients (Nos. 3-4) in Wurzburg showed no COVID-19 symptoms, and the other three patients (Nos. 1, 2, and 5) exhibited only mild symptoms such as fever, cough, and nausea, which did not require an intensive care unit (ICU) admission. Interestingly, approximately 3 weeks after diagnosis, as the patient No. 6 was discharged and the swab was also negative for COVID-19, both COVID-19 IgM and IgG were tested negative in this patient. In four patients from Wurzburg, we also performed COVID-19 antibody test after recovery, and three of them (Nos. 1, 2, and 5) showed positive IgG, while one patient (No. 3) did not develop IgG or IgM against COVID-19. This finding suggested inadequate humoral immune response in MM patients, probably due to secondary immune deficiency caused by the treatments or the disease itself. This observation suggested that it might be a nosocomial infection in this patient. After recovery, two patients from Wurzburg received MM therapy, i.e., lenalidomide maintenance in one patient and DARA-VRCD (daratumumab, bortezomib, lenalidomide, cyclophosphamide, and dexamethasone) in another patient with NDMM.

Computational and Experimental Studies Reveal That Thymoquinone Blocks the Entry of Coronaviruses Into In Vitro Cells.

INTRODUCTION: Since December 2019, severe acute respiratory syndrome-related coronavirus-2 (SARS-CoV-2) has caused the coronavirus disease 2019 (COVID-19) pandemic in China and worldwide. New drugs for the treatment of COVID-19 are in urgent need. Considering the long development time for new drugs, the identification of promising inhibitors from FDA-approved drugs is an imperative and valuable strategy. Recent studies have shown that the S1 and S2 subunits of the spike protein of SARS-CoV-2 utilize human angiotensin-converting enzyme 2 (hACE2) as the receptor to infect human cells. METHODS: We combined molecular docking and surface plasmon resonance (SPR) to identify potential inhibitors for ACE2 from available commercial medicines. We also designed coronavirus pseudoparticles that contain the spike protein assembled onto green fluorescent protein or luciferase reporter gene-carrying vesicular stomatitis virus core particles. RESULTS: We found that thymoquinone, a phytochemical compound obtained from the plant Nigella sativa, is a potential drug candidate. SPR analysis confirmed the binding of thymoquinone to ACE2. We found that thymoquinone can inhibit SARS-CoV-2, SARS-CoV, and NL63 pseudoparticles infecting HEK293-ACE2 cells, with half-maximal inhibitory concentrations of 4.999, 7.598, and 6.019 µM, respectively. The SARS-CoV-2 pseudoparticle inhibition had half-maximal cytotoxic concentration of 35.100 µM and selection index = 7.020. CONCLUSION: Thymoquinone is a potential broad-spectrum inhibitor for the treatment of coronavirus infections.

A collaborative online AI engine for CT-based COVID-19 diagnosis (preprint)

Artificial intelligence can potentially provide a substantial role in streamlining chest computed tomography (CT) diagnosis of COVID-19 patients. However, several critical hurdles have impeded the development of robust AI model, which include deficiency, isolation, and heterogeneity of CT data generated from diverse institutions. These bring about lack of generalization of AI model and therefore prevent it from applications in clinical practices. To overcome this, we proposed a federated learning-based Unified CT-COVID AI Diagnostic Initiative (UCADI, http://www.ai-ct-covid.team/), a decentralized architecture where the AI model is distributed to and executed at each host institution with the data sources or client ends for training and inferencing without sharing individual patient data. Specifically, we firstly developed an initial AI CT model based on data collected from three Tongji hospitals in Wuhan. After model evaluation, we found that the initial model can identify COVID from Tongji CT test data at near radiologist-level (97.5% sensitivity) but performed worse when it was tested on COVID cases from Wuhan Union Hospital (72% sensitivity), indicating a lack of model generalization. Next, we used the publicly available UCADI framework to build a federated model which integrated COVID CT cases from the Tongji hospitals and Wuhan Union hospital (WU) without transferring the WU data. The federated model not only performed similarly on Tongji test data but improved the detection sensitivity (98%) on WU test cases. The UCADI framework will allow participants worldwide to use and contribute to the model, to deliver a real-world, globally built and validated clinic CT-COVID AI tool. This effort directly supports the United Nations Sustainable Development Goals' number 3, Good Health and Well-Being, and allows sharing and transferring of knowledge to fight this devastating disease around the world.

Predicting COVID-19 malignant progression with AI techniques (preprint)

Background: The coronavirus disease 2019 (COVID-19) has become a worldwide pandemic since mid-December 2019, which greatly challenge public medical systems. With limited medical resources, it is a natural strategy, while adopted, to access the severity of patients then determine the treatment priority. However, our work observes the fact that the condition of many mild outpatients quickly worsens in a short time, i.e. deteriorate into severe/critical cases. Hence, it has been crucial to early identify those cases and give timely treatment for optimizing treatment strategy and reducing mortality. This study aims to establish an AI model to predict mild patients with potential malignant progression. Methods: A total of 133 consecutively mild COVID-19 patients at admission who was hospitalized in Wuhan Pulmonary Hospital from January 3 to February 13, 2020, were selected in this retrospective IRB-approved study. All mild patients at admission were categorized into groups with or without malignant progression. The clinical and laboratory data at admission, the first CT, and the follow-up CT at severe/critical stage of the two groups were compared with Chi-square test, Fisher's exact test, and t test. Both traditional logistic regression and deep learning-based methods were used to build the prediction models. The area under ROC curve (AUC) was used to evaluate the models. Results: The deep learning-based method significantly outperformed logistic regression (AUC 0.954 vs. 0.893). The deep learning-based method achieved a prediction AUC of 0.938 by combining the clinical data and the CT data, significantly outperforming its counterpart trained with clinical data only by 0.141. By further considering the temporal information of the CT sequence, our model achieved the best AUC of 0.954. The proposed model can be effectively used for finding out the mild patients who are easy to deteriorate into severe/critical cases, so that such patients get timely treatments while alleviating the limitations of medical resources.