A comprehensive benchmark for COVID-19 predictive modeling using electronic health records in intensive care.

The COVID-19 pandemic highlighted the need for predictive deep-learning models in health care. However, practical prediction task design, fair comparison, and model selection for clinical applications remain a challenge. To address this, we introduce and evaluate two new prediction tasks-outcome-specific length-of-stay and early-mortality prediction for COVID-19 patients in intensive care-which better reflect clinical realities. We developed evaluation metrics, model adaptation designs, and open-source data preprocessing pipelines for these tasks while also evaluating 18 predictive models, including clinical scoring methods and traditional machine-learning, basic deep-learning, and advanced deep-learning models, tailored for electronic health record (EHR) data. Benchmarking results from two real-world COVID-19 EHR datasets are provided, and all results and trained models have been released on an online platform for use by clinicians and researchers. Our efforts contribute to the advancement of deep-learning and machine-learning research in pandemic predictive modeling.

Light intensity differentially mediates the life cycle of lepidopteran leaf feeders and stem borers.

BACKGROUND: Leaf feeders, such as Spodoptera frugiperda and Spodoptera litura, and stem borers Ostrinia furnacalis and Chilo suppressalis, occupy two different niches and are well adapted to their particular environments. Borer larvae burrow and inhabit the interior of stems, which are relatively dark. By contrast, the larvae of leaf feeders are exposed to sunlight during feeding. We therefore designed series of experiments to evaluate the effect of light intensity (0, 2000, and 10 000 lx) on these pests with different feeding modes. RESULTS: The development of all four pests was significantly delayed at 0 lx. Importantly, light intensity affected the development of both male and female larvae of borers, but only significantly affected male larvae of leaf feeders. Furthermore, the proportion of female offspring of leaf feeders increased with increasing light intensity (S. frugiperda: 33.89%, 42.26%, 57.41%; S. litura: 38.90%, 51.75%, 65.08%), but no significant differences were found in stem borers. This research also revealed that the survival rate of female leaf feeders did not vary across light intensities, but that of males decreased with increasing light intensity (S. frugiperda: 97.78%, 85.86%, 61.21%; S. litura: 95.83%, 73.54%, 58.99%). CONCLUSION: These results improve our understanding of how light intensity affects sex differences in important lepidopteran pests occupying different feeding niches and their ecological interactions with abiotic factors in agroecosystems. © 2024 Society of Chemical Industry.

A case of cardiac tamponade occurred 1 day after transcatheter atrial septal defect closure.

Transesophageal echocardiography (TEE) shows pericardial effusion and a gap between the left atrium and the aortic sinus by atrial septal defect occluder.

Association of carotid wall shear stress measured by vector flow mapping technique with ba-PWV: a pilot study.

Objective: Arterial stiffness is an important tissue biomarker of the progression of atherosclerotic diseases. Brachial-ankle pulse wave velocity (ba-PWV) is a gold standard of arterial stiffness measurement widely used in Asia. Changes in vascular wall shear stress (WSS) lead to artery wall remodeling, which could give rise to an increase in arterial stiffness. The study aimed to explore the association between ba-PWV and common carotid artery (CCA) WSS measured by a newly invented vascular vector flow mapping (VFM) technique. Methods: We included 94 subjects free of apparent cardiovascular disease (CVD) and divided them into a subclinical atherosclerosis (SA) group (N = 47) and non subclinical atherosclerosis (NSA) group (N = 47). CCA WSS was measured using the VFM technique. Bivariate correlations between CCA WSS and other factors were assessed with Pearson's, Spearman's, or Kendall's coefficient of correlation, as appropriate. Partial correlation analysis was conducted to examine the influence of age and sex. Multiple linear stepwise regression was used for the analysis of independent determinants of CCA WSS. Receiver operating characteristic (ROC) analysis was performed to find the association between CCA WSS and 10-year CVD risk. Results: The overall subjects had a mean age of 47.9 ± 11.2 years, and males accounted for 52.1%. Average systolic CCA WSS was significantly correlated with ba-PWV (r = -0.618, p < 0.001) in the SA group. Multiple linear stepwise regression analysis confirmed that ba-PWV was an independent determinant of average systolic CCA WSS (ß = -0.361, p = 0.003). The area under the curve (AUC) of average systolic CCA WSS for 10-year CVD risk ≥10% was 0.848 (p < 0.001) in the SA group. Conclusions: Average systolic CCA WSS was significantly correlated with ba-PWV and was associated with 10-year CVD risk ≥10% in the SA group. Therefore, CCA WSS measured by the VFM technique could be used for monitoring and screening subjects with potential CVD risks.

Mortality prediction with adaptive feature importance recalibration for peritoneal dialysis patients.

The study aims to develop AICare, an interpretable mortality prediction model, using electronic medical records (EMR) from follow-up visits for end-stage renal disease (ESRD) patients. AICare includes a multichannel feature extraction module and an adaptive feature importance recalibration module. It integrates dynamic records and static features to perform personalized health context representation learning. The dataset encompasses 13,091 visits and demographic data of 656 peritoneal dialysis (PD) patients spanning 12 years. An additional public dataset of 4,789 visits from 1,363 hemodialysis (HD) patients is also considered. AICare outperforms traditional deep learning models in mortality prediction while retaining interpretability. It uncovers mortality-feature relationships and variations in feature importance and provides reference values. An AI-doctor interaction system is developed for visualizing patients' health trajectories and risk indicators.

Multi-faceted analysis and prediction for the outbreak of pediatric respiratory syncytial virus.

OBJECTIVES: Respiratory syncytial virus (RSV) is a significant cause of pediatric hospitalizations. This article aims to utilize multisource data and leverage the tensor methods to uncover distinct RSV geographic clusters and develop an accurate RSV prediction model for future seasons. MATERIALS AND METHODS: This study utilizes 5-year RSV data from sources, including medical claims, CDC surveillance data, and Google search trends. We conduct spatiotemporal tensor analysis and prediction for pediatric RSV in the United States by designing (i) a nonnegative tensor factorization model for pediatric RSV diseases and location clustering; (ii) and a recurrent neural network tensor regression model for county-level trend prediction using the disease and location features. RESULTS: We identify a clustering hierarchy of pediatric diseases: Three common geographic clusters of RSV outbreaks were identified from independent sources, showing an annual RSV trend shifting across different US regions, from the South and Southeast regions to the Central and Northeast regions and then to the West and Northwest regions, while precipitation and temperature were found as correlative factors with the coefficient of determination R2≈0.5, respectively. Our regression model accurately predicted the 2022-2023 RSV season at the county level, achieving R2≈0.3 mean absolute error MAE < 0.4 and a Pearson correlation greater than 0.75, which significantly outperforms the baselines with P-values <.05. CONCLUSION: Our proposed framework provides a thorough analysis of RSV disease in the United States, which enables healthcare providers to better prepare for potential outbreaks, anticipate increased demand for services and supplies, and save more lives with timely interventions.

Activation of peroxymonosulphate using a highly efficient and stable ZnFe2O4 catalyst for tetracycline degradation.

Tetracycline (TC) is a widely used antibiotic that adversely affects ecosystems and, therefore, must be removed from the environment. Owing to their strong ability to oxidise pollutants, including antibiotics, and selectivity for these pollutants, an improved oxidation method based on sulphate radicals (SO4·-) has gained considerable interest. In this study, a novel technique for removing TC was developed by activating peroxymonosulphate (PMS) using a ZnFe2O4 catalyst. Using the co-precipitation method, a ZnFe2O4 catalyst was prepared by doping zinc into iron-based materials, which increased the redox cycle, while PMS was active and facilitated the production of free radicals. According to electron paramagnetic resonance spectroscopy results, a ZnFe2O4 catalyst may activate PMS and generate SO4·-, HO·, O2·-, and 1O2 to eliminate TC. This research offers a new method for creating highly effective heterogeneous catalysts that can activate PMS and destroy antibiotics. The study proposes the following degradation pathways: hydroxylation and ring-opening of TC based on the products identified using ultra-performance liquid chromatography-mass spectrometry. These results illustrated that the prepared ZnFe2O4 catalyst effectively removed TC and exhibited excellent catalytic performance.

Application and evaluation of surgical tool and tool tip recognition based on Convolutional Neural Network in multiple endoscopic surgical scenarios.

BACKGROUND: In recent years, computer-assisted intervention and robot-assisted surgery are receiving increasing attention. The need for real-time identification and tracking of surgical tools and tool tips is constantly demanding. A series of researches focusing on surgical tool tracking and identification have been performed. However, the size of dataset, the sensitivity/precision, and the response time of these studies were limited. In this work, we developed and utilized an automated method based on Convolutional Neural Network (CNN) and You Only Look Once (YOLO) v3 algorithm to locate and identify surgical tools and tool tips covering five different surgical scenarios. MATERIALS AND METHODS: An algorithm of object detection was applied to identify and locate the surgical tools and tool tips. DarkNet-19 was used as Backbone Network and YOLOv3 was modified and applied for the detection. We included a series of 181 endoscopy videos covering 5 different surgical scenarios: pancreatic surgery, thyroid surgery, colon surgery, gastric surgery, and external scenes. A total amount of 25,333 images containing 94,463 targets were collected. Training and test sets were divided in a proportion of 2.5:1. The data sets were openly stored at the Kaggle database. RESULTS: Under an Intersection over Union threshold of 0.5, the overall sensitivity and precision rate of the model were 93.02% and 89.61% for tool recognition and 87.05% and 83.57% for tool tip recognition, respectively. The model demonstrated the highest tool and tool tip recognition sensitivity and precision rate under external scenes. Among the four different internal surgical scenes, the network had better performances in pancreatic and colon surgeries and poorer performances in gastric and thyroid surgeries. CONCLUSION: We developed a surgical tool and tool tip recognition model based on CNN and YOLOv3. Validation of our model demonstrated satisfactory precision, accuracy, and robustness across different surgical scenes.

Discovery of dual S-RBD/NRP1-targeting peptides: structure-based virtual screening, synthesis, biological evaluation, and molecular dynamics simulation studies.

Both receptor-binding domain in spike protein (S-RBD) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and human neuropilin-1 (NRP1) are important in the virus entry, and their concomitant inhibition may become a potential strategy against the SARS-CoV-2 infection. Herein, five novel dual S-RBD/NRP1-targeting peptides with nanomolar binding affinities were identified by structure-based virtual screening. Particularly, RN-4 was found to be the most promising peptide targeting S-RBD (Kd = 7.4 ± 0.5 nM) and NRP1-BD (the b1 domain of NRP1) (Kd = 16.1 ± 1.1 nM) proteins. Further evidence in the pseudovirus infection assay showed that RN-4 can significantly inhibit the SARS-CoV-2 pseudovirus entry into 293 T cells (EC50 = 0.39 ± 0.09 µM) without detectable side effects. These results suggest that RN-4, a novel dual S-RBD/NRP1-targeting agent, holds potential as an effective therapeutic to combat the SARS-CoV-2 infection.

Evidence-driven spatiotemporal COVID-19 hospitalization prediction with Ising dynamics.

In this work, we aim to accurately predict the number of hospitalizations during the COVID-19 pandemic by developing a spatiotemporal prediction model. We propose HOIST, an Ising dynamics-based deep learning model for spatiotemporal COVID-19 hospitalization prediction. By drawing the analogy between locations and lattice sites in statistical mechanics, we use the Ising dynamics to guide the model to extract and utilize spatial relationships across locations and model the complex influence of granular information from real-world clinical evidence. By leveraging rich linked databases, including insurance claims, census information, and hospital resource usage data across the U.S., we evaluate the HOIST model on the large-scale spatiotemporal COVID-19 hospitalization prediction task for 2299 counties in the U.S. In the 4-week hospitalization prediction task, HOIST achieves 368.7 mean absolute error, 0.6 [Formula: see text] and 0.89 concordance correlation coefficient score on average. Our detailed number needed to treat (NNT) and cost analysis suggest that future COVID-19 vaccination efforts may be most impactful in rural areas. This model may serve as a resource for future county and state-level vaccination efforts.

Study on mechanism of effect of flowing water and transferring heat on rock mass temperature in curved fracture.

Domestically and internationally, the effect of fracture flowing water and transferring heat on the temperature field of surrounding rock in high-level radioactive waste repositories is a popular research area. Compared with straight fracture flowing water and transferring heat, there are few relevant literatures about the heat transfer of curved fracture water flow. Based on the conceptive model of flowing water and transferring heat in curved fractured rock mass, the influence of flowing water and transferring heat in "I", "L", , and shaped fractures on the temperature field of rock mass is calculated by using discrete element program. The findings indicate that: When the model goes into a stable state under four working conditions, the rock on the x = 0-2 m mostly forms a heat transfer path from left to right; the x = 2-4 m primarily forms a heat transfer path from bottom to top, and the temperature gradient reveals that the isotherm of 40-45 °C is highly similar to the shape of four different fractures, indicating that flowing water and transferring heat in the fracture configuration dominate the temperature field of the right side rock mass. The direction of the flowing water and transferring heat of the fracture exerts a dominant effect on the temperature of the rock mass than the length.

GREM1 is a novel serum diagnostic marker and potential therapeutic target for pancreatic ductal adenocarcinoma.

Objective: Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant neoplasm with rising incidence worldwide. Gremlin 1 (GREM1), a regulator of bone morphogenetic protein (BMP) signaling, fine-tunes extensive biological processes, including organ morphology, cellular metabolism, and multiple pathological developments. The roles of GREM1 in PDAC remain unknown. Methods: Varieties of public databases and online software were employed to analyze the expressions at transcription and protein levels of GREM1 in multiple malignant neoplasms including PDAC, and in addition, its potential pro-tumoral functions in PDAC were further evaluated. A total of 340 serum samples of pancreatic disease, including PDAC, low-grade malignant pancreatic neoplasm, benign pancreatic neoplasm, pancreatitis, and 132 healthy controls, were collected to detect GREM1. The roles of serum GREM1 in the diagnosis and prediction of survival of PDAC after radical resection were also analyzed. Results: Bioinformatics analyses revealed that GREM1 was overexpressed in PDAC and predicted a poorer survival in PDAC. A higher protein level of GREM1 in PDAC correlated with stroma formation and immunosuppression by recruiting varieties of immunosuppressive cells, including T regulatory cells (Tregs), M2 macrophages, myeloid-derived suppressor cells (MDSCs), and exhaustion T cells into the tumor microenvironment. A higher level of serum GREM1 was observed in PDAC patients, compared to healthy control (p < 0.001). Serum GREM1 had a good diagnostic value (area under the curve (AUC) = 0.718, p < 0.001), and its combination with carbohydrate antigen 199 (CA199) achieved a better diagnostic efficacy (AUC = 0.914, p < 0.001), compared to CA199 alone. The cutoff value was calculated by receiver operating characteristic (ROC) analysis, and PDAC patients were divided into two groups of low and high GREM1. Logistic analyses showed serum GREM1 positively correlated with tumor size (hazard ratio (HR) = 7.097, p = 0.032) and histopathological grades (HR = 2.898, p = 0.014). High-level serum GREM1 (1,117.8 pg/ml) showed a shorter postoperative survival (p = 0.0394). Conclusion: Higher intra-tumoral expression of GREM1 in PDAC contributes to tumor stroma and immunosuppressive tumor microenvironment, presenting its therapeutic potential. High-level serum GREM1 predicts poorer survival after resection. A combination of serum CA199 and GREM1 shows a stronger diagnostic efficacy in PDAC.

MedML: Fusing medical knowledge and machine learning models for early pediatric COVID-19 hospitalization and severity prediction.

The COVID-19 pandemic has caused devastating economic and social disruption. This has led to a nationwide call for models to predict hospitalization and severe illness in patients with COVID-19 to inform the distribution of limited healthcare resources. To address this challenge, we propose a machine learning model, MedML, to conduct the hospitalization and severity prediction for the pediatric population using electronic health records. MedML extracts the most predictive features based on medical knowledge and propensity scores from over 6 million medical concepts and incorporates the inter-feature relationships in medical knowledge graphs via graph neural networks. We evaluate MedML on the National Cohort Collaborative (N3C) dataset. MedML achieves up to a 7% higher AUROC and 14% higher AUPRC compared to the best baseline machine learning models. MedML is a new machine learnig framework to incorporate clinical domain knowledge and is more predictive and explainable than current data-driven methods.

Automatic bleeding detection in laparoscopic surgery based on a faster region-based convolutional neural network.

Background: Laparoscopic surgery has been in great demand over the past decades; it has also brought several obstacles, such as increasing difficulty in maintaining hemostasis, changes in surgical approach, and reduced field of vision. Locating the bleeding point can help surgeons to control bleeding quickly, however, to date, there have been no tools designed for automatic bleeding tracking in laparoscopic operations. Herein, we have proposed a spatiotemporal hybrid model based on a faster region-based convolutional neural network (RCNN) for bleeding point detection in laparoscopic surgery videos. Methods: Laparoscopic videos performed at our hospital were retrieved and images containing bleeding events were extracted. Spatiotemporal features were extracted by using red-green-blue (RGB) frames and optical flow maps and a spatiotemporal hybrid model was developed based on the faster RCNN. The proposed model contributed to (I) providing real-time bleeding point detection which directly assist surgeons, (II) showing the blood's optical flow which improved bleeding point detection, and (III) detecting both arterial and venous bleeding. Results: In this study, 12 different bleeding videos were included for deep learning model training. Compared with models containing a single RGB or a single optical flow map, our model combining RGB and optical flow achieved great detection results (precision rate of 0.8373, recall rate of 0.8034, and average precision of 0.6818). Conclusions: Our approach performs well in bleeding point location and recognition, indicating its potential value in helping to maintain and re-establish hemostasis during operations.

Pathological complete response in a patient with locally advanced pancreatic adenocarcinoma treated with neoadjuvant gemcitabine and S-1: a case report and literature review.

Background: Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal cancers worldwide. Radical resection is currently the only potential curative treatment. However, over 80% of patients present with unresectable tumor at the time of diagnosis. It is recommended that patients with unresectable pancreatic cancers be offered neoadjuvant treatment. A combination of gemcitabine and S-1 (GS-1) has been reported to be an effective regimen for unresectable pancreatic cancers, however, there have been no reports of pathological complete response up until now. Case Description: Herein, we present a 67-year-old male who presented with a 4-month history of upper abdominal and back pain, as well as unintentional weight loss. Abdominal computed tomography (CT) confirmed a hypovascular mass in the pancreas neck consistent with unresectable pancreatic cancer. Positron emission tomography (PET)/CT also revealed a high fludeoxyglucose (FDG)-avid lesion in the pancreas neck without evidence of distant metastasis. Pancreatic adenocarcinoma was confirmed with ultrasound-guided fine-needle aspiration cytology. The patient was recommended to undergo treatment with gemcitabine and S-1. After 5 cycles of neoadjuvant chemotherapy, CT and PET/CT both revealed the disappearance of the lesion and a pancreaticoduodenectomy was offered as a potentially curative treatment. Histological assessment revealed no evidence of residual adenocarcinoma [ypT0N0 (0/38)]. The tumor marker cancer antigen (CA)125 increased one month after the surgery, resulting in two additional cycles of GS-1. This patient remained disease-free for 21 months after surgery. Conclusions: This report is the first to present a case of a pathological complete response in a patient with locally advanced pancreatic cancer following GS-1 treatment, suggesting radical resection after GS-1 chemotherapy might be a potential curative treatment strategy for unresectable PDAC.

Near 100% external quantum efficiency 1550-nm broad spectrum photodetector.

We report InGaAs/InP based p-i-n photodiodes with an external quantum efficiency (EQE) above 98% from 1510 nm to 1575 nm. For surface normal photodiodes with a diameter of 80 µm, the measured 3-dB bandwidth is 3 GHz. The saturation current is 30.5 mA, with an RF output power of 9.3 dBm at a bias of -17 V at 3 GHz.

Genetic and Clinicopathologic Characteristics of Papillary Thyroid Carcinoma in the Chinese Population: High BRAF Mutation Allele Frequency, Multiple Driver Gene Mutations, and RET Fusion May Indicate More Advanced TN Stage.

PURPOSE: To describe the genetic landscape and clinical characteristics of Chinese patients diagnosed with papillary thyroid cancer (PTC) and to determine which high-risk genetic characteristics suggest a likelihood of lymph node metastasis (LNM) and lateral lymph node metastasis (LLNM). PATIENTS AND METHODS: Data from previously untreated patients with PTC collected between May 2018 and December 2020 from 14 hospitals in China were analyzed retrospectively. High-risk pathologic characteristics were defined as T3/T4, N(+), and N1b(+) stages. All patients were tested for 57 genes by second-generation sequencing. The t-test, chi-square test, and Fisher's exact test were performed for statistical analysis. RESULTS: Overall, 395 patients were enrolled in this study. The prevalence of BRAF mutation was 78.53%. BRAF mutant allele frequency (MAF) >16.93% was associated with a significantly higher risk of LNM, LLNM, and T3 + T4 stage compared with a low-risk group, defined by a MAF <2.54% (odd ratios [ORs] for each risk=3.38, 3.46, and 8.54, respectively), and an intermediate-risk group, defined by a MAF of 2.54% to 16.93% (ORs=2.04, 2.07, and 4.07, respectively). The population with RET fusion had higher T, N, and N1b stages (ORs for each stage=10.40, 7.60, and 8.77, respectively) compared with a RET-negative population. Similar conclusions about T, N, and N1b stages were observed in relation to multiple driver gene mutations (ORs for each stage=7.48, 2.80, and 7.04, respectively) compared with population without multiple driver mutations. These genetic characteristics may be suggestive of high clinical risk. However, regardless of genetic profiles, patients younger than age 45 years had greater rates of LNM and LLNM. CONCLUSION: The main driver gene in this study, BRAF, differs significantly between the United States (79% vs 51%) and other countries. The Chinese population in this study that experienced more aggressive tumor biology had a BRAF MAF greater than 16.93%, exhibited RET fusion events, and had multiple driver gene mutations; thus, these traits may be considered high-risk genetic characteristics in PTC that could warrant aggressive treatment in such population.

A new mathematical modeling approach for thermal exploration efficiency under different geothermal well layout conditions.

The water temperature at the outlet of the production well is an important index for evaluating efficient geothermal exploration. The arrangement mode of injection wells and production wells directly affects the temperature distribution of the production wells. However, there is little information about the effect of different injection and production wells on the temperature field of production wells and rock mass, so it is critical to solve this problem. To study the influence mechanism of geothermal well arrangement mode on thermal exploration efficiency, the conceptual model of four geothermal wells is constructed by using discrete element software, and the influence law of different arrangement modes of four geothermal wells on rock mass temperature distribution is calculated and analyzed. The results indicated that the maximum water temperature at the outlet of the production well was 84.0 °C due to the thermal superposition effect of the rock mass between the adjacent injection wells and between the adjacent production wells. Inversely, the minimum water temperature at the outlet of the production well was 50.4 °C, which was determined by the convection heat transfer between the water flow and the rock between the interval injection wells and the interval production wells. When the position of the model injection well and production well was adjusted, the isothermal number line of rock mass was almost the same in value, but the direction of water flow and heat transfer was opposite. The study presented a novel mathematical modeling approach for calculating thermal exploration efficiency under various geothermal well layout conditions.