Polyoxometalate Clusters Confined in Reduced Graphene Oxide Membranes for Effective Ion Sieving and Desalination.

Efficient 2D membranes play a critical role in water purification and desalination. However, most 2D membranes, such as graphene oxide (GO) membranes, tend to swell or disintegrate in liquid, making precise ionic sieving a tough challenge. Herein, the fabrication of the polyoxometalate clusters (PW12) intercalated reduced graphene oxide (rGO) membrane (rGO-PW12) is reported through a polyoxometalate-assisted in situ photoreduction strategy. The intercalated PW12 result in the interlayer spacing in the sub-nanometer scale and induce a nanoconfinement effect to repel the ions in various salt solutions. The permeation rate of rGO-PW12 membranes are about two orders of magnitude lower than those through the GO membrane. The confinement of nanochannels also generate the excellent non-swelling stability of rGO-PW12 membranes in aqueous solutions up to 400 h. Moreover, when applied in forward osmosis, the rGO-PW12 membranes with a thickness of 90 nm not only exhibit a high-water permeance of up to 0.11790 L m-2 h-1 bar-1 and high NaCl rejection (98.3%), but also reveal an ultrahigh water/salt selectivity of 4740. Such significantly improved ion-exclusion ability and high-water flux benefit from the multi-interactions and nanoconfinement effect between PW12 and rGO nanosheets, which afford a well-interlinked lamellar structure via hydrogen bonding and van der Waals interactions.

A novel T-shaped high-power waveguide phase shifter with continuous linear phase adjustment.

A novel T-shaped high-power waveguide phase shifter is investigated in this paper. The phase shifter consists of straight waveguides, four 90° H-bend waveguides, a stretching metal plate, and a metal spacer linked with the stretching metal plate. The entire structure for the phase shifter is symmetrical along both sides of the metal spacer. The phase-shifting principle for the phase shifter is to change the microwave transmission path by moving the stretching metal plate and then realize the linear phase adjustment. An optimal design approach of the phase shifter based on the boundary element method is described in detail. On this basis, a T-shaped waveguide phase shifter prototype with a center frequency of 9.3 GHz is designed. Simulation results show that the phase shifters can achieve 0°-360° linear phase adjustment by altering the distance of the stretched metal plate to 24 mm, and the efficiency of power transmission is more than 99.6%. In the meanwhile, experiments were conducted and the test results are in good agreement with simulation results. The return loss is more than 29 dB, and the insertion loss is less than 0.3 dB at 9.3 GHz in the entire phase-shifting range.

Recent Advances in Confining Polyoxometalates and the Applications.

Polyoxometalates (POMs) are widely used in catalysis, energy storage, biomedicine, and other research fields due to their unique acidity, photothermal, and redox features. However, the leaching and agglomeration problems of POMs greatly limit their practical applications. Confining POMs in a host material is an efficient tool to address the above-mentioned issues. POM@host materials have received extensive attention in recent years. They not only inherent characteristics of POMs and host, but also play a significant synergistic effect from each component. This review focuses on the recent advances in the development and applications of POM@host materials. Different types of host materials are elaborated in detail, including tubular, layered, and porous materials. Variations in the structures and properties of POMs and hosts before and after confinement are highlighted as well. In addition, an overview of applications for the representative POM@host materials in electrochemical, catalytic, and biological fields is provided. Finally, the challenges and future perspectives of POM@host composites are discussed.

Development and deployment of interpretable machine-learning model for predicting in-hospital mortality in elderly patients with acute kidney disease.

BACKGROUND: Acute kidney injury (AKI) is more likely to develop in the elderly admitted to the intensive care unit (ICU). Acute kidney disease (AKD) affects ∼45% of patients with AKI and increases short-term mortality. However, there are no studies on the prognosis of AKD in the elderly. METHODS: Data from 2666 elderly patients with AKD in the Medical Information Mart for Intensive Care IV were used for model development and 535 in the eICU Collaborative Research Database for external validation. Based on 5 machine learning algorithms, 33 noninvasive parameters were extracted as features for modeling. RESULTS: In-hospital mortality of AKD in the elderly was 29.6% and 31.8% in development and validation cohorts, respectively. The comprehensive best-performing algorithm was the support vector machine (SVM), and a simplified online application included only 10 features employing SVM (AUC: 0.810 and 0.776 in the training and external validation cohorts, respectively) was deployed. Model interpretation by SHapley Additive exPlanation (SHAP) values revealed that the difference (AKD day - ICU day) in sequential organ failure assessment (delta SOFA), Glasgow coma scale (GCS), delta GCS, delta peripheral oxygen saturation (SpO2), and SOFA were the top five features associated with prognosis. The optimal target was determined by SHAP values from partial dependence plots. CONCLUSIONS: A web-based tool was externally validated and deployed to predict the early prognosis of AKD in the elderly based on readily available noninvasive parameters, assisting clinicians in intervening with precision and purpose to save lives to the greatest extent.

Deep Active Learning Framework for Lymph Node Metastasis Prediction in Medical Support System.

Assessing the extent of cancer spread by histopathological analysis of sentinel axillary lymph nodes is an important part of breast cancer staging. With the maturity and prevalence of deep learning technology, building auxiliary medical systems can help to relieve the burden of pathologists and increase the diagnostic precision and accuracy during this process. However, such histopathological images have complex patterns that are difficult for ordinary people to understand and require professional medical practitioners to annotate. This increases the cost of constructing such medical systems. To reduce the cost of annotating and improve the performance of the model as much as possible, in other words, using as few labeled samples as possible to obtain a greater performance improvement, we propose a deep learning framework with a three-stage query strategy and novel model update strategy. The framework first trains an auto-encoder with all the samples to obtain a global representation in a low-dimensional space. In the query stage, the unlabeled samples are first selected according to uncertainty, and then, coreset-based methods are employed to reduce sample redundancy. Finally, distribution differences between labeled samples and unlabeled samples are evaluated and samples that can quickly eliminate the distribution differences are selected. This method achieves faster iterative efficiency than the uncertainty strategies, representative strategies, or hybrid strategies on the lymph node slice dataset and other commonly used datasets. It reaches the performance of training with all data, but only uses 50% of the labeled. During the model update process, we randomly freeze some weights and only train the task model on new labeled samples with a smaller learning rate. Compared with fine-tuning task model on new samples, large-scale performance degradation is avoided. Compared with the retraining strategy or the replay strategy, it reduces the training cost of updating the task model by 79.87% and 90.07%, respectively.

Auxiliary Medical Decision System for Prostate Cancer Based on Ensemble Method.

Prostate cancer (PCa) is one of the main diseases that endanger men's health worldwide. In developing countries, due to the large number of patients and the lack of medical resources, there is a big conflict between doctors and patients. To solve this problem, an auxiliary medical decision system for prostate cancer was constructed. The system used six relevant tumor markers as the input features and employed classical machine learning models (support vector machine and artificial neural network). Stacking method aimed at different ensemble models together was used for the reduction of overfitting. 1,933,535 patient information items had been collected from three first-class hospitals in the past five years to train the model. The result showed that the auxiliary medical system could make use of massive data. Its performance is continuously improved as the amount of data increases. Based on the system and collected data, statistics on the incidence of prostate cancer in the past five years were carried out. In the end, influence of diet habit and genetic inheritance for prostate cancer was analyzed. Results revealed the increasing prevalence of PCa and great negative impact caused by high-fat diet and genetic inheritance.