Abstract: Provided are an anti-CD47/anti-PD-L1 antibody, a pharmaceutical composition of the anti-CD47/anti-PD-L1 antibody, and applications thereof. The anti-CD47/anti-PD-L1 antibody provides antitumor activity, is free of significant red blood cell toxicity, and is applicable in preparing an antitumor medicament.

Abstract: A ventilation adjustment method and a high-frequency ventilation system, which ensure stable and accurate oxygen concentration control within an oxygen concentration setting range, are disclosed. The ventilation adjustment method includes: determining a first gas flow rate control value and a second gas flow rate control value according to a target output flow rate and an oxygen concentration setting value; determining whether the first gas flow rate control value falls into a first dead zone range and whether the second gas flow rate control value falls into a second dead zone range; if the first gas flow rate control value falls into the first dead zone range, maintaining a first gas flow rate controller turned on in an expiratory phase; and if the second gas flow rate control value falls into the second dead zone range, maintaining a second gas flow rate controller turned on in the expiratory phase.

Abstract: The disclosure relates to a method for making a photocatalytic structure, the method comprising: providing a carbon nanotube structure comprising a plurality of carbon nanotubes intersected with each other; a plurality of openings being defined by the plurality of carbon nanotubes; forming a photocatalytic active layer on the surface of the carbon nanotube structure; applying a metal layer pre-form on the surface of the photocatalytic active layer; and annealing the metal layer pre-form.

Abstract: A method for defending against an adversarial sample in image classification includes: denoising, by an adversarial denoising network, an input image to acquire a reconstructed image; acquiring, by a target classification model, a predicted category probability distribution of the reconstructed image; acquiring, by the target classification model, a predicted category probability distribution of the original input image; calculating an adversarial score of the input image, and determining the input image as an adversarial sample or a benign sample according to a threshold; outputting a category prediction result of the reconstructed image if the input image is determined as the adversarial sample; and outputting a category prediction result of the original input image if the input image is determined as the benign sample. A system for defending against an adversarial sample in image classification, and a data processing terminal are further provided.

Abstract: A MXene support for a noble metal that forms a catalyst having active sites comprising single metal-layer nanostructures. The catalyst is stable under conditions for methane conversion to higher hydrocarbons and provides reduced coke formation. The results show a supported metal catalyst using the MXene where Pt atoms form one or more layers of atoms on the surface of the Mo2TiC2Tx support after it is reduced at 750° C. The catalyst shows high selectivity for C2-hydrocarbons with reduced coke formation, which can cost effectively convert methane into other valuable products.

Abstract: Methods for gas-phase deoxygenation of a bio-oil are provided. In embodiments, such a method comprises exposing a bio-oil vapor comprising hydrocarbon compounds having oxygenated aromatic groups, to hydrogen gas in the presence of catalyst under conditions to induce deoxygenation of the oxygenated aromatic groups to provide a deoxygenated aromatic species, wherein the catalyst is a transition metal-incorporated mesoporous silicate having platinum deposited thereon and the transition metal is selected from Nb, W, Zr, and combinations thereof. The transition metal-incorporated mesoporous silicate catalysts are also provided.

Abstract: Provided is an anti-B7-H3 antibody or antigen-binding fragment thereof, a pharmaceutical composition containing the anti-B7-H3 antibody or antigen-binding fragment thereof, and an application thereof. The anti-B7-H3 antibody or antigen-binding fragment thereof of the present invention has high affinity and stability, and can be used for preparing antitumor drugs.

Abstract: A method and a device for acquiring biomechanical parameters based on an ultrasonic elastomyogram are provided, wherein the method includes: synchronously collecting a dynamic myodynamics image sequence and a dynamic elasticity image sequence of a single skeletal muscle under continuous stretching; acquiring a myodynamics parameter corresponding to each myodynamics image in the dynamic myodynamics image sequence and an elasticity modulus value corresponding to each elasticity image in the dynamic elasticity image sequence respectively; and generating an ultrasonic elastomyogram curve with the myodynamics parameter as the abscissa and the synchronized elasticity modulus value as the ordinate, and estimating a muscle biomechanical parameter based on the ultrasonic elastomyogram curve. Dynamically changing biomechanical parameters can be obtained, and the obtained muscle biomechanical parameters have relatively high accuracy.

Abstract: A system is described herein for controlling an inverter comprising an artificial neural network (ANN) and a space vector pulse-width modulation converter. The ANN comprises an input layer, plurality of hidden layers, an output layer, and a processor. The ANN receives a plurality of input signals and produces output signals that control a SVPWM converter. The ANN is trained to minimize a cost function and to implement optimal control based on ADP. Further, the ANN can be configured to handle rated current and PWM saturation constraints in providing volt/VAR control functions. Finally, the ANN uses averaged feedback signals instead of instantaneous feedback signals to improve volt/VAR control performance of parallel inverters. Performance evaluation shows that an ANN controller has a strong ability to maintain volt and VAR control at the grid edge and prevent fighting between inverters thereby allowing an inverter to work effectively in parallel with other ANN-controlled inverters.

Abstract: A trusted graph data node classification method includes: (1) inputting a topological graph and node features, and calculating a discrete Ricci curvature of the discrete topological graph; (2) preprocessing the curvature and the node features; (3) mapping the curvature, reconstructing original features, and performing a semi-supervised training on graph data containing adversarial examples; and (4) performing a classification on unlabeled nodes. The new method uses a discrete curvature to extract topological information, and uses a residual network to reconstruct node feature vectors without knowing the technical details of the adversarial examples, and without using a large number of adversarial examples for adversarial training. Hence, the system effectively defends against attacks from adversarial examples on the graph data, outperforms the existing mainstream models in terms of accuracy when used in data without adversarial examples, and is thus a trusted node classification system.

Abstract: The disclosure relates to a photocatalytic structure. The photocatalytic structure includes a substrate, a photocatalytic active layer, and a metal layer. The substrate, the photocatalytic active layer, and the metal layer are arranged in succession. The substrate includes a base and a patterned bulge layer on a surface of the base. The patterned bulge layer is a net-like structure comprising a plurality of strip-shaped bulges intersected with each other and a plurality of indents defined by the plurality of strip-shaped bulges. The plurality of strip-shaped bulges is an integrated structure. The photocatalytic active layer is on the surface of the patterned bulge layer. The metal layer includes a plurality of nanoparticles located on the surface of the photocatalytic active layer away from the substrate.

Abstract: The disclosure discloses an AI-based condition classification system for patients with novel coronavirus, which includes: a classification model acquisition module for training one or more binary classification models that classify patient conditions according to patient data, and obtain the most accurate binary classification model as the target model, and determine the interpretable features in the patient data; a preprocessing module is configured to extract the interpretable features in the patient data to be classified, and then perform preprocessing on the extracted features to fill in the missing values and replace the abnormal values in the data, so as to obtain the features to be classified after the preprocessing is completed; a condition classification module is configured to use the features to be classified as the input for the target model, and the target model is used to complete condition classification for the patients to be classified.

Abstract: A thin film transistor includes a gate electrode, a insulating medium layer and at least one Schottky diode unit. The at least one Schottky diode unit is located on a surface of the insulating medium layer. The at least one Schottky diode unit includes a first electrode, a semiconductor structure and a second electrode. The semiconductor structure comprising a first end and a second end. The first end is laid on the first electrode, the second end is located on the surface of the insulating medium layer. The semiconducting structure includes a nano-scale semiconductor structure. The second electrode is located on the second end.

Abstract: A rotor for a motor is provided. The rotor has a rotor core, first permanent magnets mounted in first grooves of the rotor core, and second permanent magnets mounted in second grooves of the rotor core. The rotor core is provided with groove groups distributed in a peripheral direction of the rotor core. Each groove group has an air groove, a first groove and a second groove. The ends of the air groove, the first groove and the second groove, which are close to a center point of the rotor core, are close to one another. The end of the air groove, the first groove and the second groove, which are away from the center point of the rotor core, are away from one another. Multiple magnetic isolation structures are provided in the rotor core.

Abstract: An ultrasound deep brain stimulation method and system, the ultrasound deep brain stimulation method comprises: medically imaging a head of an animal or a human being, to generate image data; creating a head 3D digital model according to the image data; creating a 3D digital model of an ultrasound transducer array according to structure, density and acoustic parameters information of the ultrasound transducer array; generating a first ultrasound transmitting sequence according to the head 3D digital model, the 3D digital model of the ultrasound transducer array, structure, density and acoustic parameters of the skull and brain tissues, and structure, density and acoustic parameters of the ultrasound transducer array; and controlling the ultrasound transducer array to transmit ultrasound waves in accordance with the first ultrasound transmitting sequence, to implement ultrasound deep brain stimulation to the brain nucleus to be stimulated.

Abstract: A method and device for evaluating muscle tension, the method comprising: obtaining the elastic modulus values of the measured skeletal muscle under different joint angles and each of the joint angles (S11); determining the changing trend of the elastic modulus value of the measured skeletal muscle according to the elastic modulus values of the measured skeletal muscle under different joint angles and each of the joint angles (S12); and evaluating the muscle tension level of the measured skeletal muscle according to the changing trend of the elastic modulus value of the measured skeletal muscle (S13). The method can be used to evaluate the muscle tension level.

Abstract: Described are systems and methods that solve localization problems using Machine Learning models to compute country vectors for each linked content item and present content items in response to requests based on the country vectors. For example, a request from a user in Country A may be processed to determine candidate content items responsive to the request and to determine Country A as the country corresponding to the request. The candidate content items may then be processed to determine, for each candidate content item, a country vector corresponding to Country A as indicative of the relevance of the content item to Country A. Content items that are more likely than not to be relevant to the country of the request (e.g., Country A), as indicated by the respective country vector, may be considered as responsive and all other candidate content items discarded.