Abstract: This application provides a video labeling method performed by a server, and the method includes: receiving a video extraction instruction transmitted by a terminal, and obtaining a to-be-extracted video according to the video extraction instruction; extracting event information from video frames of the to-be-extracted video; forming at least one event information flow by using the event information; capturing, based on the at least one event information flow, at least one first clip that meets a plot trigger condition from the to-be-extracted video and obtaining a plot labeling tag of the at least one first clip; and transmitting the at least one first clip and the corresponding plot labeling tag to the terminal, wherein the terminal displays the at least one first clip and the corresponding plot labeling tag in a preset display region of a display interface in which the to-be-extracted video is displayed.

Abstract: A model training method applied to the field of artificial intelligence is disclosed. The method includes: sending a first submodel to a first device, where the first submodel is obtained by compressing a to-be-trained model; receiving a first gradient sent by the first device, where the first gradient is obtained when the first device trains the first submodel; and performing model training on the to-be-trained model based on at least the first gradient, to obtain an updated to-be-trained model. In the method, a server compresses the to-be-trained model and delivers the to-be-trained model to a terminal device, so that the terminal device does not need to train a large model with a same scale as that of the server.

Abstract: Described herein are systems, methods, and instrumentalities associated with segmenting and/or determining the shape of an anatomical structure. An artificial neural network (ANN) is used to perform these tasks based on a statistical shape model of the anatomical structure. The ANN is trained by evaluating and backpropagating multiple losses associated with shape estimation and segmentation mask generation. The model obtained using these techniques may be used for different clinical purposes including, for example, motion estimation and motion tracking.

Abstract: A microfluidic apparatus, and its drive circuit and drive method are provided in the present disclosure. The drive circuit includes at least one switch unit, where the switch unit includes a first signal input terminal, a signal output terminal, first and second control signal terminals, a signal terminal, a first module, a second module, and a third module. The first module is electrically connected to the first signal input terminal, the signal output terminal and the second module; and the second module is electrically connected to the first control signal terminal. The first control signal terminal is configured to control the first module and the second module to be in conduction or disconnection, and the second control signal terminal is configured to control the third module to in conduction or disconnection, thereby controlling the signal output terminal to output a first or second signal.

Abstract: Described herein are neural network-based systems, methods and instrumentalities associated with cardiac assessment. An apparatus as described herein may obtain electrocardiographic imaging (ECGI) information associated with a human heart and magnetic resonance imaging (MRI) information associated with the human heart, and integrate the ECGI and MRI information using a machine-learned model. Using the integrated ECGI and MRI information, the apparatus may predict target ablation sites, estimate electrophysiology (EP) measurements, and/or simulate the electrical system of the human heart.

Abstract: The present application relates to a posture display method, a device and a system for a guiding channel, and readable storage medium. The posture display method of a guiding channel includes determining an original position relation between a three-dimensional affected limb image and a virtual guiding channel displayed in a monitor according to an original position relation between the guiding channel and an affected limb and obtaining posture change data of the guiding channel. The method further includes adjusting the posture of the virtual guiding channel based on the posture change data, so as to the relative position relation between the virtual guiding channel and the three-dimensional affected limb image matches with the relative position relation between the guiding channel and the affected limb.

Abstract: A nuclear magnetic resonance (NMR) system-based substance measurement method, including: acquiring several echo signals of an NMR pulse sequence varying in echo spacing from a substance to be measured followed by processing to obtain several signals varying in transverse relaxation and diffusion attenuation; and fitting, in combination with the prior knowledge, the signals to obtain the diffusion coefficient, transverse relaxation time or/and content weight of individual components of the substance to be measured. This application further provides a substance measurement system including a console, a magnet module, and an NMR system.

Abstract: In Multiplex MRI image reconstruction, a hardware processor acquires sub-sampled Multiplex MRI data and reconstructs parametric images from the sub-sampled Multiplex MRI data. A machine learning model or deep learning model uses the subsampled Multiplex MRI data as the input and parametric maps calculated from the fully sampled data, or reconstructed fully sample data, as the ground truth. The model learns to reconstruct the parametric maps directly from the subsampled Multiplex MRI data.

Abstract: A neural network system implements a model for generating an output image based on a received input image. The model is learned through a training process during which parameters associated with the model are adjusted so as to maximize a difference between a first image predicted using first parameter values of the model and a second image predicted using second parameter values of the model, and to minimize a difference between the second image and a ground truth image. During a first iteration of the training process the first image is predicted and during a second iteration the second image is predicted. The first parameter values are obtained during the first iteration by minimizing a difference between the first image and the ground truth image, and the second parameter values are obtained during the second iteration by maximizing the difference between the first image and the second image.

Abstract: A method for detecting a storage life of a pre-crosslinked material is provided. Tableting is performed on an unaged pre-crosslinked material to obtain crosslinked polyethylene. A crosslinking degree and a mechanical property of the crosslinked polyethylene are measured to obtain reference data. The pre-crosslinked material is heated to obtain a fast-aged pre-crosslinked material. The crosslinking degree and mechanical property of crosslinked polyethylene obtained from the fast-aged pre-crosslinked material are measured to obtain measurement results, which are compared with the reference data. If comparison results all fall within corresponding ranges, the time period of heating is increased by a step to repeat the above steps until the comparison results do not all fall within the corresponding ranges. A result obtained by subtracting the step from the time period of heating is converted into a time period of storage at the room temperature.

Abstract: Disclosed herein are systems, methods, and instrumentalities associated with reconstructing magnetic resonance (MR) images based on multi-slice, under-sampled MRI data (e.g., k-space data). The multi-slice MRI data may be acquired using a simultaneous multi-slice (SMS) technique and MRI information associated with multiple MRI slices may be entangled in the multi-slice MRI data. A neural network may be trained and used to disentangle the MRI information and reconstruct MRI images for the different slices. A data consistency component may be used to estimate k-space data based on estimates made by the neural network, from which respective MRI images associated with multiple MRI slices may be obtained by applying a Fourier transform to the k-space data.

Abstract: Provided are a keyword extraction method applicable to a Word text, an electronic device, and a computer-readable storage medium. The method includes: acquiring a text and extracting a body of the text; extracting a set number of keywords in the body by a TFIDF algorithm and a set number of keywords in the body by a TextRank algorithm, respectively; acquiring a text name and a text title of the text and segmenting the text name and the text title into words of the text name and words of the text title; constructing text feature vectors of the text and inputting the text feature vectors into a trained keyword extraction model; and extracting a keyword set from the keywords extracted by the TextRank algorithm by using the keyword extraction model to achieve text keyword extraction.

Abstract: Provided is a secondary battery, including: an electrode assembly, a packing bag, an electrode lead and an insulation part; the electrode assembly is housed in the packing bag having a sealing part on edge, and the electrode lead is connected to the electrode assembly and passes through the sealing part. The sealing part includes a main body area, a first step area and a first transition area which are located on same side of the electrode lead along width direction, and the main body area, first transition area and first step area are successively arranged along direction approaching the electrode lead; the insulation part is wrapped around the electrode lead, and has a first portion which is located on a side of the electrode lead close to the main body area along width direction and covered by the first step area on both sides in thickness direction.

Abstract: A method for measuring a gradient field of a nuclear magnetic resonance (NMR) system based on a diffusion effect uses a non-uniform field magnet, an NMR spectrometer, a radio frequency (RF) power amplifier, an RF coil, and a standard quantitative phantom with known apparent diffusion coefficient (ADC) and time constant for decay of transverse magnetization after RF-pulse (T2). A plurality of sets of signals are acquired by an NMR sequence with different diffusion-sensitive gradient durations or different echo spacings and the magnitude of the gradient field is calculated by fitting based on the plurality of sets of signals. The method does not require an additional dedicated magnetic field detection device, has a short measurement time, is easy to use with the NMR system, and is convenient to complete gradient field measurement at the installation site, thereby improving the installation and service efficiency of the NMR system.

Abstract: The invention provides a bioactive substance composition, a serum-free medium comprising the composition and the uses thereof. The bioactive substance composition is used for serum-free medium and/or composition and the preparation thereof; The serum-free medium and/or composition can be used for primary culture and secondary culture of cells and/or tissues. The cells are selected from any one or more of tendon and/or ligament derived cells, chondrocytes, meniscus stem cells, mesenchymal stem cells, skeleton stem cells, and muscle stem cells. The tissue is the musculoskeletal system tissue. The bioactive substance composition and/or serum-free medium and/or the composition can be used to prepare drugs for tissue and/or organ injury treatment; The tissue or organ injury is selected from the tissue or organ injury of the musculoskeletal system.

Abstract: Provided are a text duplicate checking method, an electronic device and a computer-readable storage medium. The method includes storing a fingerprint set and a corresponding text ID in a byte data manner to obtain a fingerprint library; acquiring a target text and creating a target fingerprint; obtaining a comparison fingerprint set from map memories according to the target fingerprint, and calculating a similarity between the target fingerprint and each comparison fingerprint in the comparison fingerprint set separately; and based on a determination result that a number of 1 s in binary values of one similarity is less than or equal to a preset value, querying a text ID corresponding to the one similarity, to implement duplicate checking of the target text.

Abstract: A method for measuring a gradient field of a nuclear magnetic resonance (NMR) system based on a diffusion effect uses a non-uniform field magnet, an NMR spectrometer, a radio frequency (RF) power amplifier, an RF coil, and a standard quantitative phantom with known apparent diffusion coefficient (ADC) and time constant for decay of transverse magnetization after RF-pulse (T2). A plurality of sets of signals are acquired by an NMR sequence with different diffusion-sensitive gradient durations or different echo spacings and the magnitude of the gradient field is calculated by fitting based on the plurality of sets of signals. The method does not require an additional dedicated magnetic field detection device, has a short measurement time, is easy to use with the NMR system, and is convenient to complete gradient field measurement at the installation site, thereby improving the installation and service efficiency of the NMR system.

Abstract: This application provides a method and apparatus for sequence processing, relates to the field of artificial intelligence, and specifically relates to the field of sequence data processing. The method includes: receiving an input sequence (S410); performing self-attention calculation on a first element in the input sequence by using an element included in M windows, to obtain a representation of the first element, where each window includes one element or a plurality of consecutive elements in the input sequence, there is an interval of at least one element between different windows, at least one of the M windows does not include the first element, and M is an integer greater than or equal to 1 (S420); and obtaining, based on the representation of the first element, an output sequence corresponding to the input sequence (S430).

Abstract: A drain programmed read-only memory includes a diffusion region that spans a width of a bitcell and forms a drain of a first transistor and a second transistor. A bit line lead in a metal layer adjacent the diffusion region extends across the width of the bitcell. A first via extends from an upper half of the bit line lead and couples to a drain of the first transistor. Similarly, a second via extends from a lower half of the bit line and couples to a drain of the second transistor.

Abstract: Deep learning based systems, methods, and instrumentalities are described herein for registering images from a same imaging modality and different imaging modalities. Transformation parameters associated with the image registration task are determined using a neural ordinary differential equation (ODE) network that comprises multiple layers, each configured to determine a respective gradient update for the transformation parameters based on a current state of the transformation parameters received by the layer. The gradient updates determined by the multiple ODE layers are then integrated and applied to initial values of the transformation parameters to obtain final parameters for completing the image registration task. The operations of the ODE network may be facilitated by a feature extraction network pre-trained to determine content features shared by the input images.