Abstract: Methods and systems are described herein for facilitating dynamic selection of training routines for training machine learning models based on data shift severity. The system may detect, in a production dataset, a data shift from a training dataset used to train a machine learning model. The system may provide the production dataset to an adversarial network to cause the adversarial network to generate synthetic data based on the production dataset. The system may determine that a magnitude of the data shift satisfies a first threshold and does not satisfy a second threshold and, in response, may perform a first training routine. The first training routine may involve training a second machine learning model using the training dataset, the production dataset, and the synthetic data. The system may then replace the first machine learning model with the second machine learning model in a production environment.

Abstract: Methods and systems are described herein for facilitating model training related to data shifts. The system may detect, in a production dataset, a data shift from a training dataset used to train a machine learning model. The system may provide the training dataset and the production dataset to an adversarial network to train a first classifier and a second classifier, respectively. The system may provide synthetic data derived from the production dataset to the adversarial network to cause the first classifier and the second classifier to classify the synthetic data. Based on the classifications received from the adversarial network, the system may exclude the synthetic data from an updated training dataset for updating the machine learning model.

Abstract: Methods and systems are described herein for identifying malicious activity using deep-linked items related to stochastic images. The system may receive event data associated with an event performed in connection with a token. The system may generate a token embedding based on the event data and may obtain, via a stochastic machine learning model based on the token embedding, an image related to the event. The system may generate, for display, the image and the event data. In some embodiments, the image may be deep-linked to functionality for submitting feedback relating to the event. The system may receive feedback related to the image indicating an invalid event. Based on the feedback related to the image, the system may perform a remedial action related to the token or to the event.

Abstract: Methods and systems are described herein for facilitating token use authentication for access tokens using stochastic images. The system may detect an authentication request to authenticate use of an access token. The access token may be associated with a first image previously displayed to an authenticated user of the access token. The system may retrieve the first image previously displayed to the authenticated user and input parameters previously used to generate the first image. The system may obtain, from a stochastic machine learning model based on the input parameters, a second image different from the first image. The system may generate, for display, an image set including the first image and the second image and may receive a selection of the first image from the image set. The system may then grant the authentication request based on the selection of the first image.

Abstract: Methods and systems are described herein for generating deep-linked stochastic image representations of access tokens that embed token access deep links on a mobile application interface. The system may obtain, in connection with a request to register an access token with an account, token data associated with the access token and event data associated with one or more events performed with the access token. The system may generate, for input to a stochastic machine learning model, input vectors using the token data and the event data. The system may obtain, via the stochastic machine learning model based on the input vectors, an image for the access token and may generate, for display on a user interface associated with the account, an image representation of the access token including the image and a deep link to functionality associated with the access token.

Abstract: In some implementations, a style system may receive, from a repository, a plurality of files associated with an entity. The style system may apply a machine learning model to the plurality of files to determine a set of rules associated with images or text included in the plurality of files. The style system may generate a document that indicates the set of rules and may output, to a user device, the document.

Abstract: Methods and systems for generating synthetic training data based on abandoned web activity data are described herein. In some aspects, the system determines that a user abandoned a user activity included in web activity data for the user. The system processes, using a first machine learning model, the abandoned web activity data to generate a probability for each entry in the abandoned web activity data. Each probability indicates a likelihood that the abandoned user activity would have been completed. The system generates a synthetic dataset that includes abandoned web activity with a probability above a threshold. The system uses the synthetic dataset to train a second machine learning model. The system tests the machine learning model using a testing dataset based on completed user activities.

Abstract: Methods and systems are described herein for generating images based on generated clusters. The system retrieves, from an account of a user, historic events associated with the user and determines clusters of the historic events based on location and time data associated with the historic events. The system generates an input for a machine learning model for a first cluster of historic events, where the machine learning model has been trained to generate images based on clusters of historic events. The system then inputs the input into the machine learning model to cause the machine learning model to output images depicting locations associated with the first cluster of historic events. The system outputs the images in conjunction with the first cluster of historic events.

Abstract: This disclosure discloses a method and system for predicting disease quantification parameters for an anatomical structure. The method includes extracting a centerline structure based on a medical image. The method further includes predicting the disease quantification parameter for each sampling point on the extracted centerline structure by using a GNN, with each node corresponds to a sampling point on the extracted centerline structure and each edge corresponds to a spatial constraint relationship between the sampling points. For each node, a local feature is extracted based on the image patch for the corresponding sampling point by using a local feature encoder, and a global feature is extracted by using a global feature encoder based on a set of image patches for a set of sampling points, which include the corresponding sampling point and have a spatial constraint relationship defined by the centerline structure.

Abstract: The present disclosure relates to a method and a system for generating anatomical labels of an anatomical structure. The method includes receiving an anatomical structure with an extracted centerline, or a medical image containing the anatomical structure with the extracted centerline; and predicting the anatomical labels of the anatomical structure based on the centerline of the anatomical structure, by utilizing a trained deep learning network. The deep learning network includes a branched network, a Graph Neural Network, a Recurrent Neural Network and a Probability Graph Model, which are connected sequentially in series. The branched network includes at least two branch networks in parallel. The method in the disclosure can automatically generate the anatomical labels of the whole anatomical structure in medical image end to end and provide high prediction accuracy and reliability.

Abstract: Embodiments of the disclosure provide systems and methods for analyzing a medical image containing a vessel structure using a sequential model. An exemplary system includes a communication interface configured to receive the medical image and the sequential model. The sequential model includes a vessel extraction sub-model and a lesion analysis sub-model. The vessel extraction sub-model and the lesion analysis sub-model are independently or jointly trained. The exemplary system also includes at least one processor configured to apply the vessel extraction sub-model on the received medical image to extract location information of the vessel structure. The at least one processor also applies the lesion analysis sub-model on the received medical image and the location information extracted by the vessel extraction sub-model to obtain a lesion analysis result of the vessel structure. The at least one processor further outputs the lesion analysis result of the vessel structure.

Abstract: Embodiments of the disclosure provide methods and systems for multi-modality joint analysis of a plurality of vascular images. The exemplary system may include a communication interface configured to receive the plurality of vascular images acquired using a plurality of imaging modalities. The system may further include at least one processor, configured to extract a plurality of vessel models for a vessel of interest from the plurality of vascular images. The plurality of vessel models are associated with the plurality of imaging modalities, respectively. The at least one processor is also configured to fuse the plurality of vessel models associated with the plurality of imaging modalities to generate a fused model for the vessel of interest. The at least one processor is further configured to provide a diagnostic analysis result based on the fused model of the vessel of interest.

Abstract: Embodiments of the disclosure provide methods and systems for joint abnormality detection and physiological condition estimation from a medical image. The exemplary method may include receiving, by at least one processor, the medical image acquired by an image acquisition device. The medical image includes an anatomical structure. The method may further include applying, by the at least one processor, a joint learning model to determine an abnormality condition and a physiological parameter of the anatomical structure jointly based on the medical image. The joint learning model satisfies a predetermined constraint relationship between the abnormality condition and the physiological parameter.

Abstract: The disclosure provides a method, device and a computer-readable medium for performing three-dimensional blood vessel reconstruction. The device includes an interface configured to receive a single-view two-dimensional image of a blood vessel of a patient, where the single-view two-dimensional image is a projection image acquired in a predetermined projection direction. The device further includes a processor configured to estimate three-dimensional information of the blood vessel from the single-view two-dimensional image using an inference model, and reconstruct a three-dimensional model of the blood vessel based on the three-dimensional information.

Abstract: Methods and Systems for generating a centerline for an object in an image and computer readable medium are provided. The method includes receiving an image containing the object. The method also includes generating the centerline of the object, by a processor, using a reinforcement learning network configured to predict movement of a virtual agent that traces the centerline in the image. The reinforcement learning network is further configured to perform at least one auxiliary task that detects a bifurcation in a trajectory of the object. The reinforcement learning network is trained by maximizing a cumulative reward and minimizing an auxiliary loss of the at least one auxiliary task. Additionally, the method includes displaying the centerline of the object.

Abstract: The disclosure relates to a computer-implemented method for analyzing an image sequence of a periodic physiological activity, a system, and a medium. The method includes receiving the image sequence from an imaging device, and the image sequence has a plurality of images. The method further includes identifying at least one feature portion in a selected image, which moves responsive to the periodic physiological activity. The method also includes detecting, by a processor, the corresponding feature portions in other images of the image sequence and determining, by the processor, a phase of a the selected image in the image sequence based on the motion of the feature portion.

Abstract: A computer-implemented method for predicting a blood vessel stenosis is disclosed. The method may include extracting a blood vessel path and its centerline based on the image of the blood vessel. The method may further include determining a candidate stenosis for the blood vessel path and identifying image blocks along the centerline of the blood vessel path within a range of candidate stenosis for the blood vessel path determined based on the candidate stenosis. The method may also include determining a degree of stenosis for the blood vessel path by applying a trained learning network comprising a convolutional neural network and a recurrent neural network on the image blocks within the range of candidate stenosis.

Abstract: Embodiments of the disclosure provide systems and methods for generating a diagnosis report based on a medical image of a patient. The system includes a communication interface configured to receive the medical image acquired by an image acquisition device. The system further includes at least one processor. The at least one processor is configured to detect a medical condition based on the medical image and automatically generate text information describing the medical condition. The at least one processor is further configured to construct the diagnosis report, where the diagnosis report includes at least one image view showing the medical condition and a report view including the text information describing the medical condition. The system also includes a display configured to display the diagnosis report.

Abstract: Embodiments of the disclosure provide systems and methods for detecting a medical condition of a subject. The system includes a communication interface configured to receive a sequence of images acquired from the subject by an image acquisition device and an end-to-end multi-task learning model. The end-to-end multi-task learning model includes an encoder, a Convolutional Recurrent Neural Network (ConvRNN), and at least one of a decoder and a classifier. The system further includes at least one processor configured to extract feature maps from the images using the encoder, capture contextual information between adjacent images in the sequence using the ConvRNN, and detect medical condition of the subject using the classifier based on the extracted feature maps of the image slices and the contextual information or segment each image slice using the decoder to obtain a region of interest indicative of the medical condition based on the extracted feature maps.

Abstract: Systems and methods for generating a centerline for an object in an image are provided. An exemplary method includes receiving an image containing the object. The method also includes detecting at least one bifurcation of the object using a trained bifurcation learning network based on the image. The method further includes extracting the centerline of the object based on a constraint condition that the centerline passes through the detected bifurcation.