Lawrence Livermore National Security, LLC

Arizona State University

Rushil Anirudh

Jayaraman Jayaraman Thiagarajan

Peer-Timo Bremer

Hyojin Kim

Kyle Champley

Andreas Spanias

Suren Jayasuriya

Vivek Sivaraman Narayanaswamy

Kadri Aditya Mohan

Bhavya Kailkhura

Albert William Reed

Mikel L Landajuela

Robert C Blake

Thomas J O'Hara

A computer-based system and process are disclosed for reconstructing the internal electrical behavior of a patient's heart based partly or wholly on the patient's electrocardiogram (ECG). The output of the process may include, for example, a cardiac activation map, and/or a representation of transmembrane potentials over time. The process advantageously does not require any medical imaging of the patient, and does not require any special medical equipment. For example, the patient's activation map and transmembrane potentials may be reconstructed based solely on a preexisting or newly-obtained 12-lead cardiac ECG of the patient. The process makes use of a machine learning model, such as a neural network based model, trained with actual and/or simulated ECGs and intracardiac electrical data (typically transmembrane potentials) of many thousands of patients. Because an insufficient quantity of such data exists for actual patients, model training may be performed using ECGs and intracardiac electrical data obtained through computer simulations.

Machine learning based reconstruction of intracardiac electrical behavior based on electrocardiograms

A system for generating 2D slices of a 3D image of a target volume is provided. The system receives a target sinogram collected during a computed tomography scan of the target volume. The system inputs the target sinogram to a convolutional neural network (CNN) to generate predicted 2D slices of the 3D image. The CNN is trained using training 2D slices of training 3D images. The system initializes 2D slices to the predicted 2D slices. The system reconstructs 2D slices of the 3D image from the target sinogram and the initialized 2D slices.

Few-view computed tomography reconstruction using deep neural network inference

Various embodiments of a system and associated method for audio source separation based on generative priors trained on individual sources. Through the use of projected gradient descent optimization, the present approach simultaneously searches in the source-specific latent spaces to effectively recover the constituent sources. Though the generative priors can be defined in the time domain directly, it was found that using spectral domain loss functions leads to good-quality source estimates.

Systems and methods for unsupervised audio source separation using generative priors

A system for generating a 4D representation of a scene in motion given a sinogram collected from the scene while in motion. The system generates, based on scene parameters, an initial 3D representation of the scene indicating linear attenuation coefficients (LACs) of voxels of the scene. The system generates, based on motion parameters, a 4D motion field indicating motion of the scene. The system generates, based on the initial 3D representation and the 4D motion field, a 4D representation of the scene that is a sequence of 3D representations having LACs. The system generates a synthesized sinogram of the scene from the generated 4D representation. The system adjusts the scene parameters and the motion parameters based on differences between the collected sinogram and the synthesized sinogram. The processing is repeated until the differences satisfy a termination criterion.

Reconstruction of dynamic scenes based on differences between collected view and synthesized view

A system for classifying a target image with segments having attributes is provided. The system generates a graph for the target image that includes vertices representing segments of the image and edges representing relationships between the connected vertices. For each vertex, the system generates a subgraph that includes the vertex as a home vertex and neighboring vertices representing segments of the target image within a neighborhood of the segment represented by the home vertex. The system applies an autoencoder to each subgraph to generate latent variables to represent the subgraph. The system applies a machine learning algorithm to a feature vector comprising a universal image representation of the target image that is derived from the generated latent variables of the subgraphs to generate a classification for the target image.

Universal image representation based on a multimodal graph

Methods and systems are provided to generate an uncorrupted version of an image given an observed image that is a corrupted version of the image. In some embodiments, a corruption mimicking ('CM') system iteratively trains a corruption mimicking network ('CMN') to generate corrupted images given modeled images, updates latent vectors based on differences between the corrupted images and observed images, and applies a generator to the latent vectors to generate modeled images. The training, updating, and applying are performed until modeled images that are input to the CMN result in corrupted images that approximate the observed images. Because the CMN is trained to mimic the corruption of the observed images, the final modeled images represented the uncorrupted version of the observed images.

Growing community of inventors

Dublin, CA, United States of America

Rushil Anirudh