公开(公告)号：US20230342996A1

公开(公告)日：2023-10-26

申请号：US18305697

申请日：2023-04-24

申请人： The General Hospital Corporation ,  Trustees of Boston University

发明人： Danyal Fareed Bhutto ,  Matthew S. Rosen ,  Neha Koonjoo ,  Bo Zhu

IPC分类号： G06T11/00 ,  G16H30/40 ,  A61B5/055

CPC分类号： G06T11/006 ,  G16H30/40 ,  A61B5/055 ,  G06T2211/421 ,  G06T2210/41

摘要： Systems, methods, and media for complex input data configurations for imaging applications. Complex data optimization can be provided to improve accuracy of models (e.g., neural networks) used to reconstruct medical images from raw sensor data, for example. Complex data optimization can include applying raw sensor data to an input layer of a neural network to generate an input vector ordered such that real components and imaginary components of samples in the raw sensor data are adjacent. The input vector can then be applied to convolutional layer of the neural network.

公开(公告)号：US11620772B2

公开(公告)日：2023-04-04

申请号：US16326910

申请日：2017-09-01

申请人： THE GENERAL HOSPITAL CORPORATION

发明人： Matthew S. Rosen ,  Bo Zhu ,  Bruce R. Rosen

IPC分类号： G06T11/00 ,  G01R33/56 ,  G06N3/08 ,  G01R33/12 ,  G06K9/62 ,  G06N3/04 ,  G16H30/40 ,  A61B5/00 ,  A61B5/055 ,  A61B6/03 ,  A61B6/00 ,  A61B8/00 ,  A61B8/08 ,  G01R33/48

摘要： A system may transform sensor data from a sensor domain to an image domain using data-driven manifold learning techniques which may, for example, be implemented using neural networks. The sensor data may be generated by an image sensor, which may be part of an imaging system. Fully connected layers of a neural network in the system may be applied to the sensor data to apply an activation function to the sensor data. The activation function may be a hyperbolic tangent activation function. Convolutional layers may then be applied that convolve the output of the fully connected layers for high level feature extraction. An output layer may be applied to the output of the convolutional layers to deconvolve the output and produce image data in the image domain.

公开(公告)号：US20180203081A1

公开(公告)日：2018-07-19

申请号：US15872449

申请日：2018-01-16

申请人： THE GENERAL HOSPITAL CORPORATION

发明人： Ouri Cohen ,  Bo Zhu ,  Matthew S. Rosen

IPC分类号： G01R33/48 ,  G01R33/56 ,  G06N3/08 ,  G06N3/04

CPC分类号： G01R33/4828 ,  G01R33/50 ,  G01R33/5608 ,  G06N3/04 ,  G06N3/08 ,  G06N3/084

摘要： Disclosed is a system and method for estimating quantitative parameters of a subject using a magnetic resonance (“MR”) system using a dictionary. The dictionary may include a plurality of signal templates that sparsely sample acquisition parameters used when acquiring data. The acquired data is compared with the dictionary using a neural network. Thus, systems and methods are provided that are more computationally efficient, and have reduced data storage requirements than traditional MRF reconstruction systems and methods.

公开(公告)号：US20230215161A1

公开(公告)日：2023-07-06

申请号：US18183639

申请日：2023-03-14

申请人： The General Hospital Corporation

发明人： Matthew S. Rosen ,  Bo Zhu ,  Bruce R. Rosen

IPC分类号： G06T11/00

CPC分类号： G06T11/006 ,  G06T2207/20084 ,  G06T2207/20081 ,  G06T2207/20024

摘要： A system may transform sensor data from a sensor domain to an image domain using data-driven manifold learning techniques which may, for example, be implemented using neural networks. The sensor data may be generated by an image sensor, which may be part of an imaging system. Fully connected layers of a neural network in the system may be applied to the sensor data to apply an activation function to the sensor data. The activation function may be a hyperbolic tangent activation function. Convolutional layers may then be applied that convolve the output of the fully connected layers for high level feature extraction. An output layer may be applied to the output of the convolutional layers to deconvolve the output and produce image data in the image domain.

公开(公告)号：US20190213761A1

公开(公告)日：2019-07-11

申请号：US16326910

申请日：2017-09-01

申请人： THE GENERAL HOSPITAL CORPORATION

发明人： Matthew S. Rosen ,  Bo Zhu ,  Bruce R. Rosen

IPC分类号： G06T11/00 ,  G01R33/56 ,  G01R33/48 ,  G06K9/62 ,  A61B5/00 ,  A61B5/055 ,  A61B6/03 ,  A61B6/00 ,  A61B8/08 ,  A61B8/00 ,  G16H30/40 ,  G06N3/08

CPC分类号： G06T11/006 ,  A61B5/0035 ,  A61B5/0059 ,  A61B5/055 ,  A61B5/7267 ,  A61B5/7425 ,  A61B6/032 ,  A61B6/037 ,  A61B6/463 ,  A61B6/5247 ,  A61B8/463 ,  A61B8/5261 ,  G01R33/12 ,  G01R33/4818 ,  G01R33/4824 ,  G01R33/5608 ,  G06K9/623 ,  G06K9/6274 ,  G06K2209/05 ,  G06N3/0454 ,  G06N3/08 ,  G06T2210/41 ,  G16H30/40

摘要： A system may transform sensor data from a sensor domain to an image domain using data-driven manifold learning techniques which may, for example, be implemented using neural networks. The sensor data may be generated by an image sensor, which may be part of an imaging system. Fully connected layers of a neural network in the system may be applied to the sensor data to apply an activation function to the sensor data. The activation function may be a hyperbolic tangent activation function. Convolutional layers may then be applied that convolve the output of the fully connected layers for high level feature extraction. An output layer may be applied to the output of the convolutional layers to deconvolve the output and produce image data in the image domain.

公开(公告)号：US20230342995A1

公开(公告)日：2023-10-26

申请号：US18304974

申请日：2023-04-21

申请人： The General Hospital Corporation ,  Trustees of Boston University

发明人： Neha Koonjoo ,  Matthew S. Rosen ,  Bo Zhu ,  Danyal Fareed Bhutto

IPC分类号： G06T11/00 ,  G06T3/40 ,  G06N3/08

CPC分类号： G06T11/005 ,  G06T3/40 ,  G06N3/08 ,  G06T2210/41

摘要： Systems, methods, and media for patch-based medical image generation for complex input datasets. Patch-based medical image generation can include creating a training dataset with an image patch and corresponding sensor data patch and training a neural network using the training dataset. Then, sensor data acquired from a patient using a medical imaging system can be applied as input to the neural network, and a medical image of the patient can be generated based on an output of the neural network.