-
公开(公告)号：US20200012904A1
公开(公告)日：2020-01-09
申请号：US16143703
申请日：2018-09-27
Applicant: General Electric Company
Inventor： Qian Zhao , Min Zhang , Gopal Avinash
Abstract: Systems and techniques for classification based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The plurality of images is associated with a plurality of masks, a plurality of image level labels, and/or a bounding box. The system also generates a first loss function based on the plurality of masks, a second loss function based on the plurality of image level labels, and a third loss function based on the bounding box. Furthermore, the system generates a fourth loss function based on the first loss function, the second loss function and the third loss function, where the fourth loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.
-
公开(公告)号：US20200012895A1
公开(公告)日：2020-01-09
申请号：US16046084
申请日：2018-07-26
Applicant: General Electric Company
Inventor： Qian Zhao , Min Zhang , Gopal Avinash
IPC: G06K9/62
Abstract: Systems and techniques for classification and localization based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The training data is associated with a plurality of patients from at least one imaging device. The plurality of images is associated with a plurality of masks from a plurality of objects. The convolutional neural network comprises a decoder consisting of at least one up-sampling layer and at least one convolutional layer. The system also generates a loss function based on the plurality of masks, where the loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.
-
公开(公告)号：US10438354B2
公开(公告)日：2019-10-08
申请号：US16359647
申请日：2019-03-20
Applicant: General Electric Company
Inventor： Jiang Hsieh , Gopal Avinash , Saad Sirohey
Abstract: Methods and apparatus for monitoring and improving imaging system operation are provided. An example apparatus includes a first deployed deep learning network (DLN) which operates with an acquisition engine to generate an imaging device configuration. The example apparatus includes a second deployed DLN which operates with a reconstruction engine based on acquired image data. The example apparatus includes a first assessment engine with a third deployed DLN. The assessment engine receives output from at least one of the acquisition engine or the reconstruction engine to assess operation of the respective at least one of the acquisition engine or the reconstruction engine and to provide feedback to the respective at least one of the acquisition engine or the reconstruction engine. The first deployed DLN and the second deployed DLN are generated and deployed from first and second training DLNS, respectively.
-
4.发明申请公开(公告)号：US20190026608A1
公开(公告)日：2019-01-24
申请号：US16126762
申请日：2018-09-10
Applicant: General Electric Company
Inventor： Jiang Hsieh , Gopal Avinash , Saad Sirohey , Xin Wang , Zhye Yin , Bruno De Man
CPC classification number: G06K9/6265 , G06K9/036 , G06K9/4604 , G06K9/4628 , G06N3/04 , G06N3/0454 , G06N3/08 , G06N3/084 , G06T7/0002 , G06T2207/10081 , G06T2207/10088 , G06T2207/10104 , G06T2207/10116 , G06T2207/10132 , G06T2207/30168
Abstract: Methods and apparatus to automatically generate an image quality metric for an image are provided. An example method includes automatically processing a first medical image using a deployed learning network model to generate an image quality metric for the first medical image, the deployed learning network model generated from a digital learning and improvement factory including a training network, wherein the training network is tuned using a set of labeled reference medical images of a plurality of image types, and wherein a label associated with each of the labeled reference medical images indicates a central tendency metric associated with image quality of the image. The example method includes computing the image quality metric associated with the first medical image using the deployed learning network model by leveraging labels and associated central tendency metrics to determine the associated image quality metric for the first medical image.
-
公开(公告)号：US20220284579A1
公开(公告)日：2022-09-08
申请号：US17751349
申请日：2022-05-23
Applicant: General Electric Company
Inventor： Katelyn Rose Nye , Gireesha Rao , Gopal Avinash , Ravi Soni
Abstract: Apparatus, systems, and methods to improve imaging quality control, image processing, identification of findings, and generation of notification at or near a point of care are disclosed and described. An example imaging apparatus includes a processor to at least: process the first image data using a trained learning network to generate a first analysis of the first image data; identify a clinical finding in the first image data based on the first analysis; compare the first analysis to a second analysis, the second analysis generated from second image data obtained in a second image acquisition; and, when comparing identifies a change between the first analysis and the second analysis, generate a notification at the imaging apparatus regarding the clinical finding to trigger a responsive action.
-
Patent Agency Ranking
公开(公告)号：US10885400B2
公开(公告)日：2021-01-05
申请号：US16143703
申请日：2018-09-27
Applicant: General Electric Company
Inventor： Qian Zhao , Min Zhang , Gopal Avinash
Abstract: Systems and techniques for classification based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The plurality of images is associated with a plurality of masks, a plurality of image level labels, and/or a bounding box. The system also generates a first loss function based on the plurality of masks, a second loss function based on the plurality of image level labels, and a third loss function based on the bounding box. Furthermore, the system generates a fourth loss function based on the first loss function, the second loss function and the third loss function, where the fourth loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.
-
公开(公告)号：US20200349394A1
公开(公告)日：2020-11-05
申请号：US16928462
申请日：2020-07-14
Applicant: General Electric Company
Inventor： Qian Zhao , Min Zhang , Gopal Avinash
IPC: G06K9/62
Abstract: Systems and techniques for classification and localization based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The training data is associated with a plurality of patients from at least one imaging device. The plurality of images is associated with a plurality of masks from a plurality of objects. The convolutional neural network comprises a decoder consisting of at least one up-sampling layer and at least one convolutional layer. The system also generates a loss function based on the plurality of masks, where the loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.
-
8.发明申请公开(公告)号：US20200211694A1
公开(公告)日：2020-07-02
申请号：US16233670
申请日：2018-12-27
Applicant: General Electric Company
Inventor： Katelyn Nye , Gireesha Rao , Gopal Avinash , Christopher Austin
Abstract: Apparatus, systems, and methods to improve automated identification, monitoring, processing, and control of a condition impacting a patient using image data and artificial intelligence classification are disclosed. An example image processing apparatus includes an artificial intelligence classifier to: process first image data for a patient from a first time to determine a first classification result indicating a first severity of a condition for the patient; and process second image data for the patient from a second time to determine a second classification result indicating a second severity of the condition for the patient. The example image processing apparatus includes a comparator to compare the first classification result and the second classification result to determine a change and a progression of the condition associated with the change. The example image processing apparatus includes an output generator to trigger an action when the progression corresponds to a worsening of the condition.
-
9.发明申请公开(公告)号：US20200097773A1
公开(公告)日：2020-03-26
申请号：US16697904
申请日：2019-11-27
Applicant: General Electric Company
Inventor： Jiang Hsieh , Gopal Avinash , Saad Sirohey , Xin Wang , Zhye Yin , Bruno DeMan
Abstract: Methods and apparatus to automatically generate an image quality metric for an image are provided. An example method includes automatically processing a first medical image using a deployed learning network model to generate an image quality metric for the first medical image, the deployed learning network model generated from a digital learning and improvement factory including a training network, wherein the training network is tuned using a set of labeled reference medical images of a plurality of image types, and wherein a label associated with each of the labeled reference medical images indicates a central tendency metric associated with image quality of the image. The example method includes computing the image quality metric associated with the first medical image using the deployed learning network model by leveraging labels and associated central tendency metrics to determine the associated image quality metric for the first medical image.
-
公开(公告)号：US20200012898A1
公开(公告)日：2020-01-09
申请号：US16058984
申请日：2018-08-08
Applicant: General Electric Company
Inventor： Qian Zhao , Min Zhang , Gopal Avinash
Abstract: Systems and techniques for classification based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The training data is associated with a plurality of patients from at least one imaging device. The plurality of images is associated with a plurality of masks from a plurality of objects. The system also generates a loss function based on the plurality of masks, where the loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.
-
-
-
-
-
-
-
-
-
Search Results
31
records
(took 0.04 seconds)
