公开(公告)号：US10679346B2

公开(公告)日：2020-06-09

申请号：US15884151

申请日：2018-01-30

Applicant: General Electric Company

Inventor： Eric Michael Gros ,  David Erik Chevalier

IPC: A61B5/00 ,  G06T7/00 ,  G06N3/08 ,  G16H30/20 ,  G06K9/46 ,  G06K9/62 ,  G06N3/02

Abstract: Methods and systems are provided for capturing deep learning training data from imaging systems. In one embodiment, a method for an imaging system comprises performing a scan of a subject to acquire imaging data, inputting the imaging data to a deep neural network, displaying an output of the deep neural network and an image reconstructed from the imaging data, and transmitting an intermediate representation of the imaging data generated by the deep neural network to a server for training a central deep neural network. In this way, imaging data may be leveraged for training and developing global deep learning models without transmitting the imaging data itself, thereby preserving patient privacy.

公开(公告)号：US10383590B2

公开(公告)日：2019-08-20

申请号：US14868146

申请日：2015-09-28

Applicant: General Electric Company

Inventor： Michael Sarju Vaz ,  Elizabeth Janus Nett ,  David Joseph Pitterle ,  David Erik Chevalier ,  Christine Carol Hammond ,  Chelsey Lewis

IPC: A61B6/03 ,  A61B6/00

Abstract: Methods and systems are provided for adaptive scan control. In one embodiment, a method comprises: while performing a scan of a scan subject, processing acquired projection data to measure a contrast level; responsive to the contrast level increasing above a first threshold, automatically switching the scan from a first scan protocol to a second scan protocol; responsive to the contrast level decreasing below a second threshold, automatically switching the scan from the second scan protocol to the first scan protocol; and responsive to the contrast level decreasing below a third threshold, automatically ending the scan. In this way, multiple scan protocols, such as angiography and perfusion scan protocols, can be interleaved within a single scan without the use of a separate timing bolus scan.

公开(公告)号：US10755407B2

公开(公告)日：2020-08-25

申请号：US15884081

申请日：2018-01-30

Applicant: General Electric Company

Inventor： Eric Michael Gros ,  David Erik Chevalier

IPC: G06K9/00 ,  G06T7/00 ,  G06N3/08 ,  G16H30/20

Abstract: Methods and systems are provided for generating deep learning training data with an imaging system. In one embodiment, a method for an imaging system comprises performing a scan of a subject to acquire imaging data, training a deep neural network on the imaging data to obtain updates to the deep neural network, and transmitting the updates to a server for training a central deep neural network. In this way, imaging data may be leveraged for training and developing global deep learning models without transmitting the imaging data itself, thereby preserving patient privacy.

公开(公告)号：US20190236773A1

公开(公告)日：2019-08-01

申请号：US15884081

申请日：2018-01-30

Applicant: General Electric Company

Inventor： Eric Michael Gros ,  David Erik Chevalier

IPC: G06T7/00 ,  G06N3/08 ,  G16H30/20

Abstract: Methods and systems are provided for generating deep learning training data with an imaging system. In one embodiment, a method for an imaging system comprises performing a scan of a subject to acquire imaging data, training a deep neural network on the imaging data to obtain updates to the deep neural network, and transmitting the updates to a server for training a central deep neural network. In this way, imaging data may be leveraged for training and developing global deep learning models without transmitting the imaging data itself, thereby preserving patient privacy.

公开(公告)号：US20190236774A1

公开(公告)日：2019-08-01

申请号：US15884151

申请日：2018-01-30

Applicant: General Electric Company

Inventor： Eric Michael Gros ,  David Erik Chevalier

IPC: G06T7/00 ,  G06N3/08 ,  G16H30/20

Abstract: Methods and systems are provided for capturing deep learning training data from imaging systems. In one embodiment, a method for an imaging system comprises performing a scan of a subject to acquire imaging data, inputting the imaging data to a deep neural network, displaying an output of the deep neural network and an image reconstructed from the imaging data, and transmitting an intermediate representation of the imaging data generated by the deep neural network to a server for training a central deep neural network. In this way, imaging data may be leveraged for training and developing global deep learning models without transmitting the imaging data itself, thereby preserving patient privacy.

公开(公告)号：US20170086772A1

公开(公告)日：2017-03-30

申请号：US14868146

申请日：2015-09-28

Applicant: General Electric Company

Inventor： Michael Sarju Vaz ,  Elizabeth Janus Nett ,  David Joseph Pitterle ,  David Erik Chevalier ,  Christine Carol Hammond ,  Chelsey Lewis

IPC: A61B6/00 ,  A61B6/03

CPC classification number: A61B6/504 ,  A61B6/032 ,  A61B6/481 ,  A61B6/486 ,  A61B6/501 ,  A61B6/507 ,  A61B6/5217 ,  A61B6/54 ,  A61B6/542

Abstract: Methods and systems are provided for adaptive scan control. In one embodiment, a method comprises: while performing a scan of a scan subject, processing acquired projection data to measure a contrast level; responsive to the contrast level increasing above a first threshold, automatically switching the scan from a first scan protocol to a second scan protocol; responsive to the contrast level decreasing below a second threshold, automatically switching the scan from the second scan protocol to the first scan protocol; and responsive to the contrast level decreasing below a third threshold, automatically ending the scan. In this way, multiple scan protocols, such as angiography and perfusion scan protocols, can be interleaved within a single scan without the use of a separate timing bolus scan.