公开(公告)号：WO2018187020A1

公开(公告)日：2018-10-11

申请号：PCT/US2018/023074

申请日：2018-03-19

Applicant: GENERAL ELECTRIC COMPANY

Inventor： FU, Lin ,  RAMANI, Sathish ,  DE MAN, Bruno Kristiaan Bernard ,  RUI, Xue ,  AHN, Sangtae

IPC: G06T11/00

Abstract: The present approach relates to the use of machine learning and deep learning systems suitable for solving large-scale, space-variant tomographic reconstruction and/or correction problems. In certain embodiments, a tomographic transform of measured data obtained from a tomography scanner is used as an input to a neural network. In accordance with certain aspects of the present approach, the tomographic transform operation(s) is performed separate from or outside the neural network such that the result of the tomographic transform operation is instead provided as an input to the neural network. In addition, in certain embodiments, one or more layers of the neural network may be provided as wavelet filter banks.

公开(公告)号：EP3608877A1

公开(公告)日：2020-02-12

申请号：EP19188957.5

申请日：2019-07-29

Applicant: General Electric Company ,  Rensselaer Polytechnic Institute

Inventor： FU, Lin ,  RAMANI, Sathish ,  TANG, Jie ,  DE MAN, Bruno Kristiaan Bernard ,  PACK, Jed Douglas ,  HSIEH, Jiang ,  WANG, Ge

IPC: G06T11/00

Abstract: The present disclosure relates to image reconstruction with favorable properties in terms of noise reduction, spatial resolution, detail preservation and computational complexity. The disclosed techniques may include some or all of: a first-pass reconstruction, a simplified datafit term, and/or a deep learning denoiser. In various implementations, the disclosed technique is portable to different CT platforms, such as by incorporating a first-pass reconstruction step.

公开(公告)号：EP3671648A1

公开(公告)日：2020-06-24

申请号：EP19216339.2

申请日：2019-12-13

Applicant: General Electric Company

Inventor： YANOFF, Brian David ,  WU, Mingye ,  FU, Lin ,  EDIC, Peter Michael ,  RUI, Xue ,  FU, Geng ,  JIN, Yannan ,  GRONBERG, Fredrik

IPC: G06T11/00

Abstract: A method for imaging an object to be reconstructed includes acquiring projection data corresponding to the object. Furthermore, the method includes generating a measured sinogram based on the acquired projection data and formulating a forward model, where the forward model is representative of a characteristic of the imaging system. In addition, the method includes generating an estimated sinogram based on an estimated image of the object and the forward model and formulating a statistical model based on at least one of pile-up characteristics and dead time characteristics of a detector of the imaging system. Moreover, the method includes determining an update corresponding to the estimated image based on the statistical model, the measured sinogram, and the estimated sinogram and updating the estimated image based on the determined update to generate an updated image of the object. Additionally, the method includes outputting a final image of the object.

公开(公告)号：EP3607529A1

公开(公告)日：2020-02-12

申请号：EP18715880.3

申请日：2018-03-19

Applicant: General Electric Company

Inventor： FU, Lin ,  RAMANI, Sathish ,  DE MAN, Bruno Kristiaan Bernard ,  RUI, Xue ,  AHN, Sangtae

IPC: G06T11/00