General Electric Company

Roy A Nilsen

Grant Morey Stevens

Jiang Hsieh

Mary Sue Kulpins

Brian Edward Nett

Dominic Joseph Crotty

Evgeny Drapkin

Thomas L Toth

Saad Ahmed Sirohey

Jiahua Fan

Sandeep Dutta

Holly A McDaniel

John Howard Londt

John Irvin Jackson

Eric Michael Gros

Zhoubo Li

Guangzhi Cao

Franco Rupcich

Jie Tang

Meghan Lynn Yue

Brian William Thomsen

Scott Slavic

Aleksandar Zavaljevski

Elizabeth Janus Nett

Suresh Narayanan Narayanan

Rajendra Kurady

Ann Soderman

Victor Eduardo McCormack

Darin R Okerland

Systems and methods for improved image denoising using a deep learning network model are disclosed. An example system includes an input data processor to process a first patient image of a first patient to add a first noise to the first patient image to form a noisy image input. The example system includes an image data denoiser to process the noisy image input using a first deep learning network to identify the first noise. The image data denoiser is to train the first deep learning network using the noisy image input. When the first deep learning network is trained to identify the first noise, the image data denoiser is to deploy the first deep learning network as a second deep learning network model to be applied to a second patient image of the first patient to identify a second noise in the second patient image.

Patient-specific deep learning image denoising methods and systems

An imaging system is provided including a selectable pre-object filter module, a detector, and a processing unit. The selectable pre-object filter module is configured to absorb radiation from the X-ray source to control distribution of X-rays passed to an object to be imaged. The selectable pre-object filter module has plural pre-object filter configurations providing corresponding X-ray distributions, and is selectable between the plural configurations to provide a selected pre-object filter configuration for a scan of the object. The detector is configured to receive X-rays that have passed through the object. The processing unit is operably coupled to the selectable pre-object filter module and the detector, and is configured to identify an anatomy to be imaged, determine a corresponding image quality and radiation dose for each of the plural pre-object filter configurations; and select the selected pre-object filter configuration based upon the determined corresponding image qualities and radiation doses.

Systems and methods for selecting bowtie filter configuration

Methods and systems are provided for direct monochromatic image generation for spectral computed tomography. In one embodiment, a method comprises acquiring projection data during a scan of a subject, reconstructing a low energy image and a high energy image from the projection data, and generating a monochromatic image from the low energy image and the high energy image. In this way, a monochromatic image may be generated directly from low and high energy images with a substantial reduction in image noise, especially when compared to a monochromatic image generated indirectly from material density images.

Direct monochromatic image generation for spectral computed tomography

Methods and systems are provided for automatic exposure control in computed tomography imaging systems. In one embodiment, a method comprises in response to an object-specific dose metric input by a user via a user interface operably coupled to a scanner, predicting an image quality of a diagnostic scan, calculating an output level of a source of the scanner for the image quality, and performing the diagnostic scan at the output level when the image quality is higher than a threshold. In this way, patient-specific image quality settings may be determined.

Methods and systems for computed tomography

An imaging system includes an identification module, an analysis module, and a determination module. The identification module is configured to identify a scanning mode of operation. The analysis module is configured to determine an attenuation for an object for a scan to be performed on the object. The determination module is configured to determine an image contrast for each of plural setting combinations, determine a corresponding tolerable noise for the image contrast for each of the setting combinations based on the scanning mode of operation, and determine a corresponding diagnostic dosage for each setting combination, the diagnostic dosages corresponding to the image contrast and tolerable noise for the corresponding setting combination. The determination module is also configured to select an operational setting for the scan to be performed using the dosages determined.

Systems and methods for selecting parameters using contrast and noise

An imaging system allows for modification of data acquisition parameters, where the parameters affect dose distribution. The system includes a library of exams, and a computer programmed to obtain a set of scanning parameters for image data acquisition and image reconstruction, obtain scan data from the library of exams that is based on a similarity metric for a patient to be scanned, simulate a new scan of the patient using the obtained scan data and using the obtained set of scanning parameters, and generate a new set of scanning parameters based on the simulation.

System and method of X-ray dose distribution for computed tomography based on simulation

Methods and systems are provided that obtain a scan prescription and receive physiologic feature or interest (FOI) measurements of an object of interest during a scan time. The methods and systems further compare at least one of the received physiologic FOI measurements with a predetermined threshold range, and determine a scan setting by evaluating an acquisition rule set based on the physiologic FOI measurement.

Systems and methods for adaptive computed tomography acquisition

An imaging system includes a computed tomography (CT) acquisition unit and a processing unit. The CT acquisition unit includes an X-ray source and a CT detector configured to collect CT imaging data of an object to be imaged. The processing unit includes at least one processor operably coupled to the CT acquisition unit. The processing unit is configured to control the CT acquisition unit to collect at least one sample projection during rotation of the CT acquisition unit about the object to be imaged, compare an intensity of the at least one sample projection to an intensity of a reference projection, select a time to perform an imaging scan based on the comparison of the intensity of the at least one sample projection to the intensity of the reference projection, and control the CT acquisition unit to perform the imaging scan.

Systems and methods for imaging phase selection for computed tomography imaging

An imaging system includes an identification module, a determination module, and a display module. The identification module is configured to identify one or more first scanning parameters and one or more first display parameters corresponding to a first image, and to identify one or more second scanning parameters corresponding to scanning information acquired during a second scan. The determination module is configured to determine, based on the one or more first scanning parameters and the one or more second scanning parameters, one or more second display parameters so that the scanning information acquired during the second scan may be used to provide a second image appearing more similar to the first image. The display module is configured to use the one or more second display parameters to provide the second image configured to be displayed to a viewer.

Systems and methods for selecting image display parameters

A method and apparatus for multi-phase cardiac CT imaging includes an adaptive, selective reconstruction process that, when possible, implements an image or non-phase location driven reconstruction process to reconstruct cardiac CT images. If the hardware parameters support an image location driven reconstruction for the particular imaging session then the image location driven reconstruction is implemented. However, if image location driven reconstruction is not supported, a default phase location driven reconstruction is employed. Image location driven reconstruction improves system throughput and improves image quality as more cardiac phase locations can routinely be generated to better 'freeze' the motion of the heart of a patient.

Method and apparatus of multi-phase cardiac imaging

A method and apparatus for processing raw image data to create processed images. Raw image data is acquired. The raw image data is decomposed by a data decomposer into N subsets of raw image data. The number N is based on a number of available image generation processors. The N subsets of raw image data are processed by at least one image generation processor to create processed image data. If more than one image generation processor is available, the image generation processors perform image processing on the raw image data in parallel with respect to each other.

Method and apparatus for image reconstruction using data decomposition for all or portions of the processing flow

The present invention, in one form, is a system which, in one embodiment, corrects for position errors of an x-ray beam caused by dynamic movement and thermal drift of the imaging system. More specifically, signals from a detector array are utilized to determine a dynamic movement error to correct for errors caused by the rotation of a gantry and a thermal drift error to correct for imaging system x-ray beam movement errors. In one embodiment, a pre-patient collimator is utilized to alter the z-axis position of the x-ray beam as determined by the dynamic movement and the temperature error profiles.

Growing community of inventors

Menomonee Falls, WI, United States of America

Roy A Nilsen