Rensselaer Polytechnic Institute

Qingsong Yang

Ge Wang

Wenxiang Cong

Lars Arne Gjesteby

Hongming Shan

Guang Li

Fenglin Liu

Systems and methods for reconstructing images for computed tomography are provided. Image reconstruction can be based on a realistic polychromatic physical model, and can include use of both an analytical algorithm and a single-variable optimization method. The optimization method can be used to solve the non-linear polychromatic X-ray integral model in the projection domain, resulting in an accurate decomposition for sinograms of two physical basis components.

Image reconstruction method for computed tomography

A deep neural network for metal artifact reduction is described. A method for computed tomography (CT) metal artifact reduction (MAR) includes receiving a first CT image data; receiving a second CT image data; and generating, by an artificial neural network (ANN), CT output image data configured to include fewer artifacts compared to the first and second CT image data. The ANN includes at least two parallel CT image data streams and a CT output image data stream. A first of the at least two parallel CT image data streams is based, at least in part, on the first CT image data, a second of the at least two parallel CT image data stream is based, at least in part, on the second CT image data. The CT output image data stream is based, at least in part, on respective outputs of the at least two parallel CT image data streams.

Deep neural network for CT metal artifact reduction

A deep neural network for metal artifact reduction is described. A method for computed tomography (CT) metal artifact reduction (MAR) includes generating, by a projection completion circuitry, an intermediate CT image data based, at least in part, on input CT projection data. The intermediate CT image data is configured to include relatively fewer artifacts than an uncorrected CT image reconstructed from the input CT projection data. The method further includes generating, by an artificial neural network (ANN), CT output image data based, at least in part, on the intermediate CT image data. The CT output image data is configured to include relatively fewer artifacts compared to the intermediate CT image data. The method may further include generating, by detail image circuitry, detail CT image data based, at least in part, on input CT image data. The CT output image data is generated based, at least in part, on the detail CT image data.

Methods and systems for stationary computed tomography are disclosed. The methods and systems include a gantry having alternating x-ray sources and x-ray detectors that are stationary during operation of the system. The gantry and pairs of x-ray sources and detectors substantially surrounds an object positioned inside the gantry during operation of the system. Dynamically adjustable collimators are positioned between the x-ray sources and the object. Each of the x-ray sources projects an x-ray beam through the collimators and through the object and the x-ray detectors receive the x-ray beam. The x-ray detectors include means for converting the x-ray beam to raw image data. One or more microprocessors control the x-ray sources and the process raw image data. A data storage device stores instructions, which upon execution by the microprocessor, control the x-ray sources and process the raw image data by converting the raw image data to a digital image.

Methods and systems for stationary computed tomography

A stationary in-vivo grating-enabled micro-CT (computed tomography) architecture (SIGMA) system includes CT scanner control circuitry and a number of imaging chains. Each imaging chain includes an x-ray source array, a phase grating, an analyzer grating and a detector array. Each imaging chain is stationary and each x-ray source array includes a plurality of x-ray source elements. Each imaging chain has a centerline, the centerlines of the number of imaging chains intersect at a center point and a first angle between the centerlines of a first adjacent pair of imaging chains equals a second angle between the centerlines of a second adjacent pair of imaging chains. A plurality of selected x-ray source elements of a first x-ray source array is configured to emit a plurality of x-ray beams in a multiplexing fashion.

Stationary in-vivo grating-enabled micro-CT architecture (sigma)

Tomographic/tomosynthetic image reconstruction systems and methods in the framework of machine learning, such as deep learning, are provided. A machine learning algorithm can be used to obtain an improved tomographic image from raw data, processed data, or a preliminarily reconstructed intermediate image for biomedical imaging or any other imaging purpose. In certain cases, a single, conventional, non-deep-learning algorithm can be used on raw imaging data to obtain an initial image, and then a deep learning algorithm can be used on the initial image to obtain a final reconstructed image. All machine learning methods and systems for tomographic image reconstruction are covered, except for use of a single shallow network (three layers or less) for image reconstruction.

Tomographic image reconstruction via machine learning

Systems and methods for determining an attenuation sinogram for a time-of-flight (TOF) positron emission tomography (PET) scan using only TOF PET data, and including use of the total amount of tracer provided to the subject of the TOF PET scan, are provided. The total amount of injected tracer can be used to determine the otherwise unknown constant shift present when an attenuation sinogram is estimated using the gradient of the attenuation sinogram. The attenuation sinogram can therefore be accurately and stably determined without any additional knowledge on the attenuation sinogram or map.

Attenuation map reconstruction from TOF PET data

Dynamic bowties, imaging systems including a bowtie, and methods of imaging including such bowties or systems are provided. A bowtie can be a three-dimensional (3-D) dynamic bowtie and can include a highly-attenuating bowtie (HB) and a weakly-attenuating bowtie (WB). The HB can be filled with a liquid, and the WB can be immersed in the liquid of the HB.

Dynamic bowtie filter for cone-beam/multi-slice CT

Advantageous methods, systems, and computer-readable media are provided. An attenuation map for an image can be reconstructed with no known sub-region of the attenuation map prior to performing the method. Gradient data of an attenuation sinogram can be estimated, and an algorithm can be performed to reconstruct the attenuation map from the estimated gradient data of the attenuation sinogram.

Growing community of inventors

Troy, NY, United States of America

Qingsong Yang