Canon Medical Systems Corporation

Toshiba Medical Systems Corporation

University of Chicago

Yan Liu

Zhou Yu

Jian Zhou

Ruoqiao Zhang

Patrick Jean La Riviere

Dimple Modgil

A deep learning (DL) network corrects/performs sinogram completion in computed tomography (CT) images based on complementary high- and low-kV projection data generated from a sparse (or fast) kilo-voltage (kV)-switching CT scan. The DL network is trained using inputs and targets, which respectively generated with and without kV switching. Another DL network can be trained to correct sinogram-completion errors in the projection data after a basis/material decomposition. A third DL network can be trained to correct sinogram-completion errors in reconstructed images based on the kV-switching projection data. Performance of the DL network can be improved by dividing a 3D convolutional neural network (CNN) into two steps performed by respective 2D CNNs. Further, the projection data and DLL can be divided into high- and low-frequency components to improve performance.

Apparatus and method that uses deep learning to correct computed tomography (CT) with sinogram completion of projection data

A medical image processing apparatus according to an embodiment comprises a memory and processing circuitry. The memory is configured to store a plurality of neural networks corresponding to a plurality of imaging target sites, respectively, the neural networks each including an input layer, an output layer, and an intermediate layer between the input layer and the output layer, and each generated through learning processing with multiple data sets acquired for the corresponding imaging target site. The processing circuitry is configured to process first data into second data using, among the neural networks, the neural network corresponding to the imaging target site for the first data, wherein the first data is input to the input layer and the second data is output from the output layer.

Medical image processing apparatus having a plurality of neural networks corresponding to different fields of view

A method and apparatus is provided that uses a deep learning (DL) network to reduce noise and artifacts in reconstructed medical images, such as images generated using computed tomography, positron emission tomography, and magnetic resonance imaging. The DL network can operate either on pre-reconstruction data or on a reconstructed image. The DL network can be an artificial neural network or a convolutional neural network (e.g., using a three-channel volumetric kernel architecture). Different neural networks can be trained depending on the noise level, scanning protocol, or the anatomic, diagnostic or clinical objective of the reconstructed image (e.g., by partitioning the training data into noise-level range and training respective DL networks for each range). Further, the DL networks can be trained to mitigate artifacts, such as the cone-beam artifact.

Apparatus and method for medical image reconstruction using deep learning for computed tomography (CT) image noise and artifacts reduction

An apparatus and method are described using a forward model to correct pulse pileup in spectrally resolved X-ray projection data from photon-counting detectors (PCDs). To calibrate the forward model, which represents each order of pileup using a respective pileup response matrix (PRM), an optimization search determines the elements of the PRMs that optimize an objective function measuring agreement between the spectra of recorded counts affected by pulse pileup and the estimated counts generated using forward model of pulse pileup. The spectrum of the recorded counts in the projection data is corrected using the calibrated forward model, by determining an argument value that optimizes the objective function, the argument being either a corrected X-ray spectrum or the projection lengths of a material decomposition. Images for material components of the material decomposition are then reconstructed using the corrected projection data.

Method and apparatus for computed tomography (CT) and material decomposition with pile-up correction calibrated using real pulse pileup effect and detector response

A method and apparatus uses multi-material decomposition of three or more material components to generate material-component images from spectral images reconstructed from spectral computed tomography data. In three-component material decomposition e.g., the Mendon&#xe7;a method is used for multi-material decomposition when the attenuation values satisfy an assumed volume fraction condition (i.e., for a given voxel, the attenuation values are within a triangle having vertices given by unit volume fractions of three respective material components). However, when the volume fraction condition fails (e.g., the attenuation values are outside the triangle), either a shortest-Hausdorff-distance method or a closest-edge method is used for multi-material decomposition. For example, the attenuation values of the voxel are projected onto a lower-dimensional space (e.g., the space of a closest edge) and decomposed into a pair/single material component(s) of the lower-dimensional space.

Method and apparatus for spectral computed tomography (CT) with multi-material decomposition into three or more material components

A deep learning (DL) network reduces artifacts in computed tomography (CT) images based on complementary sparse-view projection data generated from a sparse kilo-voltage peak (kVp)-switching CT scan. The DL network is trained using input images exhibiting artifacts and target images exhibiting little to no artifacts. Another DL network can be trained to perform image-domain material decomposition of the artifact-mitigated images by being trained using target images in which beam hardening is corrected and spatial variations in the X-ray beam are accounted for. Further, material decomposition and artifact mitigation can be integrated in a single DL network that is trained using as inputs reconstructed images having artifacts and as targets material images without artifacts with beam-hardening corrections, etc. Further, the target material images can be transformed using a whitening transform to decorrelate noise.

Apparatus and method for dual-energy computed tomography (CT) image reconstruction using sparse kVp-switching and deep learning

Medical image processing apparatus and medical image processing system

A method and apparatus is provided to denoise computed tomography (CT) sinograms and reconstructed images by minimizing a penalized-weighted-least-squares objective function that includes a joint-edge-preserving regularization term. Common information shared among a series of related image and/or sinograms can beneficially be used to more accurately identify edges common to all of the image and/or sinograms. Thus, a joint edge preserving regularizer that relies on information from a series of three-dimensional constituents of a four-dimensional image, such as a time series of images in a CT profusion study or an energy series of images in spectral CT, can improve image quality by better preserving edges while simultaneously denoising the constituent images.

Apparatus and method for joint-edge-preserving regularization to reduce noise in four-dimensional computed tomography images

A method and apparatus is provided to denoise material-decomposition data generated using projection data from a spectral computed tomography scanner. A whitening transform is used to transform the material-decomposition data into uncorrelated components and perform denoising on the uncorrelated components. Using different denoising parameters for the various uncorrelated components, a flattening can be achieved for the standard deviation of the noise as a function of X-ray energy, which can be determined using mono-energetic images derived from the material-decomposition data. The whitening transformation and the denoising can be applied the material-decomposition sinograms and/or to material-decomposition images reconstructed from the material-decomposition sinograms.

Growing community of inventors

Vernon Hills, IL, United States of America

Yan Liu