Abstract: A neural network is trained and implemented to simultaneously remove noise and artifacts from medical images using a Generalized noise and Artifact Reduction Network (“GARNET”) method for training a convolutional neural network (“CNN”) or other suitable neural network or machine learning algorithm. Noise and artifact realizations from phantom images are used to synthetically corrupt images for training. Corrupted and uncorrupted image pairs are used for training GARNET. Following the training phase, GARNET can be used to improve image quality of routine medical images by way of noise and artifact reduction.

Abstract: In accordance with some embodiments, systems, methods, and media for material decomposition and virtual monoenergetic imaging from multi-energy computed tomography data are provided. In some embodiments, a system comprises: at least one hardware processor configured to: receive a multi-energy computed tomography (MECT) image of a subject; provide the MECT data to a trained convolutional neural network (CNN); receive output from the trained CNN indicative of predicted material mass density for each of a plurality of materials at each pixel location of the MECT data, wherein the plurality of materials includes at least four materials; and generate a transformed version of the MECT data using the output.

Abstract: A system and method for determining stenosis severity in a subject's vasculature using multi-energy computer tomography (MECT) imaging is provided. In some aspects, the method includes acquiring MECT data using a CT system, performing a material decomposition process on acquired MECT data to generate one or more material density maps, and selecting, using the one or more material density maps, one or more regions of interest (ROIs) encompassing at least one vessel cross-section. The method also includes measuring an iodine content in the one or more ROIs, and determining a stenosis severity based on the measured iodine content. The method further includes generating a report indicating the stenosis severity associated with the subject's vasculature.

Abstract: Images are reconstructed from data acquired using an ultra-fast-pitch acquisition with a CT system. As an example, an ultra-fast-pitch acquisition mode in single-source helical CT (p?1.5) can be used to acquire data. A trained machine learning algorithm, such as a neural network, is used to reconstruct images in which artifacts associated with insufficient data acquired in the ultra-fast-pitch mode are reduced. An example neural network can include customized functional modules using both local and non-local operators, as well as the z-coordinate of each image, to effectively suppress the location- and structure-dependent artifacts induced by the ultra-fast-pitch mode. The machine learning algorithm can be trained using a customized loss function that involves image-gradient-correlation loss and feature reconstruction loss.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: A system and method is provided for performing material decomposition using a computed tomography (CT) system. The method includes acquiring CT imaging data of an object including data subsets corresponding to at least two different energy spectral bins and using the CT imaging data at each of the at least two different energy spectral bins to form a series of equations for basis material decomposition. The method also includes using a general physical constraint, which quantifies how each basis material in the object is mixed together to form the object, within the series of equations. The method also includes determining at least one basis material density of the object using the physical constraint and the CT imaging data and generating an image of the object using the CT imaging data and the mass densities of at least one basis material.

Abstract: A fully image-based framework for CT image, or other medical image, quality evaluation and virtual clinical trial using deep-learning techniques is provided. This framework includes deep learning-based noise insertion, lesion insertion, and model observer, which enable efficient, objective, and quantitative image quality evaluation and virtual clinical trial directly performed on patient images.

Abstract: System and methods are provided for producing computed tomography (CT) images. In some aspects, a method includes obtaining medical image data sets acquired using the multiple energies of irradiating radiation and analyzing the medical image data sets for spatial and spectral features. The method also includes comparing the spatial and spectral features of the medical image data sets to identify similarities and using the similarities, weighting the medical image data sets to generate images of the subject having reduced noise compared to images of the subject produced from the medical image data sets without weighting.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: Systems and methods are provided for synthesizing information from multiple image series of different kernels into a single image series, and also for converting a single baseline image series of a kernel reconstructed by a CT scanner to image series of various other kernels, using deep-learning based methods. For multi-kernel synthesis, a single set of images with desired high spatial resolution and low image noise can be synthesized from multiple image series of different kernels. The synthesized kernel is sufficient for a wide variety of clinical tasks, even in circumstances that would otherwise require many separate image sets. Kernel conversion may be configured to generate images with arbitrary reconstruction kernels from a single baseline kernel. This would reduce the burden on the CT scanner and the archival system, and greatly simplify the clinical workflow.

Abstract: A system and method is provided for high fidelity multi-energy CT processing. This system and method exploits prior knowledge, where prior knowledge may include redundant information existing in the CT images, such as spatial redundancy between a thick slice and a thin slice encompassed by or close to the thick slice, or the spatiospectral redundancy between the image output of multi-energy CT processing and the source multi-energy CT images. The system and method retains structural details, spatial resolution, spectral fidelity, and noise texture while achieving noise reduction. The method reduces image noise and increases the contrast-to-noise ratio in processed images, while simultaneously maintaining image details and natural appearance of the image to enhance detectability and facilitate reader acceptance.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: A fully image-based framework for CT image, or other medical image, quality evaluation and virtual clinical trial using deep-learning techniques is provided. This framework includes deep learning-based noise insertion, lesion insertion, and model observer, which enable efficient, objective, and quantitative image quality evaluation and virtual clinical trial directly performed on patient images.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: A system and method is provided for performing material decomposition using a computed tomography (CT) system. The method includes acquiring CT imaging data of an object including data subsets corresponding to at least two different energy spectral bins and using the CT imaging data at each of the at least two different energy spectral bins to form a series of equations for basis material decomposition. The method also includes using a general physical constraint, which quantifies how each basis material in the object is mixed together to form the object, within the series of equations. The method also includes determining at least one basis material density of the object using the physical constraint and the CT imaging data and generating an image of the object using the CT imaging data and the mass densities of at least one basis material.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: A system and method for determining stenosis severity in a subject's vasculature using multi-energy computer tomography (MECT) imaging is provided. In some aspects, the method includes acquiring MECT data using a CT system, performing a material decomposition process on acquired MECT data to generate one or more material density maps, and selecting, using the one or more material density maps, one or more regions of interest (ROIs) encompassing at least one vessel cross-section. The method also includes measuring an iodine content in the one or more ROIs, and determining a stenosis severity based on the measured iodine content. The method further includes generating a report indicating the stenosis severity associated with the subject's vasculature.

Abstract: System and methods are provided for producing computed tomography (CT) images. In some aspects, a method includes obtaining medical image data sets acquired using the multiple energies of irradiating radiation and analyzing the medical image data sets for spatial and spectral features. The method also includes comparing the spatial and spectral features of the medical image data sets to identify similarities and using the similarities, weighting the medical image data sets to generate images of the subject having reduced noise compared to images of the subject produced from the medical image data sets without weighting.

Abstract: Described here are systems and methods for optimization techniques for automatically selecting x-ray beam spectra, energy threshold, energy bin settings, and other imaging technique parameters for photon-counting detector computed tomography (“PCCT”). The techniques described here are generally based on subject or object size, material of interest, and location of the target material. Advantageously, the optimizations can be integrated with different PCCT systems to automatically select optimal imaging technique parameters before scanning a particular subject or object.

Abstract: A method for creating an energy series of images acquired using a multi-energy computed tomography (CT) imaging system having a plurality of energy bins includes acquiring, with the multi-energy CT imaging system, a series of energy data sets, where each energy data set is associated with at least one of the energy bins. The method includes producing a conglomerate image using at least a plurality of the energy data sets and, using the conglomerate image, reconstructing an energy series of images, each image in the energy series of images corresponding to at least one of the energy data sets.