Automatic method of material identification for computed tomography
| DWPI Title: Automatic method of material identification for computed tomography by creating profile for pixels belong to spectral radiographic image, clustering pixels, and assigning labels to defined clusters, where image is stored in digital medium |
| Abstract: A method is provided for isolating and labeling discrete features in a spectral radiographic image recorded as a set of images in different energy channels. The disclosed method involves creating a profile for each of at least some pixels in the spectral radiographic image. The profiles are sequences of pixel values, in which each pixel value is a photon count or a similar radiographic exposure value indicative of the attenuation of a portion of the scanning beam in a respective energy channel. Iterative hierarchical clustering is used to cluster the pixels on the basis of their respective profiles. Labels are assigned to one or more of the resulting clusters. In implementations, each label can be associated with an inferred material composition or with an inference that the material composition is unknown. |
| Use: The method is used for automatic of material identification for computed tomography, and used for isolating and labeling discrete features in a spectral radiographic image recorded as a set of images in different energy channels. |
| Advantage: The method combines techniques of machine learning with high-energy spectral x-ray computed tomography data to identify material compositions within each reconstructed voxel in the field-of-view of a CT apparatus. The method is not required to analyze attenuation based on the Lambert-Beer law, and identifies material signatures using unsupervised clustering procedures. The method compares scan of unidentified materials with a scan of previously identified materials. |
| Novelty: The method includes (a) creating a profile for each pixel in pixels that belong to a spectral radiographic image, (b) clustering the pixels, where the clustering is based on the pixel profiles, and by them one or more clusters are defined, and (c) assigning labels to one or more of the defined clusters, by them one or more labeled clusters are defined, where: each of the profiles comprises a vector of radiographic exposure values taken in different energy channels; the spectral radiographic image is stored in a digital medium accessible by a computer; the profiles are stored in a digital medium accessible by a computer; and the clustering comprises iterative hierarchical clustering (HC) in which each of one or more iterations of an HC procedure eliminates one or more clusters from a subsequent iteration of the HC procedure. |
| Filed: 6/28/2019 |
| Application Number: US16456235A |
| Tech ID: SD 14601.0 |
| This invention was made with Government support under Contract No. DE-NA0003525 awarded by the United States Department of Energy/National Nuclear Security Administration. The Government has certain rights in the invention. |
| Data from Derwent World Patents Index, provided by Clarivate All rights reserved. Republication or redistribution of Clarivate content, including by framing or similar means, is prohibited without the prior written consent of Clarivate. Clarivate and its logo, as well as all other trademarks used herein are trademarks of their respective owners and used under license. |