Back to results
US 8,045,774 B2Granted
Method for creating material-selective volume images
AssigneeSiemens Aktiengesellschaft
InventorsWolfgang Härer, Ernst-Peter Rührnschopf, Bernhard Scholz
Life of this patent
Estimating the term…
Abstract
A method for creating material-selective volume images of various material components of an object to be examined is proposed. The method is based on multi-spectral projected images captured from various directions of projection, the images being captured using an X-ray machine, and makes it possible to quantitively exactly determine the material-selective volume images by way of iteration.
Claims19
A method for creating material-selective volume images, comprising:
- generating radiation in different energy fields using a radiation source;
- x-raying an object, composed of various material components, in different energy fields and from different directions of projection;
- loading a detector with the radiation and capturing projected images in different energy fields by way of the detector device;
- creating material-selective volume images via an evaluation unit connected downstream of the detector;
- creating a volume image of the object using a series of projected images captured from different directions of projection;
- segmenting a volume image of at least one secondary component out of the volume image in addition to a number of main components, which corresponds to at most the number of energy fields;
- creating mass occupancy partial images linked to the at least one secondary component in the different energy fields by back projection of the volume image of the at least one secondary component in the different energy fields;
- producing corrected projected images in different energy fields from the captured projected images by removing fractions corresponding to the mass occupancy partial images;
- creating mass occupancy partial images of the main components using the corrected projected images by inversion of a multi-dimensional attenuation function;
- and
- creating material-selective volume images of the main components using the mass occupancy partial images.
8 more elements
Dependencies are inferred from the claim text.