Method for artifact reduction using monoenergetic data in computed tomography

Abstract

A method for artifact correction in computed tomography, the method comprising: (1) acquiring a plurality of data sets associated with different X-ray energies (i.e., D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); (2) generating a plurality of preliminary images from the different energy data sets acquired in Step (1) (i.e., I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n); (3) using a mathematical function to operate on the preliminary images generated in Step (2) to identify the sources of the image artifact (i.e., the artifact source image, or ASI, where ASI=f(I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n)); (4) forward projecting the ASI to produce ASD=fp(ASI); (5) selecting and combining the original data sets D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n in order to produce a new subset of the data associated with the artifact, whereby to produce the artifact reduced data, or ARD, where ARD=f(ASD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); (6) generating a repaired data set (RpD) to keep low-energy data in artifact-free data and introduce high-energy data in regions impacted by the artifact, where RpD=f(ARD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); and (7) generating a final reduced artifact image (RAI) from the repaired data, RAI=bp(RpD), where the function bp is any function which generates an image from data.

Claims

1. A method for image improvement in computed tomography, the method comprising: (1) acquiring a plurality of data sets associated with different energies of radiation following ray-driven paths (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); (2) generating a plurality of preliminary images from the different energy data sets acquired in Step (1) (I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n); (3) using a mathematical function to operate on the preliminary images generated in Step (2) to identify sources of image degradation (the degraded source image, or ASI, where ASI=f(I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n)); (4) forward projecting the ASI to produce artifact source data, or ASD, where ASD=fp(ASI); (5) selecting and combining the plurality of data sets associated with different energies of radiation following ray-driven paths acquired in Step (1) (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n) in order to produce a new subset of the data associated with the image degradation, whereby to produce degradation reduced data, or ARD, where ARD=f(ASD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); (6) generating a repaired data set (RpD) to keep low-energy data in degradation-free data and introduce high-energy data in regions impacted by the degradation, where RpD=f(ARD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n); and (7) generating a final reduced degradation image (RAI) from the repaired data, RAI=bp(RpD), where the function bp is any function which generates an image from data.

2. A method according to claim 1 wherein the plurality of data sets associated with different energies of radiation following ray-driven paths (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n) are monoenergetic.

3. A method according to claim 1 wherein the plurality of data sets associated with different energies of radiation following ray-driven paths (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n) are polyenergetic with different energy distributions.

4. A method according to claim 3 wherein the plurality of data sets associated with different energies of radiation following ray-driven paths (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n) are polyenergetic with polychromatic high-, mid- and low-energy energy distributions.

5. A method according to claim 1 wherein the mathematical function used in Step (3) produces a binary image locating regions of degradation-generating objects.

6. A method according to claim 1 wherein the mathematical function used in Step (3) produces a measure of the degree of degradation impact where stronger values occur where an object has larger negative impact.

7. A method according to claim 1 wherein image degradation is produced by a point in an object having high radiation attenuation.

8. A method according to claim 7 wherein the point in the object comprises a metal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) These and other objects and features of the present invention will be more fully disclosed or rendered obvious by the following detailed description of the preferred embodiments of the invention, which is to be considered together with the accompanying drawings wherein like numbers refer to like parts, and further wherein:

(2) FIG. 1 is a schematic view showing typical elements of a CT, including an X-ray source, X-ray(s), a rotational center, an object of interest to be scanned, an array of x-ray detectors, a direction of rotation and a rotating gantry;

(3) FIG. 2 is a schematic view showing the locations of an X-ray source and a single detector as the source, with the detector being rotated about the object of interest through a series of angular positions 1-9 (source locations are denoted with an “s”, detector with a “d”);

(4) FIG. 3 is a graph showing a plot of x-ray attenuation verses angular position for the object and ray geometry shown in FIG. 2, wherein circles represent values for an ideal system, and x's represent a system where saturation effects are present;

(5) FIG. 4 is a schematic view showing an object shape which is consistent with the saturated data model of FIG. 3;

(6) FIG. 5 is a schematic view showing the object of FIG. 2 with extreme beam hardening artifacts;

(7) FIGS. 6A-6H are schematic views illustrating a process for metal identification and mixing of high-energy data into data regions which have been impacted by metal in the form of a flowchart which is to be read from top to bottom; and

(8) FIGS. 7A-7H are schematic views showing a generalized process for reducing artifacts.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(9) Metal Artifact Correction Using Monoenergetic Data

(10) In the preferred embodiment of the present invention, several monoenergetic data sets (either approximation-derived, experimentally-measured or synthetically-derived) are used together to minimize artifacts and maximize image quality. More particularly, an object imaged with a low-energy monochromatic X-ray source demonstrates a high degree of contrast between radiologically opaque and transparent objects, while an object imaged with a high-energy monochromatic X-ray source shows relatively low contrast between materials. This is anecdotally explained by the fact that high-energy X-rays tend to simply pass through all materials, whereas low-energy X-rays are more likely to interact with the material via absorption or scattering. Likewise, high-energy X-rays are less likely to be impeded by metal and the images generated from high-energy X-ray data suffer less from metal artifacts.

(11) The fundamental idea is to combine the high-energy data and the low-energy data in such a way so as to maximize the benefits of both. In regions of data where metal has not impacted measurements, low-energy data is preferred because of its high contrast characteristics, whereas in regions impacted by metal, high-energy data is preferred because of its inherent resistance to artifacts.

(12) In a CT system which is capable of generating monoenergetic images, low-energy images are generally more desirable because of their high contrast. However, these low-energy monoenergetic images are also more susceptible to metal artifacts. Therefore, the present invention is used to generate a composite image which utilizes low-energy monoenergetic data for low attenuation regions of the object being scanned and high-energy monoenergetic data for high attenuation regions of the object being scanned.

(13) More particularly, with the present invention, “metal” is first identified in a preliminary image. This image may be generated using any kind of X-ray source, e.g., polyenergetic, monoenergetic, high- or low-energy, etc. Note that for purposes of the present invention, the term “metal” is used to describe any material which generates artifacts and may include particularly dense materials, or materials with unusual attenuation properties. In one simple embodiment, the metal in the preliminary CT image is identified by subjecting the image to a threshold attenuation value, where pixels having an attenuation value higher than this threshold are considered to be metal. All other pixels containing values associated with a non-metal material are set to zero.

(14) The pixels within the image identified to contain metal are then forward projected so as to identify which of the “raw” detector readings have been affected by metal. In the preferred embodiment, only those pixels which are identified as containing metal are forward projected. Note that while there may be other advantages to forward projecting a complete image, forward projecting only those portions of the image impacted by metal leads to a computationally efficient process. The forward projection of the metal-containing portions of the image produces a simulated data set where the value at each data location is related to the amount of metal observed along a line connecting the X-ray source to the sensor, and the forward projection is a measure of the degree of impact the metal has on the data along that line. For purposes of the present invention, the forward projection of the identified metal may be referred to as the map of the metal-impacted data.

(15) In one embodiment, the data value associated with each detector is corrected according to the amount of metal intersecting the line of response connecting the X-ray source to the X-ray detector. In the preferred embodiment, the magnitude of the correction is abandoned for a binary decision where each data point is analyzed to determine whether there has been any metal interference in the path connecting the X-ray source to the X-ray detector. The practical implementation of this preferred embodiment is to qualify the simulated data set by a threshold attenuation value so as to produce a binary identifier describing which data points have been impacted by metal (which may be referred to as “metal-impacted data”). These data, negatively impacted by the presence of metal, need to be repaired in such a way so as to reduce metal artifacts while remaining geometrically consistent with the objects being reconstructed.

(16) The quality of the data identified within the map of metal-impacted data are less negatively influenced in the high-energy X-ray data than in the low-energy X-ray data. In the preferred embodiment, the metal-impacted areas of the low-energy data set are repaired by “mixing-in” a fixed fraction of the high-energy data set. By restricting the data repair of the low-energy data set to areas known to be impacted by metal, the desirable high contrast traits of the low-energy image are retained where appropriate while gaining the high-energy ability to image through metal in the appropriate regions.

(17) Thus, with the present invention, and referring now to FIGS. 6A-6H, metal artifact correction of an image of an object (FIG. 6A) may be affected by:

(18) (1) scanning the object with a high-energy monoenergetic X-ray source and a low-energy monoenergetic X-ray source so as to create two monoenergetic data sets (FIG. 6B);

(19) (2) generating a high-energy monoenergetic image and a low-energy monoenergetic image (FIG. 6C);

(20) (3) using a pixel threshold to segregate the high-energy monoenergetic image and the low-energy monoenergetic image into metal and non-metal portions (FIG. 6D);

(21) (4) forward projecting the regions of data associated with the metal so as to produce a map of metal-impacted data (FIG. 6E);

(22) (5) using the map of metal-impacted data to select and combine elements of the high-energy monoenergetic image and the low-energy monoenergetic image to produce a new subset of image data for the regions associated with the artifact (FIG. 6F);

(23) (6) integrating the “repaired data” identified in Step (5) above with the image data for the regions which are not metal-impacted so as to create a complete data set (FIG. 6G); and

(24) (7) using the complete data set to produce a new image which has reduced artifact impact (FIG. 6H).

(25) Metal Artifact Correction Using Monoenergetic and/or Polyenergetic Data (“Generalized Correction”)

(26) In the foregoing discussion, metal artifact correction was effected using two sets of monoenergetic data (i.e., high-energy monoenergetic data and low-energy monoenergetic data). However, it is also possible to effect metal artifact correction using polyenergetic data of two different energy levels.

(27) More particularly, in this form of the invention, a material with anomalous spectral properties (e.g., highly attenuating metals) is first identified. In the most generalized correction, the initial material identification can be performed by looking at how regions of the material behave across a set of monochromatic images or by identifying objects with a common set of X-ray absorption spectra. In the preferred embodiment discussed above, this method was simplified to any image regions with high attenuation. In the more generalized case, one could, for instance, preferentially identify biological tissues with spectral properties similar to lung tissue or very dense bone. This is achieved by applying a mathematical function to the set of available monochromatic images (or to several polychromatic sets) in order to select the areas within the image with the desired spectral properties.

(28) Once the regions within the image are identified as containing metal, the “metal only” regions are forward projected onto the data space, thereby quantifying the degree of impact. In the most generalized case, the forward projection of a selected material is used to quantify the total contribution of that material to the measurement at a given detector. This parameter can be used to selectively combine the monoenergetic data sets (or polychromatic data sets) into a data set with optimal properties for the selected material. For example, the forward projection of lung tissue identifies which detectors are used in measuring lung tissue. For those lung tissue data channels, a function can be used to construct an optimal data set from the set of monochromatic data sets (or polychromatic data sets) for imaging lungs. Likewise, data associated with the imaging of bone or metal can be constructed to minimize artifacts. In all cases, data for every detector channel is optimized for the materials sampled by that detector.

(29) Thus, with the present invention, and referring now to FIGS. 7A-7H, metal artifact correction of an image of an object (FIG. 7A) can be generalized to the following steps:

(30) (1) Generating or acquiring any number of data sets associated with different energies (D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n). These data sets can be monoenergetic, or polyenergetic with different energy distributions (i.e., polychromatic high-, mid- and low-energy). See FIG. 7B.

(31) (2) Generating any number of preliminary images from the different energy data sets described in Step (1) (I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n). See FIG. 7C.

(32) (3) Using a mathematical function which operates on the preliminary images of Step (2) to identify the sources of the image artifact. This function may produce a binary image (locating regions of artifact-generating objects) or a measure of the degree of artifact impact, i.e., stronger values where the object has larger negative impact. This may be referred to as the artifact source image, or ASI, where ASI=f(I.sub.1, I.sub.2, I.sub.3 . . . I.sub.n). See FIG. 7D.

(33) (4) Using a mathematical function which transposes an image into model data (i.e., a forward projector). The regions of data associated with the artifact sources are known as the artifact source data, or ASD. The artifact source data is the forward projection of the ASI, or ASD=fp(ASI). See FIG. 7E.

(34) (5) Using the artifact source data set to select and combine the original data sets D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n in order to produce a new subset of the data associated with the artifact. This is known as the artifact reduced data, or ARD, where ARD=f(ASD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n). See FIG. 7F.

(35) (6) Constructing the repaired data (RpD) into a data set in such a way as to optimize positive image characteristics (i.e., to keep low-energy data in artifact-free data and introduce high-energy data in regions impacted by metal). This is generically described as a function RpD=f(ARD, D.sub.1, D.sub.2, D.sub.3 . . . D.sub.n). See FIG. 7G.

(36) (7) Reconstructing the final reduced artifact image (RAI) from the repaired data, or RAI=bp(RpD). Here the function bp is generically any function which generates an image from data. See FIG. 7H.

MODIFICATIONS

(37) It will be appreciated that still further embodiments of the present invention will be apparent to those skilled in the art in view of the present disclosure. It is to be understood that the present invention is by no means limited to the particular constructions herein disclosed and/or shown in the drawings, but also comprises any modifications or equivalents within the scope of the invention.