Correction for polychromatic X-ray distortion in CT images

• US 4,217,641 A
• Filed: 04/28/1978
• Issued: 08/12/1980
• Est. Priority Date: 04/28/1978
• Status: Expired due to Term

• Current Assignee: US Philips Corporation (Koninklijke Philips N.V.)
• Original Assignee: US Philips Corporation (Koninklijke Philips N.V.)
• Inventors: Naparstek, Abraham
• Primary Examiner(s): Krass, Errol A.
• Application Number:
  US05/901,212
• Time in Patent Office: 837 Days
• Field of Search: 364/414, 364/571, 364/415, 364/515, 364/527, 250/445 T, 250/358 R, 250/360, 358/160, 358/166
• US Class Current: 382/131
• CPC Class Codes: A61B 6/032 Transmission computed tomography [CT]
• CPC Class Codes: G06T 11/005 Specific pre-processing for tomographic reconstruction, e.g. calibration, source positioning, rebinning, scatter correction, retrospective gating
• CPC Class Codes: Y10S 378/901 Computer tomography program or processor

Abstract

A method and apparatus which corrects the polychromatic distortion of CT images which is produced by the non-linear interaction of body constituents with a polychromatic X-ray beam. A CT image is processed to estimate the proportion of the attenuation coefficients of the constituents in each pixel element. A multiplicity of projections for each constituent are generated from the original image and are combined utilizing a multidimensional polynomial which approximates the non-linear interaction involved. An error image is then generated from the combined projections and is subtracted from the original image to correct for the poloychromatic distortion.

Claims

1. A computerized tomographic system for measuring characteristics of and imaging, body tissue comprising:

scanning means which generate projections of said tissue by directing a known spectrum of radiation through said body and measuring the spatial intensity distribution of radiation traversing said tissue along a multiplicity of paths;

image generating means which calculate the attenuation of each pixel element of a raw image array, which is representative of a tranverse section of said tissue, from said projections;

means which analyze values in said raw image array and assign to each pixel element, on the basis of those values, a specific proportion of the attenuation coefficient to each of two or more constituents of said tissue;

means which combine said element proportions to generate a multiplicity of constituent projections for each of said constituents of said tissue;

means which combine said constituent projections, in accordance with a predetermined function approximating the non-linear interaction of said constituents with said spectral intensity distribution, to generate a multiplicity of error projections of said tissue;

means which combine said error projections to generate an error image array which is representative of the polychromatic aberration in said raw image array; and

means for subtracting said error image array from said raw image array to generate a compensated image array.

2. The system of claim 1 wherein said predetermined function is a multidimensional polynomial.

3. The system of claim 1 wherein said constituents are bone and soft tissue and said function is a two-dimensional cubic.

4. The system of claim 1 wherein said means which analyze and assign function by thresholding and interpolating the values of said elements relative to the thresholds.

5. The system of claim 1 further including means for filtering said error projections.

6. A system for compensating polychromatic distortion in a computed tomographic image of body tissue comprising:

means for generating an error image wherein the numerical value of each image element is representative of the polychromatic distortion in a corresponding element of said tomographic image which means include means for producing constituent projections of each of at least two constituents of said tissue from said tomographic image and means for combining said constituent projections to generate said error image and

means for subtracting said error image, element by element, from said tomographic image.

7. The system of claim 6 wherein said means for producing constituent projections include means for assigning a proportion of each element of said tomographic image to each of said constituents.

8. The system of claim 7 wherein said means for assigning a proportion compare the value of each element with predetermined threshold values and assign said proportions on the basis of said comparison.

9. The system of claim 8 wherein said means for assigning determined said proportion by interpolating each element value between threshold values.

10. The system of claim 7 wherein said means for producing constituent projections include, for each of said projections, means for associating each of said elements with one of a multiplicity of image strips and

means for summing each of the constituent proportion values for all elements associated with each strip whereby said proportion values are projected along said strips.

11. The system of claim 6 wherein said means for combining constituent projections functions to combine said constituent projections in accordance with a multidimensional polynomial the coefficients of which are determined from a least-squares approximation of the interaction of radiation, having the same spectral content as was used to measure said tomographic image, with the spectral radiation absorption characteristics of said constituents.

12. The system of claim 11 wherein said constituents are bone and soft tissue and said multidimensional polynomial is a two-dimensional cubic.

13. The system of claim 11 further including filter means which function to filter the output of said means for combining said constituent projections.

14. A method for measuring characteristics of, and imaging, body tissue comprising the steps of:

directing a known spectrum of radiation through said tissue;

measuring the spatial intensity distribution of said radiation traversing said tissue along a multiplicity of paths;

generating a multiplicity of projections of said tissue from said spatial intensity distribution;

calculating, from said projections, an attentuation value for each element of a raw image, said raw image being representative of a transverse section of said tissue and each element of said raw image corresponding to a distinct portion of said tissue;

for each of said elements of said raw image, comparing the associated attenuation value with a threshold value and, on the basis of the result of said comparison, assigning a proportion of the associated attenuation value to each of two or more constituents of the tissue;

for each of said constituents of the tissue, combining the associated proportions of the attenuation values associated with all of said elements to generate a multiplicity of constituent projections of said tissue;

combining said constituent projections, in accordance with a predetermined function approximating the non-linear interaction of said constituents with said known spectrum of radiation, to generate a multiplicity of error projections of said tissue;

combining said error projections to generate an error image which is representative of the polychromatic aberration in said raw image; and

subtracting said error image from said raw image to generate a compensated image.

15. The method of claim 14 wherein said predetermined function is a multidimensional polynomial.

16. The method of claim 14 wherein said constituents are bone and soft tissue and said function is a two-dimensional cubic.