Multiscale adaptive system for enhancement of an image in X-ray angiography

• US 5,963,676 A
• Filed: 02/07/1997
• Issued: 10/05/1999
• Est. Priority Date: 02/07/1997
• Status: Expired due to Term

• Current Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Original Assignee: Siemens Corporate Research Incorporated (Siemens AG)
• Inventors: Wu, Zhenyu, Fang, Ming
• Primary Examiner(s): Moore, David K.
• Assistant Examiner(s): BRINICH, STEPHEN M
• Application Number:
  US08/796,331
• Time in Patent Office: 970 Days
• Field of Search: 382/128, 382/132-133, 382/274, 382/263, 382/284
• US Class Current: 382/274
• CPC Class Codes: G06T 2207/10116 X-ray image
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 2207/30101 Blood vessel; Artery; Vein; Vascular
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/92 based on global image properties

Abstract

A multi-scale approach to enhancement is taken where enhancement masks are generated at different scales and then combined together using a pyramid scheme. Each mask is computed from applying directional sensitive Laplacian kernels, which is responsible for extracting images of low contrast, followed by an adaptive non-linear mapping. This non-linearity is crucial in virtually eliminating edge overshoots. Furthermore, the input image may also be preprocessed by an edge-preserving smoothing filter to reduce the noise effect.

Claims

1. A multiscale adaptive system for enhancement of an image in x-ray angiography comprising:

an image acquisition system for acquiring an input image;

an image contrast enhancement system connected to said image acquisition system for utilizing a multiscale approach to enhancement;

a display mapping system connected to said image contrast enhancement system for modifying pixel values such that relevant image information is presented in an optimal way; and

,

an image display apparatus connected to said display mapping system for presenting an enhanced image;

wherein said image contrast enhancement system comprises;

a pyramid system wherein a pyramid of intensity images are generated at pyramid levels from said input image;

an enhancement mask generator wherein enhancement masks of said intensity images are generated at each of said pyramid levels;

a combination system wherein said enhancement masks are combined together to form a single composite mask; and

,

an enhanced image generator wherein said single composite mask is combined with said input image to form an enhanced image.

2. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 1 wherein said combination system comprises:

transformation means for transforming mask at a coarsest scale to scale of its parent image;

weighting means connected to said transformation means;

adding means for adding said weighted transformed mask to mask of said parent image repeatably to form said single composite mask that is at full resolution.

3. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 2 wherein said transformation means comprises:

wavelet synthesis means.

4. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 2 wherein said enhancement mask generator comprises:

a first computation means for computing a directional sensitive Laplacian of said intensity images; and

,

a second computation means for computing a nonlinear mapping of said directional sensitive Laplacian.

5. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 4 wherein:

said first computation means utilizes small 3×

3 kernels for computing said directional sensitive Laplacian.

6. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 4 wherein:

said first computation means utilizes four kernels which are tuned respectively to 0 degrees, 45 degrees, 90 degrees and 135 degrees to extract line-like fine structures.

7. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 4 further comprising:

Gaussian smoothing means for applying Gaussian smoothing to said intensity images before said intensity image enters said first computation means.

8. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 4 wherein:

said second computation means is represented by;

##EQU2## where I, I_lap, M denote said input intensity image, said directional sensitive Laplacian and said enhancement masks, respectively, M(x,y) denotes the mask value at a pixel (x,y), λ

is a positive gain factor and α and

β

are control parameters of non-linearity.

9. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 4 wherein;

said second computation means virtually eliminates positive edge overshoots and greatly reduces negative overshoots.

10. A multiscale adaptive system for enhancement of blood vessels in x-ray angiography as claimed in claim 1 further comprising:

an edge preserving smoothing filter for preprocessing said input image to reduce noise effect.

11. A multiscale adaptive system for enhancement of an image in x-ray angiography comprising:

an image contrast enhancement system for utilizing a multiscale approach to enhancement wherein said image contrast enhancement system comprises;

a pyramid system wherein a pyramid of intensity images are generated at pyramid levels from an input image;

an enhancement mask generator wherein enhancement masks of said intensity images are generated at each of said pyramid levels;

a combination system wherein said enhancement masks are combined together to form a single composite mask; and

,

an enhanced image generator wherein said single composite mask is combined with said input image to form an enhanced image.

12. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 11 wherein said combination system comprises:

transformation means for transforming mask at a coarsest scale to scale of its parent image;

weighting means connected to said transformation means;

adding means for adding said weighted transformed mask to mask of said parent image repeatably to form said single composite mask that is at full resolution.

13. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 12 wherein said transformation means comprises:

wavelet synthesis means.

14. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 11 wherein said enhancement mask generator comprises:

a first computation means for computing a directional sensitive Laplacian of said intensity images; and

,

a second computation means for computing a nonlinear mapping of said directional sensitive Laplacian.

15. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 14 wherein:

said first computation means utilizes small 3 x 3 kernels for computing said directional sensitive Laplacian.

16. A multiscale adaptive system for enhancement of an image in x-ray angiography as claimed in claim 14 wherein:

said first computation means utilizes four kernels which are tuned respectively to 0 degrees, 45 degrees, 90 degrees and 135 degrees to extract line-like fine structures.

17. A multiscale adaptive method for enhancing an image in x-ray angiography comprising the steps of:

acquiring an input image;

enhancing said input image through a multiscale approach;

performing display mapping to modify pixel values such that relevant image information is presented in an optimal way; and

,

displaying said enhanced image;

wherein enhancing said input image comprises the steps of;

generating a pyramid of intensity images at pyramid levels from said input image;

generating enhancement masks from said intensity images at each of said pyramid levels;

first combining said enhancement masks together to form a single composite mask; and

,

second combining said single composite mask with said input image to form an enhanced image.

18. A multiscale adaptive method for enhancing an image in x-ray angiography as claimed in claim 17 wherein first combining comprises the steps of:

transforming mask at a coarsest scale to scale of its parent image to provide a transformed mask;

performing a weighting to said transformed mask;

adding said weighted transformed mask to mask of said parent image repeatably to form said single composite mask that is at full resolution.