X-ray or gamma ray systems or devices – Specific application – Absorption
Reexamination Certificate
2002-08-09
2004-07-06
Bruce, David V (Department: 2882)
X-ray or gamma ray systems or devices
Specific application
Absorption
C378S098800
Reexamination Certificate
active
06760401
ABSTRACT:
BACKGROUND
This invention relates to a method for processing digital images, in particular digitized X-ray images of the coronary blood vessels. Furthermore this invention relates to a method of achieving this objective.
In certain applications of methods for producing images, there arises the problem of intensification of small dark structures against a background which is of no interest. A representative example of this is the X-ray diagnosis of blood vessels, in particular coronary blood vessels. For diagnostic display of blood vessels, an X-ray-opaque contrast substance is frequently injected, so that the blood vessels through which the contrast medium passes appear on the X-ray image as dark, tube-like structures against the lighter background of bones, tissues and similar. For the purpose of obtaining even better display of the blood vessels, the X-ray image obtained is post-processed by Digital Subtraction Angiography (DSA) by subtracting a mask or a mask image from the actual X-ray image, so that the background contrast, but not the contrast of the blood vessels, is reduced. Generation of the mask images however is very difficult, in particular in displays of the heart, due to the movements of the patient himself and due to the heart beat.
A method and apparatus for contrast intensification in digital images is known from EP 0 527 525 B1. In this method, the digital image is first divided into N (N=1, 2, . . . ) detailed images with different resolution levels or spatial frequency (“Multi-Scale Decomposition”), where the nth detailed images (n<N) reproduce the image contrast corresponding to their resolution levels, i.e. approximately the difference from the mean gray-scale value for the particular resolution. The last, Nth, detailed image or “remanent image” yields an approximation to the original image, but without all the variations contained in the nth detailed images. After multi-scale decomposition, at least one of the detailed images is f-modified by means of a non-linear, monotonically increasing uneven conversion function (characteristic line) which has a gradient falling gradually with distance from the zero point, corresponding to multiplication of the pixel values by a symmetrical amplification function V which has a maximum>1 for pixel value zero. The relationship between the original image pixel values x and the modified pixel values y is expressed firstly by the conversion function or characteristic line f, and secondly by the intensification function V: y=f(x)=V(x)·x. As a result of the said modification, weak contrasts—both dark structures against a light background, and light structures against a dark background—are intensified.
SUMMARY OF THE INVENTION
In view of this state of the art, it was an object of the present invention to provide an apparatus and a method to ensure improved display of image details, in particular details of coronary blood vessels.
This object is achieved by a method for processing digital X-ray images comprising the steps of decomposing the digital image into N (N=1, 2, . . . ) digital images of different spatial frequencies and where appropriate different bandwidth and modifying at least one of the detailed images, the modification including at least one of intensifying small dark structures and attenuating other structures. The method includes recombining the detailed images into a whole image after their modification.
The object of the invention is achieved by an apparatus for image processing of X-ray images comprising an X-ray detector for generating digital X-ray images and a data processing unit. The data processing unit comprising means for decomposing the digital image into N (N=1, 2, . . . ) digital images of different spatial frequencies and where appropriate different bandwidth, means for modifying at least one of the detailed images, the modification including at least one of intensifying small dark structures and attenuating other structures, and means for recombining the detailed images into a whole image after their modification.
The method according to the invention for processing of digital images and in particular digitized X-ray images of coronary blood vessels comprises the steps of:
decompositing the digital image into N detailed images of different spatial frequencies and, where appropriate, different bandwidth, where N is a real number N=1, 2, . . . ;
modifying at least one of the detailed images so obtained, such that small structures having a high absorption capacity, which appear dark in the digital image, are intensified and/or other structures are attenuated; and
combining the detailed images after their possible modification to form a complete image.
The method proposed initially produces a multiscale decomposition of the digital image, so that detailed images are obtained with different resolution levels or spatial frequencies and, where appropriate, different bandwidth. Then the interesting structures are intensified and the uninteresting structures attenuated on the detailed images, which is generally considerably simpler in terms of computing effort than processing of the original digital image. The process can accordingly be carried out on images in real time. Furthermore, by selection of the modified detailed image it is possible to decide which spatial frequency and/or bandwidth of the original image is to be processed.
In accordance with a first embodiment of the process in particular the Nth detailed image, which by definition is the detailed picture with the lowest resolution and in certain variants of the multiscale decomposition is known as the “remanent image”, is modified by subtraction of a mask of equally low resolution. In post-processing of images of coronary blood vessels by means of a mask representing the uninteresting background, there exists the problem that the correlation between the background of the actual image and the subtracted mask can be insufficient due to the movement of the patient and the heart. In order to reduce the resulting artifacts, according to the invention only such structures are considered in the mask to be subtracted which have an extent greater than the variance arising from the movement. As disclosed, this can be most simply achieved if the subtraction of a mask of equal resolution level only occurs in the Nth detailed image (remanent image) of the lowest resolution level.
More precise allowance for motion artifacts can be obtained by applying a motion estimation and motion compensation on the mask used before its subtraction from the Nth detail image (remanent image). In this way, the mask follows and adapts to the movements of the actual image background as well as possible, in order to minimize errors due to movement of the object shown.
In accordance with a preferred further development of the process, the mask used for subtraction is calculated from the dynamically obtained digital image by formation of the temporal maximum value pixel by pixel. Such a process has advantages for coronary diagnosis when a contrast substance is used, where a pixel associated with a blood vessel is subjected to darkening due to the transient occurrence of the contrast substance. If therefore the temporal maximum of the gray-scale is stored for each pixel, an approximate image free of the effects of the contrast substance is produced. Thus there is no need to take a separate exposure without contrast medium. Details of this procedure are to be found in EP 0 200 262 B1.
In accordance with a further embodiment of the process, the mask used is generated by N-times successive low pass filtering with subsequent reduction of resolution. In spatial low-pass filtering, high spatial frequencies i.e. small structures are suppressed. In the reduction of resolution, the number of pixels with which the entire image is produced is reduced. By means of said processing steps it is possible to adapt a mask which originated in the original digital image with original bandwidth and resolution, to the bandwidth and resolution level of
Aach Til
Rongen Peter Maria Johannes
Schmitz Georg
Stegehuis Herman
Bruce David V
Koninklijke Philips Electronics , N.V.
LandOfFree
Apparatus and method for processing of digital images does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for processing of digital images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for processing of digital images will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3221006