JP2002526139A - Image restoration method and apparatus - Google Patents

Abstract

(57) [Summary] An image restoration system (65) capable of accurately displaying an image of a structure and a region in an object is disclosed. One aspect of the invention comprises a processing unit (1005) configured to receive X-ray transmission information of an object and to generate image pixel information at a plurality of depths within the object. The displayed image is comprised of selected image pixel information corresponding to image pixels at multiple depths within the object.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

【Technical field】

The present invention relates to the field of diagnostic X-ray images. Above all, an object
The present invention relates to a method for generating an image representing a structure in a subject.

[0002]

[Background Art]

Real-time X-ray images are increasingly needed for medical procedures as treatment techniques improve. For example, many electrophysiological heart procedures, peripheral vascular procedures, PTCA
There is a need for the use of real-time X-ray images in (percutaneous coronary angioplasty) procedures, urological procedures and orthopedic procedures. Furthermore, modern medical procedures often require the use of instruments such as catheters that are inserted into the human body. These medical procedures often work in conjunction with precise imaging of the surrounding body using x-ray images,
It often requires the ability to accurately recognize the position of an instrument inserted into the human body.

[0003] Many real-time X-ray imaging systems are known. These known systems include those based on fluoroscopy. In the system, X-rays are projected onto the object to be imaged, and shadows caused by relatively radiopaque objects within the object are displayed on a fluoroscope located on the opposite side of the object from the X-ray source. You. However, such systems use images that identify particular structures or regions deep within the object to be imaged (ie, where the images are “focused” on particular important structures or regions within the object). Very difficult to form. This is due in part to the geometry (geometry, geometry) of the fluoroscope-based system. That is, in this arrangement, the radiopaque properties over the entire depth of the object contribute to the final image, regardless of the exact depth of the particular radiopaque structure / region within the object.

One way to generate images of specific structures or regions within an object is provided by a computed tomography (CT) imaging system. During surgery, a CT system performs a combined X-ray projection or measurement of an object imaged from multiple angles. Data from the composite projection is processed to construct an image of a particular plane / section in the object.
Multiple image planes / intercepts can have varying depths within the object by moving the CT imaging system and the object relative to each other. However, conventional CT systems cannot produce an image focused on a particular structure in an object if the structure of interest traverses multiple image planes / intercepts at various depths within the object.

[0005] Another method of x-ray image generation involves the use of an inverse geometry x-ray imaging system.
In such systems, an x-ray tube is used where an electron beam is generated and focused on a small spot on a relatively large target assembly from which the x-ray is emitted. The electron beam is deflected in a scan pattern across the target assembly. A relatively small X-ray detector is located remotely from the target assembly of the X-ray tube. The X-ray detector converts an X-ray that has hit itself into an electric signal. This electric signal indicates the amount of the X-ray flux detected by the detector. As one of the advantages afforded by inverse geometry systems, the geometry of such systems allows X-rays to be projected onto objects from multiple angles without the need for physical repositioning of the X-ray tube. I do. However, the particular X-ray detector used in such a system often limits the spatial resolution of the system and therefore the quality / range of the image obtained. In addition, known inverse geometry X-ray imaging systems do not have the ability to generate focused images of structures at various depths within an object.

Accordingly, it is an object of the present invention to provide an image generation method and system that can generate accurate and focused images of structures at various depths within an object.

[0007]

DISCLOSURE OF THE INVENTION

The present invention comprises an X-ray imaging system capable of local focusing to any depth within an object (object). According to one aspect, the present invention comprises a system and method for generating a number of data comprising image information at a plurality of depths of an object under inspection, and selecting data from the number of data to generate a display image. I have. The above display image is
Comprising selected image pixel information that matches image pixels at a plurality of depths within the object.

[0008] These and other objects, advantages and aspects of the present invention will become apparent to those skilled in the art from a consideration of the drawings, the detailed description and the claims of the invention contained herein.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION

(System Overview) FIG. 1 is a diagram showing high-level components of an embodiment of an X-ray imaging system according to the present invention. The X-ray source 10 includes an electron beam source having a power supply. The power supply can operate the X-ray source 10 at about -70 kV to -120 kV. In the present embodiment, this voltage level produces an X-ray spectrum up to 120 keV.
The electron beam 40 generated in the X-ray source 10 by the charged particle gun is directed in a predetermined pattern onto the surface of the target assembly 50 (which in one embodiment is a grounded anode); For example, it is deflected by a scanning or stepping pattern. The X-ray source 10 includes a mechanism such as a deflection yoke 20 that controls an electron beam 40 across a target assembly 50 under the control of an electron beam pattern generator 30. An advantage provided by the geometry of the X-ray source 10 is that the X-rays can be emitted at multiple angles toward the target 100 without requiring physical relocation of the X-ray source 10. A method and apparatus for generating and moving the electron beam 40 across the target assembly 50 is disclosed in commonly owned U.S. Patent No. 5,644,612.

In FIG. 1, a collimating assembly is a target assembly 5 of an X-ray source 10.
0 and the detector array 60. In a preferred embodiment, the collimating assembly comprises a target assembly 50 and the object 1 to be imaged.
00 and 00. A currently preferred collimating assembly is
The collimator grid 70 includes an opening 80 through which a plurality of X-rays arranged in a grid pattern pass. The collimator grid 70 is designed to allow the passage of x-rays forming a divergent beam 135 that is directly intercepted by the detector array 60. In one embodiment, the collimator grid 7
0 uses a cooling assembly and a beam curing filter.

In operation, the electron beam 40 is preferably applied to a point 1 on the target assembly 50.
10, said location 110 is substantially one aperture 1 of the collimator grid 70.
20 is at a position where the axis 90 intersects the target assembly 50. Electron beam 4
When 0 hits the target assembly 50 at location 110, an X-ray cascade 1
30 is released. Only a portion of the cascade of x-rays 130, whose path is substantially along the axis 90, forms an x-ray beam 135 that diffuses through the aperture 120. The shape of the X-ray beam 135 is affected by the shape of the opening 120. For example, if the aperture is square, the X-ray beam 135 will exhibit a truncated pyramid shape. If the aperture is circular, the X-ray beam 135 will have an approximately conical shape. In a preferred embodiment, the shape and area of the aperture is such that the maximum diffusion area of the x-ray beam 135 is substantially equal to the dimensions of the x-ray capture surface of the detector array 60.

The detector array 60 includes a plurality of separate detectors (herein referred to as detector elements) 61 arranged. Each detector element 61 includes an X-ray plane with a capture area for detecting X-rays. Each detector element can independently measure the amount of X-rays that strike each detector element. When the object 100 is placed between the X-ray source 10 and the detector array 60, some of the X-rays of the X-ray beam 135 pass through a portion of the object 100 and constitute the detector array 60 if not dispersed or absorbed. Hits the detector element.
X-rays that strike any of the independent detector elements comprise a portion of the x-ray beam 135, referred to herein as the x-ray beam subpath.

In a preferred embodiment, each detector element comprises components for measuring the amount of X-ray photons hitting the detector element and emitting a signal corresponding to said measurement. Alternatively, each detector element comprises a component that produces an electrical signal that is approximately proportional to the total energy of the X-rays striking the detector element. The magnitude of the generated electrical signal is consistent with the X-ray flux intensity of the X-ray beam 135 from the proper X-ray beam subpath.
Using a detector array 60 that uniquely measures the X-rays that strike each detector element,
Line transmission information is generated. The X-ray transmission information is proportional to the flux of X-rays passing through the object 100 along a particular X-ray beam sub-path. The resulting intensity data is an image of the object 100, i.e., the X-ray transmission (transmissivity
eness) can be used or processed to create an image of
The image of the object 100 can be displayed on the monitor 140.

In one embodiment, the number of openings 80 in collimator grid 70 is
Zero or equal to the number of image pixels displayed on the visual display that can be connected to the video output of the X-ray imaging system. Alternatively, the ratio of the image pixels to the openings is increased. As a result, the number of openings is smaller than the number of image pixels displayed on the display device. For the purposes of this discussion, an "object pixel" is an area in the plane of the object for which information is collected. An image pixel is a picture element that is an image representation composed of one or a plurality of object pixels. The currently preferred number of openings is 100
10,000 pieces arranged in a × 100 lattice. The number of apertures shown above is for illustrative purposes only and will depend on the particular application to which the invention is applied.

X-ray transmission information obtained from the detector element 61 associated with a particular image pixel is reconstructed by an image reconstruction system 65, as described in further detail below.
ruct). In one embodiment, the image restoration system 65 also performs the control functions of the X-ray imaging system and prepares the display. Operational instructions and control of the X-ray source 10, detector 60 and image restoration system 65 are performed via a control workstation 150. The control workstation 150 also receives operational and status information from various components of the x-ray imaging system.

In some applications, more X than can be obtained with a single release from one opening
It may be desirable or necessary to use a line flux for each area of the object 100. This means, for example, that the material of the target assembly is not able to withstand enough electron beam bombardment during a single emission to produce the desired amount of x-ray flux (e.g., the heat generated by such bombardment). May happen in case). In these applications, multiple smaller releases from one opening may be made. The added X-ray flux can produce potentially more accurate images by reducing quantum noise.

In many inverted geometry X-ray systems, the spatial resolution of the resulting image is largely determined by the capture area of a single detector. Generally speaking, non-segmented detectors with small capture areas have high spatial resolution but poor collection efficiency (ie, the ratio of useful photons to the total number of photons passing through the object). Non-segmented detectors with large capture areas, on the other hand, have high acquisition efficiency but low spatial resolution. To address this problem, the present invention utilizes a detector array with a relatively large capture area, said detector array comprising a plurality of separate detectors with a relatively small capture area.

(Image Restoration) As shown in FIG. 2, the divergent X-ray beam path 200 is
One of the openings 210 diverges and extends toward the detector array 220. The X-ray beam path 200 passes through an object 230 on the way to the detector array 220. The x-ray beam path 200 traverses the object 230 at various planes of interest, such as the object plane 250.

X-rays traveling along the X-ray beam path 200 diverge after exiting the aperture 210,
Preferably, it has a cross-sectional area that varies from a minimum area approximately equal to the size of the opening to a maximum area approximately equal to the area covering the detection surface of the detector array 220. In one embodiment, the detector array 220 is positioned such that the maximum area of the X-ray beam path 200 covers only the total capture area of the detector array 220 and does not exceed this area. This minimizes the generation of X-rays that provide meaningless image information.

The X-ray beam return paths 270a, 270b, 270c, 270d, 270e and 2
70f is defined by the capture size and amount of the various detector elements that make up the detector array 220. The three-dimensional shape of the X-ray beam return path is essentially determined by the shape of the detector element. In other words, the X-ray beam return path (e.g., X-ray beam return path 270c) has a truncated apex at aperture 210 and a detector element 2
A solid defined by an elongated shape with a base equal in area to 80 capture areas. If the capture area of the detector element is round, the shape of the x-ray beam sub-path is substantially conical. If the capture area of the detector element is rectangular, the shape of the x-ray beam sub-path is substantially pyramid.

An object pixel 240 is an area in a plane 250 of interest. When the reconstructed image is displayed, object pixels 240 are represented by particular image pixels. This particular pixel is constructed using information obtained from some x-ray beam subpath across the object pixel 240.

Each detector element 280 of detector array 220 detects X-rays that have passed through object 230 at a particular portion of X-ray beam path 200. The amount of x-rays detected by detector element 280 provides information about the x-ray transparency of object 230 at a particular object pixel 240. An X-ray beam path 200 from each aperture 210 provides information to form a group of discrete information about the X-ray transparency of the object. The number of discrete information generated for an X-ray beam diverging from a single aperture corresponds to the number of 280 detector elements in the detector array 220. The currently preferred detector array consists of 48x48 detector elements. Thus, for each X-ray beam 200 emanating from one aperture, this results in 2304 discrete information about the X-ray transparency of the object pixel for the object plane defined in the object. . The X-ray transmission information obtained by the detector element from each of the X-ray beam sub-paths serves to generate pixel information of the image.

Image data for an object pixel is generated by collecting measured x-ray transmission information for an x-ray beam subpath that intersects a particular object pixel at a particular plane of interest. Depending on the plane of interest selected, the intersection of the subpaths on the plane of interest may not be perfectly coincident, but will only partially coincide. In FIG. 3, openings 300a, 300b, 300c, 300d, 33
X-ray beam sub-paths 330a, 330b diverging from 0e and 330f, respectively
, 330c, 330d, 330e and 330f are shown. Each of these X-ray beam part paths is their respective X-ray beam part that completely or partially coincides with the object pixel 320 on the plane of interest 325. X-ray beam partial path 330a-
X-ray transmission information (object pixel 320) obtained by detectors 335a-c for f
Image data that accurately represents the x-ray transparency of the object at the object pixel 320 is taken into account.

As shown in FIG. 4, there are a number of planes parallel to the source plane 350 and the detector plane 355. Some of the planes are located where the plurality of x-ray beam sub-paths are perfectly coincident over a plurality of spaced regions in the plane. These planes are called focal planes. Regions having a certain interval in the plane can be identified as object pixels. Examples of focal planes in FIG. 4 are planes 360, 362 and 364.
It is. Each focal plane has features such as a distance from the source, a pitch of object pixels, a specific area of the object that intersects the focal plane, and the like, which are different from other focal planes. Non-focal plane 36
5 is located between any two focal planes, the X-ray beam sub-paths emanating from the source plane at these non-focal planes only partially coincide with each other.

FIG. 5 shows a method of restoring a focal plane and a non-focal plane as a two-dimensional plane of an object pixel. According to the present invention, X-ray beam source position (source location)
SOURCEX × S having a pitch of λs in the x and y axis directions
A rectangular array of OURCEy sources is used with a detector array, preferably a square array of DETx x DETy detectors with a pitch of λd in the x and y directions. In one embodiment, each source is a separate aperture in the collimation grid and emits an x-ray beam path that covers the capture surface of the detector array. Each X-ray beam path is divided into a plurality of X-ray beam sub-paths, and the X-rays in each X-ray beam sub-path provide intensity data to one sensing element. Therefore, DE per X-ray path
There are Tx * DETY x-ray beam sub-paths, and for a total of DETx * DEty * SOURCEx * SOURCEy X-ray beam sub-paths from all sources in the source array, SOURCEx * SOURCEy X-ray beam paths is there.
INTERSITY (i, j, k, l) represents intensity data for X-rays detected by the detector DET (i, j) from among X-rays generated by the source SOURCE (k, l).

The focal plane can be written by a pair of natural numbers (integer> 1) m and n.
Here, m * λd and n * λs are the lengths of the bases of similar triangles, respectively. However, any plane, focal or non-focal, can be restored using the present invention. All that is required for the plane reconstruction is that the plane under reconstruction can be described by real values of m and n that are greater than zero.

Zd represents the distance from the source plane 350 to the detector plane 355, and Zp represents the distance from the source plane 350 to the object plane 330. Therefore, the distance Zp is m, n
And can be described as follows:

[0028]

(Equation 1)

According to one embodiment of the invention, the reconstruction of a two-dimensional array of image pixels IMAGE (m, n) in the object plane defined by m and n By generating an array of Image pixel values are generated by mathematically processing each value of INTENSITY (i, j, k, l) corresponding to an individual object pixel on the object. In one embodiment, INTENS for each object pixel
An image pixel for the object plane is generated by summing the values of ITY (i, j, k, l).
In this embodiment, each value of INTENSITY (i, j, k, l) is summed and
It is an appropriate image pixel defined by E (i * n + k * m, J * n + 1 * m).

In the case of a non-focal plane, any non-integer value is preferably assigned to the appropriate image pixel based on conventional rules of rounding to an integer. For example, m = 10
, An x-ray beam from a source having an xy index (1,1) impinging on a detector element having an xy index (1,2) when restoring the plane of n = 1.33 The partial path (which is represented as INTERNITY (1,2,1,1)) has coordinates (1 * 1.33).
+ 1 * 10, 2 * 1.33 + 1 * 10), which corresponds to the object pixel at (11.3, 12.7). The X-ray beam partial path has coordinates (11, 12), (12,
12), (11, 13) and (12, 13), and contains information on the object pixel at these coordinates. Therefore, the X-ray transmission value obtained from the INTENSITY (1,2,1,1) is calculated using any of the image pixels at coordinates (11,12), (12,12), (11,13) and / or (12,13). Could be assigned to Applying the general rule of rounding, the object pixel (11.3, 12.7) is
, 13) is currently preferred. It should be noted that other methods of assigning X-ray beam sub-paths to image pixel coordinates can be used without departing from the scope of the present invention.

The maximum x-index and y-index of the array IMAGE (m, n) are respectively DETx *
n + SOURCEx * m and DETy * n + SOURCEy * m. This embodiment does not change the position of the resulting plane to be imaged by multiplying the base of a similar triangle, for example by multiplying by two or three.

The pitch λp of the object pixels in the specific object plane Zp (m, n) can be expressed as follows.

[0033]

(Equation 2) In one embodiment, all m-th detectors in the x and y directions provide intensity information used in reconstructing one or more selected object pixels in the object plane. Therefore, the detector or X-ray beam partial paths provide strength information per object pixel is ^two approximately DETx * DETy / m. Since the total number of X-ray beam partial paths in this embodiment is DETx * DETy * SOURCEx * SOURCEy, the number of object pixels on the object plane from which X-ray transmission information used in image restoration can be obtained is as follows. Is represented as

[0034]

(Equation 3) Object pixels around the intersection area do not receive complete intensity information (ie, the number of sensing elements that measure the amount of flux through these object pixels is smaller than the sensing elements for other object pixels) Thus, the number of object pixels in the object plane that provides meaningful intensity information is slightly less than the number described above.

For example, n = 1.3 in a system with a 100 × 100 source array
3, the plane of m = 10 is 1,000,000 (100 × 1) in any object plane.
00 × 10 ² ) object pixels. Furthermore, the plane with n = 1.33 and m = 10 is 23,040,000 (48 ×) when a 48 × 48 detector array is used.
(48 × 100 × 100) X-ray beam partial paths. In this example, there should be about 23 X-ray beam sub-paths that completely or partially coincide with each object pixel. However, due to the geometry of the system (geometry, geometry), the object pixels on the edges of the object plane may have fewer than 23 x-ray beam sub-paths that completely or partially match them. .

When the size of the source array is SOURCEx * λs × SOURCEy * λs, the field of view of a specific object plane can be expressed as follows.

[0037]

(Equation 4)

The field of view can be changed by using some but not all of the sources in the source array. By using a smaller number of apertures located in certain areas of the collimator grid, the area to be imaged can be smaller.

In forming each image plane, the X-ray transparency information must be processed so that it is always associated with the image pixel to which it is assigned. In addition, when using an electron beam with a suitable stepping pattern, each aperture emits X-rays more than once to create a single frame, and each detector has an X-ray assigned to the same image pixel. Provide line transmission information more than once. In this situation, the X-ray transmission information obtained by the same detector from the X-rays emitted from the same opening in the same frame is combined, and then the X-ray transmission information obtained from the other openings in the frame is obtained. It is presently preferred to combine it with the assigned X-ray transparency information assigned to the same image pixel.

The invention is not really a two-dimensional plane constituted by the mathematical combination of the relevant X-ray beam sub-paths, but rather a volume slice having a certain depth.
Is considered. The X-ray absorptivity measured by the detector element as the X-ray intensity is a measured value depending on the depth. Without measurement of X-ray absorption over the entire depth, there would be little or no difference between object areas of different densities. Thus, the reconstructed image array represents a two-dimensional object plane in the reconstructed “intercept”. A section is a substantially flat area within an object at some depth. The term "voxel" refers to a capacitive element within a section of the object to be imaged.

The image restoration method of the present invention generates information for various planes and intercepts at multiple locations between the source plane and the detector plane. The ability to reconstruct various planes / intercepts creates images of individual regions of the object by selecting the appropriate intercept near the region of interest of the object without having to change the source and detector locations Used for

This image restoration method also increases the effective field depth of the generated image by providing the ability to restore multiple planes / intercepts of the region of interest. A plurality of image planes representing the reconstructed intercept can be combined to create a single array of image pixels with high spatial resolution in the region of interest.

Referring to FIG. 6, the first X-ray beam sub-path 400 and the second X-ray beam sub-path 405 are two of many X-ray beam sub-paths radiating from the first aperture 410 of the collimator grid 412. is there. The remaining X-ray beam sub-paths are not shown for clarity and explanation. A portion of the x-ray traveling along the first x-ray beam sub-path 400 and the second x-ray beam sub-path 405 passes through the object 415 and
0 detectors 425 and 427, respectively. The information provided to the detector 425 by X-rays traveling along the first X-ray beam sub-path 400 does not code at any single point in the object 415, and conversely, Path is object 415
To form a capacitance that intersects the first section 430, the second section 435, and the third section 440. In particular, X-rays traveling along the first X-ray beam sub-path 400
Create a volume that is completely or partially coincident with voxel 445, second voxel 450, and first voxel 455. For reconstruction, the information obtained by the detector 425 from the first x-ray beam subpath 400 can be used to generate an image pixel representing the first voxel 445 in the image plane representing the section 430, and the section 435 It can be used to generate an image pixel that represents the second voxel 450 in the image plane that represents it and / or can be used to generate an image pixel that represents the third voxel 455 in the image plane that represents the intercept 440. . From this data, image planes are created using the methods described in FIGS.

For the second x-ray beam sub-path 405, the information provided by the detector 427 can be used to generate an image pixel representing the fourth voxel 460 in the image plane representing the section 430 and representing the section 435 It can be used to generate an image pixel representing the fifth voxel 465 in the image plane and / or can be used to generate an image pixel representing the sixth voxel 470 in the image plane representing the intercept 440.

The third X-ray beam sub-path 475 and the fourth X-ray beam sub-path 480 are two of many X-ray beam sub-paths radiating from the second aperture 485. Second opening 4
The remaining x-ray beam sub-paths emanating from 85 are not shown for clarity and explanation. X traveling along X-ray beam sub-path 475 and X-ray beam sub-path 480
A portion of the line passes through object 415 and strikes detectors 490 and 491, respectively. Third
The intensity information provided to the detector 490 by x-rays traveling along the x-ray beam sub-path 480 does not code at any single point in the object 415, and conversely, the intensity information is A collection of information about the capacitance across the entire plane / intercept between the collimator grid 412 and the detector array 420, including the plane / intercept with 476,477,478.

In one embodiment, image pixels are created by combining or summing voxel intensities from all detectors that detect x-rays traveling along the x-ray beam subpath. The x-ray beam sub-path completely or partially coincides with a particular voxel and has been assigned to the voxel for reconstruction. For example, the image pixel representing the sixth voxel 470 would include intensity data collected by the detector 427 from the x-ray beam sub-path 405 and intensity data collected by the detector 490 from the x-ray beam sub-path 475. .

A preferred reconstruction method is to separately reconstruct a number of sections simultaneously. In the example of FIG. 6, the slices 430, 435, 440 are individually reconstructed, and the various image pixels / voxels that make up each slice are combined to create an array of display pixels for producing an image on a display monitor or film. Or otherwise processed.

[0048] The image pixels of the reconstructed section can be stored in memory as image planes and used to display the image on a two-dimensional display. The two-dimensional display includes an array of display pixels each representing a position on the display. The display pixels are two-dimensional x and y
While the image pixel has not only the x and y coordinates, but also the z coordinate corresponding to the distance of the image pixel from the source (or detector). For example, an image pixel in an intercept closest to the source may be given a value of z of 1, and an image pixel in an image intercept furthest from the source may be given a value p of z. p is the total number of generated image slices.

When an image on a two-dimensional display is generated by combining all image planes / intercepts together, the display pixels may appear as images with unclear edges. This is because the displayed image does not only match one or two image pixels, but also all image pixels that have the same X and Y coordinates on different image planes / intercepts. In order to obtain a meaningful image, it is preferable that only one image pixel among a number of image pixels having the same X coordinate, Y coordinate and different Z coordinate is selected for display as a display pixel. The display image is formed from a set of the selected image pixels corresponding to individual X and Y coordinates on the display. When using the present invention, it should be noted that the displayed image can be focused on more than one region or structure of interest at multiple depths within the object being imaged.

Also, a portion of a displayed image can be formed by combining one or more image pixels with the same X and Y coordinates on different image planes / intercepts. The combined intensity data of image pixels having the same X and Y coordinates and a different Z coordinate is calculated using the appropriate X
, Y coordinates. Relevant information about the object under inspection is on a number of image planes / intercepts, and one display pixel represents a larger set of information than selecting only one of those image pixels for display. Where more appropriate to form, a combination of two or more plane / intercept image pixels is performed.

There are a number of methods used to select among different image pixels that have the same X and Y coordinates but different Z coordinates corresponding to different image planes / intercepts for display.
The currently preferred selection method is the maximum intensity projection algorithm. Same X coordinate and Y
For each set of image pixels having a Z coordinate different from the coordinates, the maximum X-ray intensity projection algorithm selects the image pixel with the highest intensity value from the set. This image pixel, which takes the maximum X-ray intensity or luminance value, is then displayed on the display as a display pixel with the appropriate X and Y coordinates.

Another algorithm is a minimum intensity projection algorithm. In this method, an image pixel having the minimum intensity, that is, a luminance value, is selected for display from a set of image pixels having the same X coordinate, Y coordinate, and different Z coordinate.

Another way to select from image pixels that have the same X and Y coordinates but different Z coordinates corresponding to different image planes / intercepts for display is to display an object in the field of view. There is a choice of some of the planes / intercepts that have the most contrast in a particular area in order to do so.

Yet another method of selecting from image pixels that have the same X and Y coordinates but different Z coordinates corresponding to different image planes / intercepts for display is to display an object in a field of view. There is a selection of a part of the plane / intercept that has the highest energy within a particular spatial frequency range in a particular region.

Further, as another method of selecting from image pixels having the same X coordinate and Y coordinate but different Z coordinates corresponding to different image planes / intercepts to be displayed, an object is displayed in a visual field. There is a selection of the most detailed planes / intercepts in a particular area to do so. In an embodiment using this method, for each image pixel (I _{X, Y} ), two of the nearest pixels are used to determine the slope as follows.

[0056]

(Equation 5)

In another method of determining the gradient or gradient, the absolute value of the difference between adjacent or adjacent pixels is calculated to determine the gradient. For example, the following equation can be used to determine the slope.

[0058]

(Equation 6)

The image pixel (I _{X, Y} ) with the largest gradient from the group with the same X and Y coordinates but a different Z coordinate is therefore selected for display. Other gradient calculations may be utilized within the scope of the present invention. It includes other calculation methods, such as considering image pixels that are in the same proximity and weighting image pixel gradient values according to the location of the pixel image.

In another embodiment, the display can be created by selecting all image pixels from a single image plane for display. In one embodiment, the single plane selected is the one that has the greatest overall in-plane contrast.

Further, two or more of the disclosed methods can be combined in any manner, depending on the needs of a particular use or application.

The above method for selecting among image pixels where the X and Y coordinates are the same but the Z coordinates corresponding to different image planes / intercepts for display are different, is to use different depths within the object. Note that other image generation modalities for restoring different image planes / intercepts (or other types of image data) can be similarly applied. For example, the above method can be applied to multi-section CT data to display an image of an object in view.

(Image Restoration System) FIG. 7 is a block diagram of an embodiment of the image restoration system 65. The image restoration system 65 includes a PCI interface 1010 connected to the control workstation 150. In one embodiment, the sensing module 700 comprises components of the detector array 60 and receives X-ray transmission information. Alternatively, the detector array 60 is physically separated from the image restoration system 65, and the detection module 700 includes components for receiving data signals from the detector array 60. Image restoration chassis 1
005 is an interface module 710 and one or more plane restoration modules 7
30, an image selection module 750, and an image preprocessor (preprocessing device) 76
0. The various components on the image restoration chassis 1005 are interconnected via one or more buses 1100. The bus 1100 also includes a control line. A video post-processor (post-processing device) 770 is connected to the display monitor 1080. In one embodiment, each of the above components is a field
Although implemented using a programmable gate array, circuits, chips, and hardware / software combinations capable of performing the desired functions may be used without departing from the scope of the invention.

The detection module 700 determines whether or not this detection module combines the X-ray transparency information obtained from the detector array 60, how to combine it, and when to output the X-ray transparency information to the interface module 710. It can operate in several modes to determine what to do. The modes include (1) an image acquisition mode for outputting the X-ray transparency information detected by the detector array to the interface module 710 for image processing, and (2) inputting the X-ray transparency information for diagnostic purposes. Sensor mode, (
3) Alignment mode in which the position of the electron beam on the target assembly is aligned with the aperture of the collimator grid using the X-ray transparency information obtained by the detector array 60, and (4) X-ray transparency information in the image acquisition mode There is an image test mode in which X-ray transparency information is transmitted to the interface module 710 only in response to a predetermined timing control signal.

FIG. 9 shows an embodiment of the interface module 710. The signal representing the X-ray transparency information sent from the detection module 700 to the interface module 710 is received by the data receiver 718 and is restored from the multiplexed signal to an individual signal (demultiplex). X received by data receiver 718
The line transparency information is sequentially input to the FIFO memory 720. FIFO memory 72
0 adjusts the transition of the data sent to the sensor data processor 722. When the image forming system is in the image acquisition mode, a signal representing X-ray transparency information is sent from the sensor data processor 722 to the plane restoration module 730. When the imaging system is in the sensor mode or the alignment mode, the sensor data processor 722 sends the X-ray transparency information to the sensor alignment memory 714. Sensor alignment memory 714 stores information for transmission to control workstation 150 when sensing module 700 is operating in alignment mode or sensor mode. The line pointer memory 716 stores a control signal received from the control workstation 150 and representing the order of the electron beams and the stepping pattern, and outputs a signal representing this information to the detection module 700. The interface controller 712 sends and receives signals between the interface module 710 and the control workstation 150, and sends signals to the detection module 700.

FIG. 10 shows an embodiment of the plane restoration module 730. Each plane restoration module 730 restores image pixels for a single plane at a particular depth between the X-ray source and the detector array. The image restoration system 65 includes a plurality of plane restoration modules 73 for restoring a plurality of image arrangements for planes having different depths between the X-ray source and the detector arrangement.
0 can be used. The preferred embodiment has 16 separate plane restoration modules 7
30. Each of these plane restoration modules 730 restores an image array for a single plane. In another embodiment, each plane restoration module comprises a component for restoring an image array for two or more corresponding numbers of image planes. Each plane restoration module 730 includes a restoration processor 1300, a correction look-up table 1310, an image frame buffer 1330, a selection processor 1340, and a selected plane driver 1340. The restoration processor 1300 includes circuitry for generating an image pixel array representing a particular object plane according to the method described with respect to FIGS. 2-6. Restoration processor 130
0 (which preferably operates with a 50 MHz clock) processes data received from the interface module 710 corresponding to the X-ray transparency information detected by the detector array 60. In a preferred embodiment, the X-ray sub-paths assigned to the same image pixel in one image plane are combined to generate data for a single image pixel. This process is repeated for each image pixel in the same image plane. The output of the decompression processor 1300 is an x × y array of image pixel information for one image plane.

A signal representing image pixel information is output from the restoration processor 1300 to the correction lookup table 1310. For example, image formation defects may occur due to the process of mapping items of X-ray transmission information to image pixel values. This image formation defect is one in which an image pixel has an incorrect value for other image pixels because it receives less or more X-ray transparency information than the other image pixels. As another example, various manufacturing and design defects / tolerances in the components of the detector array 60 may produce image information that is not completely accurate. Some sensing elements may have manufacturing defects, which always cause the image pixel at coordinates (50,50) to have a value that is too high by 5%. Before using the image restoration system 65, the correction look-up table 1310 is programmed with calibration settings to correct any imaging defects. Correction lookup table 131
0 normalizes the image pixel values to correct for image defects that appear when the system is operated with no objects in the field of view. Thus, for image arrays that always have image pixels that are too 5% higher, the correction look-up table 1310 is programmed to reduce the incorrect pixel by the appropriate value to produce the correct image pixel value. You.

The image frame buffer 1330 (which preferably operates with a 50 MHz clock) stores the image pixels after normalization by the correction look-up table 1310. In the preferred embodiment, this image buffer is a single memory, but may be replaced by several smaller memories or shift registers.

The selection processor 1340 receives the corrected image pixel information from the image frame buffer 1330. The selection processor 1340 includes circuitry for processing pixels in the image plane to help determine which of the pixels at a particular x, y coordinate of various different image planes to select and display. The previous section, entitled "Image Restoration", described several methods that can be used to select from image pixels that have the same x, y coordinates but different z coordinates depending on the image plane. For example, one method calculates the gradient or gradient of an image pixel using its neighboring image pixels and selects an image pixel for a particular portion of the displayed image. In a preferred embodiment, the selection processor generates a quantifiable "selection" or gradient value for each image pixel. This gradient value can be compared to the gradient values for pixels in other image planes to determine which pixels should be selected for display.

The selection values or gradient values for all the pixels restored in the image restoration system 65 are sent to the image selection module 1050 (FIG. 7). The image selection module 1050 compares the received selection or gradient values and selects one of the reconstructed image pixels for display. A specific plane restoration module 730 for restoring an image plane including the selected image pixel is provided with a selected plane driver 1340 through a selected plane driver 1340.
This selection is notified. Next, the selected plane driver 1340 for that particular plane restoration module 730 outputs image information about the selected image pixel to the image processor 1060 for display.

Each restoration processor 1300 preferably comprises a row processor 1415, a row-column translator 1420, and a column processor 1430. The reconstruction processor 1300 captures X-ray transmission information received by the detector element array with continuous irradiation by the X-ray source and converts the information into an array of image pixels. In a preferred embodiment, the X-ray source comprises a 100 × 100 array of X-ray source locations and a 48 × 48 array of detector elements. The restoration processor 1300 is 48 × 4
8 is converted to a smaller array of image pixel information. In a presently preferred embodiment, a 10 × 10 array of image pixels is generated for each 48 × 48 array of X-ray transmission information received in a single irradiation. This 10 × 10 array of image pixel information is the newest 48 × 48 array of X-ray transmission information and the 10 × 1 of the previously received 48 × 48 array of X-ray transmission information.
It is generated in combination with a portion related to an image pixel in the image pixel array of 0. Thus, for a 100 × 100 array of X-ray source locations, embodiments of the present invention produce a 1000 × 1000 array of image pixels. (10 of this image pixel
The array of 00 * 1000 is (100 * 100 X-ray source positions) * (10 * 10 image pixels for each X-ray source position). In a preferred embodiment, row processor 1415 receives a stream of data representing a 48x48 array of image information and processes this image information to combine information from various rows associated with the same image pixel to form an image. 10 rows of pixel information x
Generate an array of 48 columns. Next, the row-to-column translator 1420 and column processor 1430 capture the 10 row × 48 column output of row processor 1415,
Combine image information from various columns associated with the same image pixel. The output of column processor 1430 is an array of 10 rows × 10 columns of image pixel information for a particular X-ray source location, ie, a particular image plane for a collimator aperture.

FIG. 11 illustrates an embodiment of a row processor module 1100. FIG.
Embodiments process information about one row of detector elements. Therefore, row processor 1415 includes 48 row processor modules 1100 to process the 48 rows of information received from the detector array. The X-ray transmission information for one row of detector elements is input to input register 1610 and gate 162.
0 is received. Image information from different detector elements associated with the same image pixel is combined by summer 1600. The pixel information generated by adder 1600 is provided to a row adder memory. In one embodiment, the row adder memory comprises four separate 32-word 7-bit memories 1632, 1634, 1
This is a 128-word memory composed of 636 and 1638.

If pixel information has already been received for a particular image pixel, but that pixel information is incomplete, there is more image information that has not yet arrived for this image pixel, This incomplete pixel information is selectively stored in memory 1632,
Stored in 1634, 1636, or 1638. When new image information for this same image pixel arrives at adder 1600, then the previously stored pixel information is selectively passed through multiplexer 1640 to adder 1600.
And is combined with the new image information. If all image information that is to arrive for a particular image pixel has already been processed, the pixel information for this image pixel passes through multiplexer 1640 into word memory 1650. The individual memories 1632, 1634, 1636, or 1638 that store this information may thereafter be reused to hold pixel information for another image pixel. Word memory 1650 stores pixel information for image pixels, which does not further receive any x-ray transmission information from any of the detectors in a row of detectors coupled to row processor 1100. .
This pixel information is then passed to row-column translator 1420 and column processor 1420.
Sent to 30.

As described above, the output of row processor 1415 is an array of 10 rows × 48 columns of information. This 10 × 48 array of information is processed by row-column translator 1420 and column processor 1430 to obtain a 10 × 48 array of image pixels for a particular object plane.
Generate 10 arrays. Image pixel information is such that all of the partial image pixel information in one column of row-column translator 1420 is output to one column processor.
Output from row-column translator 1420 into column processor 1430. Preferably, the number of column processors 1430 utilized is equal to the value of m of the image plane being restored (as described in the section on image restoration). This is because this allows the most efficient processing in terms of speed. Column processor 1430 adds the partial image pixels provided by the column processor to the partial image pixels from another column processor assigned to the same image pixel, for each image pixel constituting the image plane. To form complete image pixel information.

FIG. 12 illustrates one embodiment of a circuit 1200 that combines the functions of a row-column translator 1420 and a column processor 1430. Each column input to the circuit 1200 includes an input register 1660 and a memory 1670.
In the preferred embodiment, each memory 1670 and register 1660 contains 6
Image pixel information received from one column processor module 1100 is handled. The output of each memory 1670 is output to a multiplexer 1680, which outputs a signal containing the image pixel information stored in the memory 1670.
Data for the sequence of information passes through multiplexer 1680 and to adder 170
Enter 0. Adder 1700 combines the information associated with the same image pixel together. Since there is more image information for a particular image pixel that has not yet been processed, if the pixel information that occurs for this information pixel is incomplete, the incomplete pixel information is stored in memory 1705. In one embodiment, memory 1705 includes registers and off-board memory. When additional image information for this same image pixel arrives at adder 1700, the previously stored pixel information is returned to adder 1700 and combined with this additional information. If all of the image information that is to arrive for a particular image pixel has already been processed, the pixel information for that image pixel passes to memory 1710 or memory 1715. This information is then conveyed to the correction look-up table 1310, and the corrected set of image information for the array of image pixels is conveyed to the image frame buffer 1330 (FIG. 10).

As described above, the selection processor 1340 receives the modified image pixel information from the image frame buffer 1330. The selection processor 1340 includes circuitry for processing pixels in the image plane, which determines which of the image pixels in the various image planes at a particular x and y coordinate is selected for display. Help to do. In a presently preferred embodiment, the selection processor determines a gradient value for each image pixel in the reconstructed image plane. One embodiment of the selection processor 1340 includes a selection input filter 1810, a FIFO memory 1820, a selection gradient generator 1830, and a selection output filter 1840.

The selection input filter 1810 acts to “smooth” the data by improving the difference between adjacent image pixels caused by noise. The currently preferred selection input filter 1810 is a low-pass filter, such as a "box car" filter that takes a moving average of a group of image pixels surrounding any image pixel. For example, one embodiment of a boxcar filter can generate a moving average based on a moving group of nine image pixels. To illustrate, FIG.
Shown at a is a representative group of image pixels 1357. The filter calculation for image pixel 1351 is obtained from the average of the values of the nine image pixels in box 1352. The filter calculation for image pixel 1353 is obtained from the average of the values of the nine image pixels in box 1354. Filtering is performed on all image planes by successively generating such a moving average for every image pixel in the plane.

It should be noted that if the average value for the nine image pixels in box 1352 has already been obtained, then these images are used to obtain the average of the nine image pixels in box 1354. That is, it is not necessary to add all the individual values of the pixel. Instead, the average of the nine image pixels in box 1354 is calculated using the calculated sum of the values in box 1352 and subtracting the values of the three image pixels in the circled column area 1355, Next, it can be obtained by adding the values of the three image pixels in the circled column area 1356. It will be appreciated that this optimization may also be performed when the image pixel under inspection is in the next row. Therefore, for the image pixel in this next row, the value of a group of three image pixels in the specific area of the previous row is subtracted while the three image pixels in the specific area of the next row are subtracted. An average can be obtained by adding a group of values.

[0079] Shown in FIG. 13b is an illustration of an embodiment of a selective input filter 1810 that can perform this type of filtering. Image pixel information is received through FIFO 1358. Image pixel information from the preceding region of the image pixel column is stored in FIFO 135
9 has been stored. When new information arrives through FIFO 1358, the new information is sent to subtractor 1365 through register 1361 and gate 1362. Previously stored image pixel information stored in FIFO 1359 is passed through a register 1360 to a subtractor 1365 where it is subtracted from the new image pixel information. The output of the subtractor 1365 is sent to the adder 1366, where it is added to the pre-processed information about the image pixel array area stored in the FIFO 1363. The output of adder 1366 is FIFO 1363
Where it waits for more information to be added. Further, the output of adder 1366 is sent to subtractor 1369, where the stored value from memory 1368 corresponding to the preceding region of the row of image pixels is subtracted. The output of subtractor 1369 is provided to adder 1370, where it is added to a value corresponding to the area of the image pixel.

The filtered image pixel values are sent to FIFO memory 1820 (FIG. 10). FI
FO memory 1820 controls the flow of data between select input filter 1810 and select gradient generator 1830. Selection gradient generator 1830 performs calculations to generate quantifiable selection or gradient values. The selection or gradient value indicates whether a particular image pixel at the x and y coordinates of the image array is to be selected as the display pixel at this coordinate on the display image. In a preferred embodiment, this quantifiable selection value is generated by determining a gradient (eg, a variance) of each image pixel relative to neighboring image pixels. In one embodiment, the gradient is determined by taking the absolute value of the difference between each image pixel and the image pixel above, below, right, or left of this image pixel. However, other methods for making gradient determinations can be utilized in accordance with the present invention without departing from the scope and spirit of the invention. For example, it is possible to determine the square root of the sum of the squares of the differences between each image pixel and an adjacent or other neighboring image pixel.

FIG. 14 a illustrates a process for gradient determination according to an embodiment of the present invention. Shown in FIG. 14a is a portion of the image pixel array on the first image plane. As described above, the gradient determination for an image pixel can be made by taking the absolute value of the difference between the values of two opposing neighboring image pixels. Therefore, image pixel 18
An example of a gradient determination for 02 is to calculate the absolute value of the difference between the value of its neighboring image pixel 1804 (value = 5) and the value of image pixel 1806 (value = 5). Since the absolute value of (5-5) is 0, the image pixel 1802 is represented by a low gradient value. That is, since the absolute difference value of the image pixel adjacent to the image pixel 1802 is low, the level of “detail” to be indicated by this image pixel is also low. Thus, the gradient value obtained for this image pixel indicates that this gradient value is not suitable for being selected for display. For clarity and explanation, the difference between only two neighboring image pixels is shown here, but as mentioned above, including performing the absolute difference calculation of more neighboring image pixels, Additional methods may be used to determine the slope value.

FIG. 14b shows the same image pixel coordinates as shown in FIG. 14a,
The image pixel coordinates are located on the second image plane. Therefore, image pixel 1808 in FIG. 14b has the same x and y coordinates as image pixel 1802 in FIG. 14a. The gradient value obtained for an image pixel 1808 is the value of its neighboring pixel 1809 (
Value = 10) and the absolute difference between the values of 1812 (value = 1). Since the absolute value of the difference between these two neighboring pixels is 9 (absolute value (10-1)), image pixel 1
The slope value of 808 is relatively high. Therefore, this determination indicates that image pixel 1808 has a higher gradient value than image pixel 1802 (and thus indicates a finer portion) and at its corresponding x and y coordinates on the display image It is more appropriate to be selected as an image pixel to be displayed.

FIG. 14 c illustrates an embodiment of the selection gradient generator 1830. In FIG. 14c, the selected gradient generator 1830 includes four gradient processors 1831, 1832
, 1833, and 1834. Each gradient processor calculates a gradient value for one image pixel. In this embodiment, the gradient values for a group of four image pixels are (adders 1841, 1842, 1843 and accumulator 18
44 together to produce one group slope value. With this group gradient value, it is possible to consider displaying each group of four image pixels in the image plane collectively as one group rather than individually for each pixel. This prevents the roughness of the displayed image when extraneous (and possibly inappropriate) image pixels are individually selected or displayed in the displayed image. Alternatively, the selection gradient generator 1830 may generate individual gradient values for each image pixel, so that each image pixel may be individually determined for display.

The gradient processor 1831 includes multiplexers 1851 and 1852, which receive image pixel values for neighboring image pixels along both sides of the image pixel whose gradient value is being calculated. The subtractor 1853 calculates a difference between image pixel values of neighboring image pixels. The absolute value of the difference is calculated by the addition / subtraction unit 1854, and then the final gradient value for the image pixel is
Output to adder 1841. Gradient processors 1832, 1833 and 183
4 includes a set of similar structures.

Next, the gradient information is provided to select output filter 1840 through register 1846 (FIG. 10). Select output filter 1840 preferably plays the same role as played by select input filter 1810. That is, the selection output filter 1840 takes an average value for the gradient information group corresponding to the image pixel group around the specific gradient value, and uses the average value instead of the gradient information for each image pixel. Therefore, in a preferred embodiment, the selected output filter 1840 comprises the circuitry described above, and performs the role described above in more detail in connection with the selected input filter 1810 (FIGS. 13a and 13b).

Referring to FIG. 15, the image selection module 750 includes a selection output filter 184.
From 0, a signal representing the slope value is received. Image selection module 75 during each cycle
Gradient values entered at 0 correspond to image pixels that have the same x and y coordinates but are located on different image planes. These gradient values are input into one or more magnitude selection circuits 2110. Each magnitude selection circuit 2110 compares two gradient values. The magnitude selection circuit 2110 selects the larger one of the two gradient values. Each magnitude selection circuit 2110 outputs the gradient value to the selection multiplexer 2120. The selection multiplexer 2120 includes information indicating the plane reconstruction module 730 from which each of the image pixels or gradient values was generated.

In operation, each size selection circuit 2110, except for the last size selection circuit 2115 located at the lowest position in the image selection module 750, provides a selection gradient value below each size selection circuit 2110. Is output to the size selection circuit 2110 in the section (1). Each size selection circuit 2110 also provides a signal to its corresponding selection multiplexer 2120 to provide an image plane reconstruction module 730 for which the image pixel value or gradient value has been selected.
The selection multiplexer 2120 is instructed to output the plane position information representing. The plane position information is stored in the next size selection circuit 2110 below the current size selection circuit.
Or the last selection multiplexer 2125 linked to
Is output to the output multiplexer 2130.

The currently preferred plane selection device 752 includes four rows of size selection circuits 2110. First row 2140 preferably includes a number of size selection circuits 2110 equal to half the number of image plane restoration modules 730. Therefore, if the number of image plane restoration modules is 16, the first row 2140 preferably contains 8
There are two size selection circuits. The second row 2150 is preferably the first row 2140
The third row 2160 preferably includes half the size selection circuits of the second row 2150. The last magnitude selection circuit 2115 outputs the highest magnitude gradient value for image pixels that have the same x, y coordinates but are located on different image planes.
Therefore, the last size selection circuit 2115 determines which particular image pixel is selected for display or processing.

The above structure is efficient in comparing two image pixels and discarding image pixels that cannot be selected. Therefore, the image pixels that cannot be selected are
It is automatically discarded without performing additional comparisons or using large matrices to perform calculations. This improves the processing speed of the image generation system and reduces the number of components used to perform the processing. The plane selection image buffer 754 is provided with the plane arrangement of the selected image pixels through the multiplexer 2165 and the output register 2170. Tiled image controller 2180
Thus, a fragment of each image can be displayed. The tiled image controller allows a user to select specific image pixels for display or processing, regardless of image plane or image position. For example, a user can select two or more image pixels that have the same x, y coordinates but are located on different image planes.

Referring to FIG. 8, a signal representing image pixel information for an image pixel selected for display is sent to an image preprocessor 760. Image preprocessor 760
Is the size (mag) of image pixels in various regions by the region of interest integrator 761.
nitude) and includes circuitry for modifying display brightness. The region of interest integrator 761 combines image pixels located in the same region of the display plane selected for display. The output of the region of interest integrator 761 can be used to determine whether to change the intensity of the electron beam at the source to adjust the x-ray intensity of the system. The output of the region of interest integrator 761 is also provided to a digital video interface 763 for output to a video post-processor 770 for modification of the brightness of the display monitor 1080. The preprocessor image buffer 764 provides information to a preprocessor look-up table 765, which performs a gamma correction function on the image pixels selected for display. The corrected image pixels may be provided to a digital to analog converter 766 for use with an analog display.

Digital video interface 767 provides processed image pixels from image preprocessor 760 to video postprocessor 770. The information output from the image pre-processor 760 is output to a video post-processor 770, which operates a display 1080, which is preferably a standard video monitor. Video post-processor 770 preferably also includes circuitry for storing the output image pixels for use in playback or other calculations.

While embodiments, uses, and advantages of the present invention have been described and described in detail, possible embodiments, uses, and advantages are provided without departing from the spirit of the inventive concept described herein. There are more benefits. Therefore, the present invention relates to the preferred embodiments described above,
It is not limited to description or drawings. Accordingly, the protection to be granted on this patent should be limited only in accordance with the spirit and intended scope of the appended claims.

[Brief description of the drawings]

FIG. 1 is a diagram showing components of an X-ray imaging system according to the present invention.

FIG. 2 represents one x-ray path originating from one aperture of the collimator grid and passing through an object on the way to the detector array.

FIG. 3 shows a plurality of x-ray beam sub-paths originating from a plurality of apertures in a collimator grid and passing through an object en route to a detector array.

FIG. 4 illustrates an arrangement of a plurality of planes between a source plane and a detector plane according to one embodiment of the present invention.

FIG. 5 shows a geometric arrangement between a regularly spaced array of X-ray sources and a detector array.

FIG. 6 shows a partial X-ray beam path through an object to a detector array.

FIG. 7 is a block diagram of an embodiment of an image restoration system according to the present invention.

FIG. 8 is a diagram of an image preprocessor according to one embodiment of the present invention.

FIG. 9 is a diagram of one embodiment of an interface module.

FIG. 10 is a diagram of one embodiment of an image plane restoration module.

FIG. 11 is a diagram of a row processor module of one embodiment of the present invention.

FIG. 12 is a diagram of a row-column translator and column processing unit of one embodiment of the present invention.

FIG. 13a is a diagram of selective filtering of one embodiment of the present invention.

FIG. 13b is a diagram of a selective input filter of one embodiment of the present invention.

FIG. 14a illustrates selective gradient machining of one embodiment of the present invention.

FIG. 14b depicts selective gradient machining of one embodiment of the present invention.

FIG. 14c is a diagram of a selection gradient process according to an embodiment of the present invention.

FIG. 15 is a diagram of one embodiment of an image selection module.

[Procedural Amendment] Submission of translation of Article 34 Amendment of the Patent Cooperation Treaty

[Submission date] September 29, 2000 (2000.9.29)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, TZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, CR, CU, CZ, DE, DK, DM, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID , IL, IN, IS, JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, TZ, UA, UG, UZ, VN, YU, ZA, ZW (72) Inventor Robert・ Emeren USA 95070 United States 20141 Kilbride Drive, Saratoga, California (72) Inventor Augustus P Lowell United States 94086 Sunnyvale, California Eurica Court 225 (72) Inventor Robert E. Alvarez United States of America 94043 Laura Lane No. 2369, Mountain View, California (72) Inventor Daniel Jay Latilin Ashuma operation view, Cypress Point Drive 505-298 No. F-term (reference) 4C093 AA11 AA18 CA37 CA39 CA50 EA04 EB17 EC24 FF46 5B057 AA08 BA03 CH02 CH07 CH14 DA16

Claims (33)

[Claims] An X-ray is passed through an object to be imaged, the X-ray is detected, and data on the object is formed from the detected X-ray. The data is obtained from a plurality of depths in the object. An image generation method, comprising: selecting a set of data including information and corresponding to a plurality of depths in the object to generate a display image. 2. The method of claim 1, wherein said X-rays are detected using a detector element array. 3. The method of claim 1, wherein said data comprises a plurality of image planes. 4. The method according to claim 3, wherein each of said plurality of image planes comprises an image pixel array. 5. The method according to claim 1, wherein said data comprises a plurality of image slices. 6. The method according to claim 5, wherein each of said plurality of image segments comprises a plurality of voxels. 7. The method of claim 1, wherein the selection of the set of data is made based on an intensity of the selected data. 8. The method of claim 7, wherein said selected data is selected with a maximum intensity. 9. The method of claim 7, wherein the selected data is selected with a minimum intensity. 10. The method of claim 1, wherein the selection of the set of data is made based on a relative contrast of the selected data. 11. The method of claim 1, wherein the selection of the set of data is performed based on a level of detail in the selected data. 12. The method according to claim 1, wherein the selection of the set of data is based on the energy of the selected data in a particular spatial frequency range. 13. The method of claim 1, wherein said data is formed by a computed tomography system. 14. The method of claim 1, wherein the data is formed by an inverse geometry x-ray imaging system. 15. A first processing module in which image pixel information at a plurality of depths in an object to be imaged is generated, and a specific display corresponding to various depths in the object from the image pixel information. An image restoration system comprising: a second processing module for selecting image pixel information; and a display device having a display image composed of the specific image pixel information. 16. The image restoration system according to claim 15, wherein:
A processing module configured to receive information corresponding to a position of one of a plurality of positions in the object and to generate partial image pixel information; An image restoration system comprising: one or more column units that receive the partial image pixel information from a unit and generate the image pixel information. 17. The image restoration system according to claim 15, wherein the first
The image restoration system further comprising a selection processing unit that receives the image pixel information and outputs pixel selection information. 18. The image restoration system according to claim 17, wherein said selection processing unit includes a gradient generator. 19. The image restoration system according to claim 17, wherein said selection processing unit includes a selection input filter and a selection output filter. 20. The image restoration system according to claim 15, further comprising a scanning beam X-ray imaging system, wherein the scanning beam X-ray imaging system generates X-ray transmission information, and wherein the first processing module includes the X-ray transmission information. An image restoration system characterized by receiving line transmission information. 21. The image restoration system according to claim 15, further comprising a computer tomographic image system, wherein the computer tomographic image system generates X-ray transmission information, and wherein the first processing module generates the X-ray transmission information. An image restoration system characterized by receiving. 22. The image restoration system according to claim 15, wherein said image pixel information includes image pixel array information of a plurality of image planes. 23. The image restoration system according to claim 15, wherein said image pixel information includes image voxel arrangement information of a plurality of image segments. 24. The image restoration system according to claim 15, wherein the display image is based on a gradient between neighboring image pixels. 25. The image restoration system according to claim 15, wherein the display image is based on a maximum intensity of an image pixel. 26. The image restoration system according to claim 15, wherein the display image is based on a minimum intensity of an image pixel. 27. The image restoration system according to claim 15, wherein the display image is based on a contrast between neighboring image pixels. 28. A means for generating image pixel information at a plurality of depths in an object to be image-generated, and selecting image data for display corresponding to various depths in the object from the image pixel information. And a display for displaying an image formed from the display data. 29. The image restoration system according to claim 28, wherein the means for generating the image pixel information receives information corresponding to a position of one of a plurality of positions in the object, and An image restoration system comprising: means for generating image pixel information; and means for receiving the partial image pixel information and generating the image data. 30. The image restoration system according to claim 28, further comprising means for receiving said image pixel information and outputting pixel selection information. 31. An image restoration system according to claim 30, wherein said pixel selection information is output using gradient information. 32. The image restoration system according to claim 28, wherein the means for selecting the image data is based on projection data having a maximum intensity at coordinates in the object. 33. The image restoration system according to claim 28, wherein the means for selecting the image data is based on projection data having a minimum intensity at coordinates in the object.