JP2580748B2 - Image processing device for medical images - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[Industrial applications]

The present invention relates to an image processing apparatus for medical images that combines a surface image of a target and a CT image.

[Prior art]

2. Description of the Related Art Conventionally, for medical diagnosis, a tomographic image (CT image) on a specific slice plane of a patient has been taken using an X-ray CT (computer tomography) apparatus or the like. In this case, if data on a large number of slice planes is collected, not only CT images on each of the slice planes can be obtained, but also data on a large number of three-dimensional small cubes can be collected. Cross section of
A CT image can be created, or a specific tissue such as soft tissue or bone can be extracted to reconstruct a three-dimensional image representing its surface. Therefore, conventionally, while creating a surface image of a certain object, setting an arbitrary cross section that crosses the object, creating a CT image at the cross section, synthesizing these, as if cutting at an arbitrary cross-sectional view Display of a three-dimensional image in a form in which is visible.

[Problems to be solved by the invention]

However, depending on how the cut plane is set, the cut plane is not always visible from the viewpoint. Therefore, conventionally, the direction (viewpoint) of the three-dimensional image is restricted, or the setting of the cut plane is restricted. However, since the composition of the surface image and the CT image on the cut surface is allowed only when the cut surface is all visible, the degree of freedom is small. This is a major problem because it reduces the diagnostic efficiency in actual clinical medical diagnosis. The present invention, when synthesizing a surface image of an object viewed from a certain viewpoint and a CT image at an arbitrary cut plane crossing the object, automatically determines whether or not the cut plane is visible from the viewpoint. By synthesizing the image accordingly, it is possible to freely and rapidly synthesize the image regardless of the surface image viewed from any direction or the cut surface regardless of the setting. Accordingly, it is an object of the present invention to provide a medical image processing apparatus capable of improving medical diagnostic efficiency.

[Means for Solving the Problems]

In order to achieve the above object, in a medical image image processing apparatus according to the present invention, a first storage unit storing three-dimensional image data representing a surface of an object and also including distance information from a viewpoint, Which stores three-dimensional image data also including distance information from the same viewpoint as described above.
Storage means, and a third storage means for storing CT data at the cut plane.
A fourth storage unit in which a surface image of the object subjected to shading processing is stored, and a unit for reading distance information of a pixel at the same point from the first and second storage units. Means for determining which is farther from the viewpoint by comparing the distance information obtained, and CT data corresponding to the pixel on the cut plane when the cut plane is determined to be closer to the fourth storage means. Means for overwriting the surface image stored in the storage area.

[Action]

The three-dimensional image data also includes distance information from the viewpoint, and is also called a depth map. Therefore, when trying to combine an image representing the surface of a target and an image representing a cut plane, three-dimensional image data relating to them and including distance information from a specific viewpoint is used. By comparing the information, it is possible to determine which is far from the viewpoint. If the cut surface is farther away, it will be hidden behind the target surface image, but if the cut surface is closer, it will be visible from the viewpoint. Therefore, when it is determined that the cut plane is closer from the viewpoint, if the CT data corresponding to the pixel in the cross-sectional view is overwritten on the surface image representing the target, the cross-section is only applied to the visible part of the cut plane. Will be created. Therefore, when synthesizing the surface image of the target from a certain viewpoint and the CT image at an arbitrary cut plane crossing the target, it is automatically determined whether or not the cut plane is visible from the viewpoint. Since the image synthesis is performed only on the visible portion, the image synthesis can be freely performed regardless of the direction of the surface image viewed from any direction or the setting of the cut surface. Moreover, image synthesis can be performed by comparing the distance information and by determining whether or not data has been overwritten based on the comparison result, thereby enabling high-speed processing and display. Therefore, it is very useful for actual clinical medical diagnosis.

【Example】

Next, an embodiment of the present invention will be described with reference to the drawings. In FIG. 1, a third memory is stored in the image memory 1.
As shown in the figure, data of a three-dimensional image 21 representing the surface of the target is stored. The three-dimensional image data includes planar image data viewed from a certain viewpoint and distance information from the viewpoint. On the other hand, in the image memory 2, for example, a three-dimensional image 3 representing each of three cut planes as shown in FIG.
1, 32, and 33 data are stored. The three-dimensional image data also includes two-dimensional image data obtained by viewing three cut planes from a certain viewpoint and distance information from the viewpoint.
The viewpoint is the same as the viewpoint of the three-dimensional image 21 of the target surface. The image memory 3 stores CT images (CT value distribution images) 41, 42, and 43 formed by collecting CT values on these three cut planes as shown in FIG.
The image memory 7 stores a surface image 22 as shown in FIG. This surface image 22 is a 3D image from the same viewpoint as FIG.
Although it is a two-dimensional image, shadowing (shooting) and the like are performed in consideration of the direction of light and the direction of the viewpoint. The image shown in FIG. 2 is an image obtained as a final result. As described below, as a result of writing to the image memory 7, the image in the image memory 7 initially shown in FIG. Will be an image like Here, the address generator 4 controls these image memories 1,
An address signal designating a few identical pixels (i, j) is generated (i is a column address, and j is a line address). Then, the pixel values Z (i, j), Z '(i, j), and V (i, j) are read from the image memories 1, 2, and 3, respectively. Z (i, j) and Z '(i, j) are distance information from the viewpoint, and V (i, j) is a CT value. Each pixel value of the image in the image memory 7 is ZV (i, j). These Z (i, j) and Z '(i, j) are sent to the comparator 5 where their magnitude relations are compared. That is, it is determined whether the following condition 1 is satisfied. Z (i, j) ≧ Z ′ (i, j)... Condition 1 When Condition 1 is satisfied, a signal of “1” is generated from the comparator 5, whereby the AND gate 6 is opened and the write signal is The data is sent to the image memory 7 and the writing state is enabled. The image memory 7 stores an image 22 on which shading processing and the like of the target surface have been performed. When the image memory 7 is set in such a writable state, the address (i, i) specified by the address signal from the address generator 4 is stored. Overwriting of V (i, j) to j) (erasing the original data and writing over it) is performed. This will be described for four representative points, P1, P2, P3, and P4. It is assumed that P1 exists in a region where the target surface is present but there is no cut surface. P2 has both a target surface and a cut surface,
It is assumed that the target surface exists in a region closer to the viewpoint.
In addition, P3 and P4 have both a target surface and a cut surface, but the cut surface exists in a region closer to the viewpoint. Since P1 and P2 do not satisfy the condition 1, the image memory 7
, V (i, j) is not overwritten and ZV (i, j)
Will remain. On the other hand, for P3 and P4, condition 1
Is satisfied, V (i, j) is overwritten in the image memory 7. That is, ZV (i, j)
Is replaced by V (i, j). Therefore, different address signals are sequentially generated from the address generator 4 and such comparison is performed for those pixels, and as a result, the CT images 41, 42, and 43 on the cut surface are obtained.
Is controlled, CT images 41, 42, and 43 on the cut surface are gradually formed in the target surface image 22 (although the CT image 41 is not written in this example because it is hidden). When such an operation is completed for all the pixels, eventually, as shown in FIG.
CT image on the cut plane only for the part visible from the viewpoint 4
2, 43 will be represented.

【The invention's effect】

According to the image processing apparatus for medical images of the present invention, when combining an image of a target surface from a certain viewpoint and a CT image at an arbitrary cut surface that crosses the target surface, the cut surface is viewed from the viewpoint. Automatically determine whether the image is visible or not and perform image synthesis accordingly, so that image synthesis can be performed freely regardless of the direction of the target surface image or the settings of the cut surface. Can be. As a result, the degree of freedom in image synthesis is greatly increased, and the operator can freely synthesize images without being restricted by the restrictions. In addition, since it is determined whether or not the CT image on the cut plane is hidden by the distance information from the viewpoint, the images are synthesized, so that the processing speed is high and the synthesized image can be displayed quickly. Since the free image observation is possible and the processing can be immediately performed at high speed, the efficiency of actual clinical medical diagnosis can be greatly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention, FIG. 2 is a diagram showing a final synthesized image, FIG. 3 is a diagram of a surface image of a target stored in an image memory 1, and FIG. FIG. 5 is a view of an image representing a cut plane stored in the memory 2, FIG. 5 is a view of a CT image on the cut plane stored in the image memory 3, and FIG. 6 is a view of an object stored in the image memory 7 in an initial state. It is a surface image subjected to shading processing. 1, 2, 3, 7 ... image memory, 4 ... address generator, 5 comparator, 6 ... AND gate, 21 ... 3 on the target surface
Dimensional image, 22 ... Surface image of the object subjected to shading,
31-33: 3D image of cut plane, 41-43: On cut plane
CT image.

Claims (1)

(57) [Claims] 1. A first storage means for storing three-dimensional image data including distance information from a viewpoint, representing a surface of an object, and also including distance information from the same viewpoint, representing a cut plane. A second storage unit in which the three-dimensional image data is stored, a third storage unit in which the CT data at the cut plane is stored, and a fourth storage unit in which the target surface image subjected to the shading process is stored. Storage means, means for reading distance information of a pixel at the same point from the first and second storage means, means for comparing the read distance information to determine which is farther from the viewpoint,
Means for overwriting CT data corresponding to the pixel on the cut plane when it is determined that the cut plane is closer to the surface image stored in the fourth storage means. Image processing device for medical images.