JP2007135841A - Ultrasonic image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accelerate image processing while preventing the decline of image quality when generating images from three-dimensional data of ultrasonic echo. <P>SOLUTION: A plurality of rays 50 and 52, etc., are set to the three-dimensional data 30 for the inside of a diagnostic object obtained by three-dimensional ultrasonic diagnosing; rendering processing is performed along the respective rays, and two or more partial images (the image indicated by pixels 34 and 38, etc., displayed by shadows and the image indicated by pixels 36 and 40, etc., displayed in white) which do not overlap at least partially are generated. To the rays corresponding to the respective partial images, the rendering processing different from the one of the rays corresponding to the other partial images is set. Then, the obtained partial images are combined and the entire image is generated. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a technique for generating a three-dimensional image based on three-dimensional data of ultrasonic echoes.

  In ultrasonic diagnosis, there are cases in which the inside of a diagnosis target such as a living body is three-dimensionally diagnosed. For example, in an ultrasound diagnosis for a mother's womb, a three-dimensional diagnosis of a fetus is often performed and a moving image is generated and displayed. In general, enormous calculations are required for processing such three-dimensional data or processing four-dimensional data including time.

  FIG. 21 is a diagram for explaining a general image generation processing mode in the ultrasonic image processing apparatus. Here, two rendering processes are performed based on the three-dimensional echo data 500 obtained by transmitting and receiving ultrasonic waves. That is, the rendering 1 processing unit 502 and the rendering 2 processing unit 504 both generate images for the entire region using different rendering algorithms. Then, a weight of α is given to the image obtained by the rendering 1 processing unit 502, and a weight of β is given to the image obtained by the rendering 2 processing unit 504, and the addition is performed by the adder 506. It is processed. The composite image obtained as a result is displayed on a display unit 508 such as a display.

  FIG. 22 is a diagram illustrating another aspect of conventional image generation processing. Also in this ultrasonic image processing apparatus, the three-dimensional echo data 500 is processed by two rendering processing systems. One is a processing system by the rendering processing unit 510, and the other is a processing system including a smoothing filter 512 and a rendering processing unit 514. The smoothing filter 512 is a filter that performs smoothing processing on the original three-dimensional echo data 500 to create smooth three-dimensional data. Smoothing is performed, for example, by averaging with a size of 3 × 3 × 3. A rendering processing unit 514 that performs rendering processing on the smoothed data performs rendering according to the same rendering algorithm as the rendering processing unit 510. Therefore, in this aspect, each image calculated by the same rendering algorithm from two three-dimensional data having different resolutions is added to generate a display image.

  In each aspect shown in FIGS. 21 and 22, two different images are superimposed to generate an image. That is, in the example of FIG. 21, images obtained by different rendering algorithms are superimposed, and in the example of FIG. 22, images obtained by using the same rendering algorithm from three-dimensional data having different smoothness are superimposed. When such processing is performed, a high-quality image having the characteristics of each image can be obtained.

  However, in the processing modes shown in FIG. 21 and FIG. 22, in order to obtain one image, about twice as many operations are required as compared with the case of obtaining a single image. Therefore, it is difficult to quickly obtain an image with high spatial resolution. In addition, when a moving image is generated by this method, the frame rate is lowered.

  In addition, as a literature regarding the technique which visualizes the three-dimensional data of an ultrasonic echo, the following patent document 1 can be mentioned, for example. However, the technique of this document generates a single image by superimposing the Doppler image shaded according to the perspective and the tissue image, and is different from the aspect of the present invention in which the same kind of image is synthesized. .

Japanese Patent No. 3270158

  An object of the present invention is to establish an image processing technique for speeding up image processing while preventing deterioration in image quality when generating an image from three-dimensional data of ultrasonic echoes.

  Another object of the present invention is to develop a new technique for generating a three-dimensional ultrasonic image by rendering for each ray.

  The ultrasonic image processing apparatus of the present invention performs a rendering process along each ray by setting a plurality of rays for the three-dimensional data about the inside of the diagnosis target obtained by the three-dimensional ultrasonic diagnosis, A generating unit that generates two or more partial images that do not partially overlap, and a setting unit that sets a rendering process different from a ray corresponding to another partial image for a ray corresponding to each partial image, and Combining means for combining two or more partial images to generate an entire image.

  An ultrasonic image processing apparatus is an apparatus that performs image processing such as image generation and processing based on ultrasonic echo data. Image processing is realized by using hardware dedicated to image processing or by controlling hardware having a general-purpose arithmetic function by software. The three-dimensional data is obtained by transmitting an ultrasonic wave to a diagnostic object such as a living body and receiving the reflected wave. The ultrasonic image processing apparatus may be configured as an apparatus (ultrasonic diagnostic apparatus) having an ultrasonic transmission / reception function. The ultrasonic image processing device generates an image based on the three-dimensional data, and the generated image is typically displayed on a display device such as a liquid crystal display. The ultrasonic image processing apparatus may incorporate this display device.

  The generating means is a means for generating a partial image by projecting information of three-dimensional data of ultrasonic echoes obtained by three-dimensional measurement inside the diagnosis target onto an appropriate projection plane along the ray. That is, a ray (light ray) incident on the three-dimensional data through the projection plane is set, and at least a part of the three-dimensional data area through which the ray passes is reflected and rendered on the projection plane. Thereby, a partial image on the projection plane is generated. Here, the partial image is an image corresponding to one or a plurality of rays and is generated by the rendering process in the same mode. One partial image may be composed of a plurality of spatially separated regions. Each partial image is generated so that another partial image and all or part of the region do not overlap each other.

  The setting means sets a rendering process different from the rays corresponding to the other partial images for the rays corresponding to the partial images. Different rendering processes may be different rendering methods themselves, or different processing parameters in the same method, or by differentiating the resolution of the 3D data used for processing. It may be.

  The synthesizing unit synthesizes the plurality of generated partial images to generate an entire image. The synthesis can be performed by weighted addition for each pixel, for example. It is also possible to perform processing such as smoothing between neighboring pixels and regions. The area ratio (pixel number ratio) and weight ratio of each partial image in the composition may be equal or unequal.

  According to this aspect, a plurality of partial images that do not overlap at least partially are synthesized to form an entire image. Therefore, the entire image is a composite reflection of the characteristics of each rendering process corresponding to the partial image. Moreover, since each partial image does not overlap at least in part, the processing time can be shortened compared to the case where two images covering the entire area are simply generated and combined. As a result, it is possible to realize image formation that maintains and improves the image quality and speeds up image processing.

  In one aspect of the ultrasonic image processing apparatus of the present invention, in the setting means, the spatial expansion of the three-dimensional data contributing to the rendering result corresponds to the ray corresponding to the other partial image. Sets a different rendering process. Spatial spread may be spread in the ray direction or spread in a direction perpendicular to the ray. When the spatial extension is large, a smoothed partial image is formed, and when the spatial extension is small, a partial image reflecting a fine structure is formed.

  In one aspect of the ultrasonic image processing apparatus of the present invention, the setting means sets, for a ray corresponding to each partial image, a rendering process having the same algorithm and different parameters from the rays corresponding to the other partial images. To do. In the aspect of the ultrasonic image processing apparatus of the present invention, the setting means sets a rendering process having a different algorithm for the ray corresponding to each partial image from the ray corresponding to the other partial image. . Examples of algorithms include Volume Rendering, VOL, MIP, X-ray, and Shaded. Examples of parameters include transparency and opacity.

  In one aspect of the ultrasonic image processing apparatus of the present invention, the generation unit generates two or more partial images whose regions do not overlap each other. Thereby, the generation efficiency of partial images can be increased.

  In one aspect of the ultrasonic image processing apparatus of the present invention, the generating means generates two partial images having a checkered pattern that complement each other with one pixel as a unit of pattern formation. In other words, the checkered pattern is a chessboard pattern. One is white and the other is arranged according to a black cell pattern, and each cell has a size of one pixel.

  In one aspect of the ultrasonic image processing apparatus of the present invention, the synthesizing means performs a smoothing process to synthesize the partial images smoothly. The smoothing process is a process of reducing a value gap between surrounding pixels (regions), and is typically performed by a filter process that weights and averages the values of a plurality of pixels.

  In one aspect of the ultrasonic image processing apparatus of the present invention, when the time series data of the entire image is generated by repeating the generation unit and the synthesis unit, the setting of the rendering process is repeatedly changed in the setting unit. The time-series data (moving image data) may be generated by performing generation means and synthesis means on the time-series three-dimensional data, or the viewpoint, projection plane, etc. one after another with respect to the three-dimensional data at a certain time. It may be generated by performing the generating means and the synthesizing means while changing. In generating the time series data, the setting of the rendering process in the setting unit is changed to another setting when one frame or a plurality of frames are processed. As a result, it is possible to avoid that a part of the entire image always has the same characteristics, and to improve the image quality of the time series data.

  The ultrasonic image processing apparatus according to the present invention generates an image by performing rendering processing along a plurality of rays on three-dimensional data about the inside of a diagnosis target obtained by three-dimensional ultrasonic diagnosis. And means for setting a rendering process different from that of other rays for one or a plurality of rays.

  According to the present invention, it is possible to efficiently generate a three-dimensional image having a plurality of rendering processing features from three-dimensional data of ultrasonic echoes.

  FIG. 1 is a schematic block diagram illustrating a configuration example of the ultrasonic diagnostic apparatus 10. The ultrasonic diagnostic apparatus 10 includes a control circuit unit 12 that controls the apparatus based on a program and user input. The mechanical scanning control unit 14 mechanically controls ultrasonic scanning of the probe 16 that performs ultrasonic transmission / reception with respect to the diagnosis target, and performs three-dimensional scanning. Further, the transmission control unit 18 transmits the ultrasonic transmission signal to the probe 16 at an appropriate timing and waveform. Thereby, the probe 16 provided with the piezoelectric element transmits an ultrasonic wave, receives an echo reflected from the diagnosis target, and converts it into an electric signal.

  The receiving circuit 20 is a circuit that amplifies the received signal output from the probe 16. The signal processing unit 22 generates three-dimensional data based on the received signal. The three-dimensional data is rendered in the three-dimensional image processing unit 24, and the obtained image is displayed on the display of the display unit 26.

  This ultrasonic diagnostic apparatus 10 is used for the purpose of diagnosing the inside of a human body, for example. As a specific example, there is an aspect in which a diagnosis of a mother's womb is performed and a fetal figure is displayed as a moving image. When this moving image display is performed, each process is repeatedly executed at a high speed, including ultrasonic transmission / reception by the probe 16.

  FIG. 2 is a diagram schematically showing a processing mode in the three-dimensional image processing unit 24 shown in FIG. The three-dimensional image processing unit 24 generates an image on the projection plane 32 based on the three-dimensional data 30 generated by the signal processing unit 22. Here, each pixel of the projection surface 32 is divided into two in a checkered pattern, and a setting with high and low opacity is respectively set. For example, the pixel 34 has high Opacity, and the adjacent pixel 36 has low Opacity. The pixel 36 is set so that the pixel 38 with high Opacity is adjacent to the pixel 36 and the pixel 40 with low Opacity is adjacent to the pixel 38.

  For each pixel on the projection plane 32, a ray perpendicular to the projection plane 32 (that is, a ray with the viewpoint set to infinity) is set. More specifically, the rays 50, 54,. . . Are set and the rays 52, 56,. . . Is set. Then, at least a part of the values of the three-dimensional data 30 is referred to along such a ray, and a pixel value in the pixel is determined.

Here, as an example, a procedure for determining a pixel value in the VOL algorithm will be described. The pixel value output in the calculation process is C _out , the pixel value input is C _in , each value on the three-dimensional data ray is e _i , and the opacity given to each e _i is α _i (0 ≦ α _i ≦ In the case of 1), in the VOL algorithm,
C _out = C _in (1−α _i ) + α _i × e _i (1)
Are sequentially executed from the near side to the far side along the ray. And
Σα _i ≧ 1 (2)
When the above condition is satisfied, the calculation for that ray is terminated, and C _out is output as a pixel value.

The opacity α _i is set to a different value according to the level of Opacity. FIG. 3 is a graph for explaining setting of Opacity. The horizontal axis represents the echo value in the three-dimensional data in 256 gradations from 0 to 255, and the vertical axis represents the Opacity given to each echo value. In this graph, five setting curves 60, 62, 64, 66, and 68 are drawn. For example, the uppermost setting curve 60 shows the relationship between the echo value set to obtain the most opaque image and Opacity. For the echo value of 255 having the strongest reflection intensity, An Opacity of 1 is given to indicate no permeability. On the other hand, the lowermost setting curve 68 is for obtaining the most transparent image, and has a relatively high transparency even for the echo value of 255 having the strongest reflection intensity. Opacity is given.

  FIG. 4 compares which ray of a plurality of sample points on a ray is referred to when the Opacity is high (FIG. 4A) and when the Opacity is low (FIG. 4B). FIG. In both figures, it is assumed that the echo intensity of each data is the same.

  4A, sample points 72, 74,... Existing along a highly opaque ray 70. In FIG. . . Of these, four sample points 72, 74, 76, 78 from the surface are used for rendering. On the other hand, in FIG. 4B, the sample points 102, 104,. . . , Eleven sample points 102,. . . 110 are used for rendering. This difference is due to the fact that four sample points are sufficient to satisfy the condition of equation (2) in the case of high capacity, whereas eleven sample points are required in the case of low capacity. As a result, in the case of FIG. 4B where the Opacity is low, the value of the sample point that has entered deeply from the surface is also referred to determine the luminance value, and the depth is deeper than in the case of FIG. An image that transparently reflects is formed.

  In this way, two types of partial images constituting the checkered pattern are generated. FIG. 5 is a diagram illustrating an algorithm for calculating the entire image 134 by combining the partial image 130 occupying one area of the checkered pattern and the partial image 132 occupying the other area. Here, the pixel value is determined by performing a smoothing process between adjacent pixels. Specifically, for each pixel, the pixel value of the entire image 134 is determined by adding the pixel value of the pixel with the weight of α and the pixel value of the adjacent pixel with the weight of β. Thereby, the gap of the pixel value resulting from the difference in the generation processing mode of the partial images 130 and 132 can be reduced.

  In addition, the synthesis can be performed in various ways. For example, for each pixel, the pixel value can be determined by adding pixel values of 3 × 3 pixels or 5 × 5 pixels centered on the pixel with appropriate weighting. This synthesizing operation is excellent in that it can be smoothed isotropically. Further, it is also possible to simply adopt the pixel values of both partial images as the pixel values of the entire image without performing such smoothing processing. This aspect is excellent in that original information of pixel values of individual pixels is not lost.

  FIG. 6 is a diagram summarizing the processing flow described above. That is, each of the rays 1, 2,. . . , N are rendered respectively. A rendering algorithm or parameter is set for each ray (144). The partial image thus obtained is subjected to the processing of filtering 146 for smoothing and synthesizing and sent to the display unit (148).

  The algorithm or parameters for each ray can be varied over time. FIG. 7 is a diagram for explaining this aspect. The same components as those in FIG. 6 are given the same reference numerals to simplify the description. Here, when new three-dimensional data 140 is generated, a switching signal indicating that is inputted (150), and the rendering parameter or algorithm is changed each time (144). Typically, two parameters or algorithms are swapped with each new input. Then, the projection image 152 and the storage result (154) of the previous projection image, for example, are added together by the adder 156, and filtering 146 and display can be performed (148).

  Note that the projection image 152 obtained by the rendering 142 may be directly processed by the filtering 146 and sent to the display unit (148). Thereby, an image with a fine temporal resolution can be output. Further, in this case, the gap between the frames can be expected to be alleviated by a filter action applied unconsciously in the optic nerve system of the viewer.

  The aspect of the rendering process described above can be variously changed. For example, two partial images can be generated according to a pattern that is not a checkered pattern. In particular, when the area ratio of both partial images is set to other than 1: 1, it is generally necessary to select an arrangement pattern in which both images are sufficiently evenly distributed in place of the checkered pattern at least in the region of interest. . Also, when generating partial images by three, four or more rendering processes, it is necessary to select a pattern in which these partial images are arranged in a balanced manner.

As an aspect in which the rendering algorithm is changed in each partial image instead of the rendering parameter, for example, a combination example of the VOL algorithm and the Xray algorithm can be given. That is, one partial image is formed in accordance with the VOL algorithm shown in Expression (1) and Expression (2), and the other partial image is _defined as C _out = 1 / N × Σ ₁ ^N e _i (3)
Is formed according to the Xray algorithm defined by Thereby, the whole image which combined the image which shows the surface structure of a diagnostic object, and the image which shows the average structure of the depth direction of a diagnostic object can be obtained. In addition, it is also effective to appropriately combine each algorithm such as Volume Rendering, VOL, MIP, X-ray, and Shaded.

  Subsequently, FIGS. 8 to 20 show results of image processing performed on a trial basis. The three-dimensional data used is data about the fetal head region obtained by performing ultrasonic diagnosis on the mother's womb.

  8 to 13 are diagrams for explaining a mode of generating and synthesizing two partial images using different algorithms. FIG. 8 shows the result of visualizing the surface of the fetus based on the Shaded algorithm for all the pixels on the projection plane. In this figure, the vicinity of the head of the fetus seen from the left side slightly behind is expressed with high resolution, and the structure of the face such as the eyes, nose and mouth, and the presence of the left shoulder and left hand can be confirmed. However, in this image, fine irregularities are seen on the surface of the fetus due to the noise of ultrasonic diagnosis, and objects that are thought to be caused by floating substances in the womb are expressed in front of the forehead etc. . For this reason, there is a problem that the mother who sees the image cannot feel the fetus cute.

  FIG. 9 shows the result of visualizing the fetal surface using the Xray algorithm for all pixels on the projection plane. This image is generated using the same ray as that in FIG. 8, and expresses information about the thickness of the fetus. As a difference from FIG. 8, it can be pointed out that the detailed structure is not reflected and only a rough structure can be recognized, and the nose and hand structures lacking depth are not expressed.

  FIG. 10 shows an image obtained by adding the images of FIGS. 8 and 9 with a weight of 0.5. In this image, the irregularities on the surface of the fetus are expressed relatively smoothly, and the facial expression of the fetus is gentle. However, this image needs to be rendered twice as usual in order to obtain the diagrams of FIGS. 8 and 9, which is not preferable from the viewpoint of computational complexity.

  FIG. 11 is an image formed by cutting and fitting the images of FIG. 8 and FIG. 9 into a checkered pattern. In this image, the irregularities on the surface of the fetus are relatively smooth as in FIG. 10, and the fetus is expressed in a cute manner. In addition, this figure can be generated in half the processing time in the case of FIG. 10, and is particularly excellent in generating moving images. However, in the structure portion that is not represented in FIG. 9, a fine checkered pattern is visible, and the image quality is slightly degraded.

  When a checkered pattern becomes a problem, it is effective to synthesize both figures by applying a smoothing filter. FIG. 12 shows an example in which a 3 × 3 pixel smoothing filter is applied to the image of FIG. In this case, the checkerboard pattern is reduced to a slightly conspicuous level, and the fine structure of the image in FIG. 11 is sufficiently expressed. FIG. 13 is an example in which a smoothing filter of 5 × 5 pixels is applied to the image of FIG. In this case, the checkerboard pattern is almost completely eliminated. However, the fine structure is slightly blurred compared to the case of FIG. The presence / absence of the filter operation or the shape of the filter need not always be set to the same setting, and may be appropriately changed according to the size of the structure to be expressed.

  Next, a mode in which two partial images are generated with the same algorithm (Shaded algorithm) and different parameters and synthesized will be described with reference to FIGS. FIG. 14 is an image obtained by setting Opacity low, and FIG. 15 is an image obtained by setting Opacity high. When the Opacity is low, since the surface of the fetus is deeply reflected, the surface of the fetus is relatively smooth in FIG. On the other hand, when the Opacity is high, an image of the fetus is created only by ultrasonic echoes from the vicinity of the surface of the fetus. Therefore, in FIG. 15, the image is conspicuous on the surface of the fetus and the suspended matter around the fetus.

  FIG. 16 is a view obtained by cutting and fitting FIGS. 14 and 15 into a checkered pattern. In this figure, an image obtained by averaging FIGS. 14 and 15 is obtained. However, as in the case of FIG. 11, there is a portion where the checkered pattern is conspicuous, and there is a possibility that it is recognized as a reduction in image quality when enlarged display is performed. This checkered pattern can be erased by a smoothing filter. FIG. 17 shows an image obtained by applying a 3 × 3 smoothing filter to the image of FIG. In the case of FIG. 12 in which a 3 × 3 smoothing filter is applied to FIG. 11, the checkered pattern partially remains, but in FIG. 17, the checkered pattern has almost completely disappeared. This is probably because the degree of approximation in FIGS. 14 and 15 is higher than the degree of approximation in FIGS. 8 and 9.

  Finally, a mode in which the checkerboard phase is exchanged frame by frame will be described with reference to FIGS. FIGS. 18 and 19 are diagrams showing an image (b) obtained by combining the image by the Shaded algorithm in FIG. 8 and the image by the Xray algorithm in FIG. 9 according to the phase shown in FIG. That is, the image in FIG. 18B is synthesized according to pattern A in FIG. 18A, and the image in FIG. 19B is synthesized in accordance with pattern B in FIG. In this case, the difference between the two images is small, and an image very similar to the image shown in FIG. 11 is obtained.

  Therefore, in creating a single still image, there is little advantage of switching the checkered pattern phase. However, when displaying as a continuous moving image, this replacement exhibits a great effect. For example, when the display interval of each frame is short, if the effect of the afterimage of the viewer is used, successive frames are added together and recognized. Therefore, the checkerboard pattern remaining in both images is canceled and recognized. It is also effective to generate an image of each frame by combining two consecutive images. FIG. 20 shows an image obtained by averaging the image of FIG. 18 and the image of FIG. Similar to the image of FIG. 10, this image is an image in which the Shaded algorithm and the Xray algorithm are averaged for all pixels. As long as the temporal movement is small, the checkerboard pattern disappears and a high-quality image is obtained.

It is a block diagram which shows the outline of the structural example of an ultrasonic diagnosing device. It is a schematic diagram which shows a mode that 3D data is projected on a projection surface along a ray. It is a figure which shows the relationship between an echo value and opacity for every parameter. It is a schematic diagram which shows the relationship between the sample point on a ray and opacity. It is a figure explaining the example of a synthesis | combination of a partial image. It is a figure explaining the example of the flow of a rendering process. It is a figure explaining the modification of the flow of a rendering process. It is the image produced | generated according to the algorithm of Shaded. It is the image produced | generated according to the algorithm of Xray. It is the image produced | generated by adding FIG. 8 and FIG. 9 0.5 each. 10 is an image generated by arranging the images of FIGS. 8 and 9 in a checkered pattern. 12 is an image generated by applying a 3 × 3 smoothing filter to the image of FIG. 11. 12 is an image generated by applying a 5 × 5 smoothing filter to the image of FIG. 11. It is an image generated with a low opacity setting by the Shaded algorithm. It is an image generated with the setting of high opacity by the Shaded algorithm. 16 is an image generated by arranging the images of FIG. 14 and FIG. 15 in a checkered pattern. 17 is an image generated by applying a 3 × 3 smoothing filter to the image of FIG. 16. FIG. 10 is a checkered pattern A (a) in which the images of FIGS. 8 and 9 are arranged, and an obtained image (b). FIG. 10 is a checkered pattern B (a) in which the images of FIG. 8 and FIG. 9 are arranged, and an obtained image (b). FIG. 20 is an image generated by adding 0.5 to FIG. 18 and FIG. 19. It is a figure which shows the example of a rendering process in the past. It is a figure which shows another example of a rendering process in the past.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Ultrasonic diagnostic apparatus, 12 Control circuit part, 14 Mechanical scanning control part, 16 Probe, 18 Transmission control part, 20 Reception circuit, 22 Signal processing part, 24 Three-dimensional image processing part, 26 Display part, 30 Three-dimensional data, 32 projection planes, 34-40 pixels, 50-56 rays, 60-68 setting curve, 70 high opacity rays, 100 low opacity rays.

Claims (9)

Two or more partial images that do not overlap at least partially by setting multiple rays and rendering along each ray for the three-dimensional data inside the diagnosis target obtained by three-dimensional ultrasonic diagnosis Generating means for generating
Setting means for setting a rendering process different from the ray corresponding to the other partial image for the ray corresponding to each partial image;
Combining means for combining two or more obtained partial images to generate an entire image;
An ultrasonic image processing apparatus comprising: The ultrasonic image processing apparatus according to claim 1,
The setting means sets, for the ray corresponding to each partial image, a rendering process in which the spatial expansion of the three-dimensional data contributing to the rendering result is different from that of the ray corresponding to the other partial image. Ultrasonic image processing device. The ultrasonic image processing apparatus according to claim 1,
An ultrasonic image processing apparatus characterized in that, in the setting means, for each ray corresponding to each partial image, a rendering process having the same algorithm and different parameters is set for the rays corresponding to the other partial images. The ultrasonic image processing apparatus according to claim 1,
An ultrasonic image processing apparatus characterized in that, in the setting means, a rendering process having an algorithm different from that of a ray corresponding to another partial image is set for a ray corresponding to each partial image. The ultrasonic image processing apparatus according to claim 1,
An ultrasonic image processing apparatus characterized in that the generation means generates two or more partial images whose regions do not overlap each other. The ultrasonic image processing apparatus according to claim 5,
An ultrasonic image processing apparatus characterized in that the generating means generates two partial images having a checkered pattern complementing each other using one pixel as a pattern forming unit. The ultrasonic image processing apparatus according to claim 1,
An ultrasonic image processing apparatus characterized in that the synthesizing means performs smoothing processing to smoothly synthesize the partial images. The ultrasonic image processing apparatus according to claim 1,
An ultrasonic image processing apparatus characterized in that, when the time series data of the entire image is generated by repeating the generating means and the synthesizing means, the setting of the rendering process is repeatedly changed in the setting means. Means for generating an image by performing rendering processing along each of a plurality of rays with respect to the three-dimensional data about the inside of the diagnostic object obtained by three-dimensional ultrasonic diagnosis;
Means for setting a different rendering process for one or more rays from other rays;
An ultrasonic image processing apparatus comprising: