JP2019030575A - Ultrasound image processing apparatus and volume data processing method - Google Patents

Abstract

To allow that when a color tissue image is generated from a tissue image and a depth image generated by volume rendering, a condition of color processing (in particular, a depth parameter) is optimized automatically.SOLUTION: Site-of-interest data is identified by a template matching process on volume data 200 (refer to Code 210). Tissue-of-interest data 202 is extracted using an association relationship with the site-of-interest data. A representative depth is identified as the depth of the deepest voxel in the tissue-of-interest data 202. A condition of color processing for weighted synthesis of two color tissue images based on the tissue image is defined based on the representative depth (refer to Code 222).SELECTED DRAWING: Figure 5

Description

  The present invention relates to an ultrasonic image processing apparatus, and more particularly to volume data processing.

  The ultrasonic image processing apparatus is an apparatus that processes ultrasonic volume data (hereinafter, simply referred to as “volume data”), and includes, for example, an ultrasonic diagnostic apparatus, an information processing apparatus, and the like. In the ultrasonic image processing apparatus, a tissue image (three-dimensional ultrasonic image) that three-dimensionally represents a target tissue is generated by volume rendering based on volume data.

  In volume rendering, normally, a plurality of rays (line of sight) are set for volume data. A plurality of rays correspond to a plurality of pixels constituting the tissue image. For each individual ray, a rendering operation is performed, resulting in a pixel value. More specifically, on each ray, output light amount calculation (voxel calculation) using opacity (opacity) in units of voxels is repeatedly executed from the calculation start point along the depth direction. The output light amount when a predetermined end condition is satisfied, that is, the output light amount at the calculation end point is set as a pixel value (luminance). For example, when the opacity accumulation value reaches 1.0 or a predetermined value close to it in the middle of the rendering operation along the ray, the rendering operation ends (first end condition). Further, when the output light amount calculation reaches the last voxel, the rendering calculation ends (second end condition). The above-mentioned accumulated opacity value is generally obtained by sequentially adding the opacity used there for each output light amount calculation. Various calculation formulas are known as output light quantity calculation formulas. A plurality of luminances obtained for a plurality of rays constitute a tissue image.

  As a special volume rendering method developed from the general volume rendering method as described above, a depth image (depth map) is generated together with a tissue image (luminance map), and a color tissue image ( A volume rendering method for generating a color three-dimensional ultrasound image) is known. A specific example will be explained. As a result of the rendering calculation in units of rays, the output light amount at the calculation end point, that is, the luminance, and the depth of the calculation end point (distance from the calculation start point to the calculation end point) are specified. Is done. As a result, a tissue image composed of a plurality of luminances corresponding to a plurality of rays and a depth image composed of a plurality of depths corresponding to the plurality of rays are generated. The depth image typically represents a three-dimensional morphology of the surface of the target tissue (including a corresponding surface (eg, an inner surface along the surface)). In color processing of a tissue image, a color tissue image excellent in depth expression is generated by changing the color for each pixel (for example, a combination of RGB values) according to the depth for each pixel. For example, on the color tissue image, the fetus in the uterus is expressed in a warm color, and the uterine wall (including the placenta) existing behind the fetus is expressed in a cold color. Such an expression makes it easy to distinguish the fetus from the uterine wall on the color tissue image.

  Patent Documents 1 and 2 propose coloring according to the depth of the calculation end point in ultrasonic image processing. In addition, as patent applications related to the present application and unpublished patent applications at the time of this application, Japanese Patent Application No. 2016-135188, Japanese Patent Application No. 2017-47085, Japanese Patent Application No. 2017-54751, and Japanese Patent Application No. 2017-92853. There is a number.

Japanese Patent No. 5117490 Japanese Patent No. 5525930

  When generating a color tissue image based on a tissue image and a depth image generated by volume rendering, it is troublesome to manually set the color processing conditions. For example, if you want to express the fetus in a warm color and want to express the uterine wall on the back side of the fetus in a cold color, the depth parameter that makes the color coding criterion is the depth of the voxel at the deepest position in the fetus. It is desirable to specify a depth (or a specific depth instead). However, in general, it is difficult for the user to immediately specify the optimum value of the depth parameter, and therefore it is necessary to find the optimum value through trial and error. This is a heavy burden on the user. If the depth parameter is not properly specified, a part of the fetus will be displayed in a cool color, giving the viewer an unnatural impression, or the uterine wall behind the fetus A part of the image is displayed in a warm color similar to that of the fetus, which makes it difficult to visually distinguish the fetus from the uterine wall.

  An object of the present invention is to reduce the burden on the user when setting color processing conditions, or to enable optimization of a depth parameter that defines color processing conditions.

  The ultrasonic image processing apparatus according to the embodiment includes an analysis unit that obtains a representative depth of a target tissue in a tissue included in the three-dimensional space by analyzing volume data obtained by transmitting and receiving ultrasonic waves to and from the three-dimensional space. And, based on the volume data, an image generation means for generating a tissue image representing the tissue in three dimensions, and a depth image representing the depth of the surface of the tissue, and based on the depth image. Means for generating a color tissue image by performing color processing on the tissue image, and color processing means for determining a condition for the color processing based on the representative depth.

  According to the above configuration, the tissue image and the depth image are generated based on the volume data, and the color tissue image is generated based on them. Color processing conditions for generating a color tissue image are determined based on the representative depth of the target tissue. As described above, since the color processing condition can be set adaptively based on the information about the depth of the target tissue, the user's burden can be reduced, or the color processing condition suitable for the target tissue can be set.

  In the embodiment, the image generation unit generates a tissue image and a depth image by volume rendering for volume data. For example, a plurality of rays are set for the volume data, and a rendering operation is executed in the depth direction along each ray for each ray. A tissue image and a depth image are formed from a plurality of rendering calculation results corresponding to a plurality of rays. A depth image is an image representing the surface of a tissue, and the concept of the surface of a tissue may include a surface or a layer corresponding to the surface.

  In the embodiment, the color processing unit generates the color tissue image by weighting and combining the first color tissue image and the second color tissue image based on the tissue image based on the depth image. The color processing conditions are the weighted synthesis conditions. The first color tissue image is, for example, a tissue image expressed by a warm color, and the second color tissue image is a tissue image expressed by, for example, a cool color. These weighted synthesis conditions are set according to the representative depth obtained for the target tissue. For example, when the representative depth indicates the deepest position in the target tissue, the entire target tissue is expressed by the first color, and the non-target tissue existing behind the target tissue is expressed by the second color. The In such a case, it becomes easy to visually distinguish the target tissue and the non-target tissue.

  In the embodiment, the weighting synthesis condition is determined according to a weighting function, the weighting function includes a depth parameter, and the depth parameter is set based on the representative depth. The depth parameter is, for example, a parameter that designates a boundary between a first depth range in which the first color expression is prioritized and a second depth range in which the second color expression is prioritized. However, the depth parameter may be another parameter that defines an action that varies along the depth direction in the weighting function.

  In the embodiment, the analysis means is based on means for searching for attention site data included in the volume data, means for extracting attention tissue data as data having a connection relation with the attention portion data, and the attention tissue data. And means for determining the representative depth. According to this configuration, even if it is difficult to extract the target tissue data directly from the volume data, the target region data included in the target tissue data is searched, and the target tissue data is obtained based on the target region data. Can be extracted. The representative depth is determined based on the target tissue data.

  In the embodiment, the means for searching the attention site data searches the attention site data by a template matching process using a two-dimensional or three-dimensional template. According to the template matching process, it is possible to search attention site data using structural or morphological features of the attention site and its surroundings.

  In an embodiment, the target tissue data is data corresponding to a fetal head, and the target site data is data corresponding to a target site included in the fetal head. The site of interest is a facial part included in the head, such as eyes, nose, mouth, and the like.

  In the embodiment, the representative depth is a depth of a representative voxel included in data corresponding to the head of the fetus. In the embodiment, the representative voxel is a voxel included in data corresponding to the head of the fetus and located at the deepest position. According to this configuration, the entire fetal head is easily expressed by a specific color.

  The image processing method according to the embodiment obtains a representative depth for a target tissue in a tissue included in the three-dimensional space by analyzing volume data obtained by transmitting and receiving ultrasonic waves to and from the three-dimensional space; and Generating a tissue image representing the tissue three-dimensionally and a depth image representing the depth of the surface of the tissue by volume rendering with respect to the volume data; and the tissue image based on the depth image. A step of generating a color tissue image by performing color processing, and a step of determining a condition of the color processing based on the representative depth.

  The image processing method can be realized as a software function or a hardware function. In the former case, a program for executing the image processing method is installed in an ultrasonic image processing apparatus such as an ultrasonic diagnostic apparatus or an information processing apparatus via a network or a portable storage medium.

  According to the present invention, it is possible to reduce the burden on the user when setting color processing conditions. Alternatively, the depth parameter that defines the color processing conditions can be automatically optimized.

1 is a block diagram illustrating an ultrasonic image processing apparatus according to an embodiment. It is a figure which shows the structure | tissue image and depth image which are produced | generated by volume rendering. FIG. 2 is a diagram illustrating a configuration example of a color processing unit illustrated in FIG. 1. It is a figure which shows an example of a weighting function. It is a conceptual diagram for demonstrating the production | generation of a color organization image. It is a figure which shows the structural example of the analysis part shown in FIG. It is a figure which shows the 1st example of a template. It is a flowchart which shows the operation | movement at the time of parts search. It is a figure which shows extraction of the surface data from volume data. It is a figure which shows the 2nd example of a template. It is a figure which shows the 3rd example of a template. It is a flowchart which shows a series of image processing.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing an ultrasonic diagnostic apparatus as an ultrasonic image processing apparatus. The ultrasonic diagnostic apparatus is an apparatus that is installed in a medical institution such as a hospital and forms and displays an ultrasonic image based on a reception signal obtained by transmitting / receiving ultrasonic waves to / from a living body. In the present embodiment, the subject is a pregnant woman, and a color tissue image (color three-dimensional ultrasound image) is generated in which the head (particularly the face) of the fetus is three-dimensionally and expressed in color. Note that the tissue image is physically a two-dimensional image.

  In FIG. 1, the ultrasonic diagnostic apparatus has a 3D probe 10. The 3D probe 10 includes a probe head, a cable, and a connector. The connector is detachably attached to the ultrasonic diagnostic apparatus main body. A probe head is abutted on the abdominal surface of the pregnant woman. The probe head has a 2D array transducer composed of a plurality of vibration elements arranged two-dimensionally. An ultrasonic beam is formed by the 2D array transducer and electronically scanned in a two-dimensional manner. As a result, a three-dimensional data capturing space (three-dimensional space) including the fetus in the uterus (particularly its face) is formed. Volume data is acquired from the three-dimensional space.

  Instead of the 2D array transducer, a 1D array transducer and a mechanism for mechanically scanning the 1D array transducer may be provided. When such a configuration is adopted, the beam scanning surface is formed by the 1D array transducer, and the beam scanning surface is scanned by the mechanical scanning of the 1D array transducer. As a result, the above-described three-dimensional space is formed.

  The transmission / reception unit 12 is an electronic circuit that functions as a transmission beam former and a reception beam former. At the time of transmission, a plurality of transmission signals are supplied from the transmission / reception unit 12 to the 2D array transducer. As a result, a transmission beam is formed. At the time of reception, the reflected wave from the living body is received by the 2D array transducer. Thereby, a plurality of reception signals are output in parallel from the 2D array transducer to the transmission / reception unit 12. The transmission / reception unit 12 includes a plurality of amplifiers, a plurality of A / D converters, a plurality of delay circuits, an addition circuit, and the like. In the transmission / reception unit 12, a plurality of reception signals are subjected to phasing addition (delay addition) to form beam data corresponding to the reception beam.

  In general, the volume data is composed of a plurality of received frame data, and each received frame data is composed of a plurality of beam data. Each beam data is composed of a plurality of echo data arranged in the depth direction.

  The beam data processing unit 14 is means for applying various kinds of processing to each beam data, and is constituted by an electronic circuit. Various processing includes detection, logarithmic conversion, correlation processing, filter processing, and the like. Individual echo data constituting the beam data output from the beam data processing unit 14 is stored in the 3D memory 16. The 3D memory 16 has a data storage space corresponding to the three-dimensional space in the living body. Individual echo data is mapped to memory coordinates corresponding to the three-dimensional coordinates from which it was obtained. In the illustrated configuration example, coordinate conversion is performed when each echo data is written to the 3D memory 16. The coordinate transformation may be performed in two stages. For example, volume data may be configured by executing two-dimensional coordinate transformation on individual received frame data to generate individual cross-sectional data and placing them in the data storage space.

  In the illustrated configuration example, the control unit 34 exhibits a coordinate conversion function. Of course, a dedicated processor for coordinate conversion may be provided. In any case, the 3D memory 16 stores volume data acquired from the three-dimensional space in the living body.

  When displaying a tomographic image, the cross-section data cut out from the volume data in the 3D memory 16 is sent to the display processing unit 18. The cross-sectional data is slice data, which consists of a plurality of two-dimensionally arranged echo data (a plurality of voxel values or a plurality of pixel values). A plurality of tomographic images may be displayed simultaneously. In order to form a three-dimensional image, the volume data read from the 3D memory 16 is sent to the image processing unit 20.

  The image processing unit 20 is processing means for processing volume data. The image processing unit 20 is an electronic circuit including an image processing processor. The image processing unit 20 includes a tissue image generation unit 22 that functions as a tissue image generation unit, a depth image generation unit 24 that functions as a depth image generation unit, and an analysis unit 26 that functions as an analysis unit. Each of them may be constituted by a dedicated processor. In addition, they may be realized as functions of an operation program executed in each CPU. The tissue image generation unit 22 and the depth image generation unit 24 constitute an image generation unit or a volume rendering unit.

  Specifically, the tissue image generation unit 22 comes out of a virtual viewpoint with respect to volume data or partial data in a three-dimensional region of interest set for the volume data (which can also be called volume data). Set multiple rays and perform rendering operation in units of rays. Specifically, the output light amount calculation is repeatedly executed in units of voxels from the calculation start point to the depth direction along the ray.

In the output light quantity calculation, for example, the following equation (1) is used. The following expression (2) is an expression that defines the input light amount C _INi for the i-th voxel. Expressions (1) and (2) are both examples. Various expressions are known as output light quantity calculation expressions.

In the above equation (1), C _OUTi represents the light amount calculation result for the i-th voxel, that is, the output light amount. As described above, C _INi represents the input light amount for the i-th voxel, and actually it is the output light amount of the (i−1) -th voxel. O _i is an opacity defined for the i-th voxel. e _i is an echo value (voxel value) for the i-th voxel. Oi is determined from ei according to the opacity function (or opacity curve).

  In the rendering operation, two end conditions are defined. In the process of performing a rendering operation along a certain ray, first, when the accumulated opacity value reaches 1.0 or a predetermined value close thereto, the voxel operation for that ray is terminated (first termination). conditions). Secondly, when the voxel calculation is completed for the last voxel on the ray, the voxel calculation for the ray ends (second end condition). A tissue image (luminance map) is composed of a plurality of pixel values corresponding to a plurality of rays. The luminance image may be smoothed. Instead of the above two end conditions, one or more other end conditions may be defined. In addition to the above two end conditions, one or more other end conditions may be defined.

  The depth image generation unit 24 specifies the depth of the voxel (that is, the distance from the calculation start point to the calculation end point) for each ray when the end condition is satisfied. A depth image is composed of a plurality of depths corresponding to a plurality of rays. The depth image generally represents a three-dimensional shape of the surface of a tissue (fetal face) or a surface corresponding to the surface (for example, a surface in the surface layer). The specific content of the depth image varies depending on the form of the opacity function and the size of the threshold compared with the accumulated opacity value. Exceptionally, a three-dimensional region of interest or part of the bottom of the volume data can appear in the depth image.

  The analysis unit 26 functions as part search means, face data extraction means, and representative depth determination means. Here, when the imaging target and the target tissue are defined, the part can be said to be a target part (search target part) belonging to the target tissue. Specifically, the analysis unit 26 uses a specific part (eyes, nose, mouth, etc.) included in the fetal face and a template simulating the surroundings in the volume data to identify the face data. Search for part data. For example, data corresponding to the eyes and the surroundings are searched. Subsequently, the analysis unit 26 extracts face data using the connection relationship with the eye data in the volume data. In that case, a three-dimensional labeling process, a three-dimensional paint process, etc. are utilized. Further, the analysis unit 26 identifies the voxel at the deepest position from the viewpoint, that is, the voxel having the largest distance in the face data, and determines the depth of the target voxel as the representative depth. This representative depth is used in automatic setting of a depth parameter to be described later.

  The memory 28 stores a plurality of templates corresponding to a plurality of parts. Each template is two-dimensional data or three-dimensional data simulating a part to be searched. The size of the fetus (for example, the size of the face) and the position of the fetus (for example, the position of the face) may be specified using the template. Further, the orientation of the fetus (for example, the orientation of the face) may be specified. In the matching process using the template (template matching process), the position and size of the template are changed in stages. In some cases, the orientation of the template is also changed in stages.

  The depth of the specific voxel included in the searched part, the average depth of the searched part, the average depth of the face data, and the average depth, not the depth of the voxel at the deepest position in the face data Alternatively, a depth that is a certain distance away may be determined as the representative depth. In any case, if the depth parameter is determined according to the face data or the part data, the burden on the user can be reduced.

  The display processing unit 18 functions as a tomographic image forming unit, a graphic image generating unit, an image synthesizing unit, a color processing unit, and the like. The display processing unit 18 is realized by a dedicated processor or a function exhibited by an operation program executed on the CPU. The display processing unit 18 includes a color processing unit 30 that functions as a color processing unit.

  The color processing unit 30 generates a color tissue image by weighting and combining the front color tissue image and the back color tissue image according to the depth image and the function of generating the front color tissue image and the back color tissue image based on the tissue image. And having a function. The front color tissue image is a first provisional color tissue image. In the present embodiment, the front color is a color for expressing the fetus and is, for example, a warm color. The back color tissue image is a second provisional color tissue image. In this embodiment, the back color is a color for expressing the uterine wall existing on the back side of the fetus, for example, a cold color. The front color tissue image is composed of a plurality of front color pixel values corresponding to a plurality of pixels, and the back color tissue image is composed of a plurality of back color pixel values corresponding to a plurality of pixels. Note that the color tissue image may be generated directly from the tissue image and the depth image without actually generating the front color tissue image and the back color tissue image.

  The weighted synthesis of the front color tissue image and the back color tissue image is performed according to a weighting function (coloring function). The weighting function has a depth parameter in this embodiment. In this embodiment, the depth parameter is a parameter that defines a boundary between a depth range on the near side where the expression by the front color is prioritized and a depth range on the back side where the expression by the back color is prioritized. . However, this is merely an example, and a depth parameter that defines other conditions may be adopted. In this embodiment, the depth parameter is automatically and adaptively set according to the representative depth for the tissue of interest. Specifically, the representative depth is substituted into the depth parameter. Thereby, the burden on the user is reduced, and the depth parameter can be optimized according to the target tissue.

  A color tissue image generated by the weighted synthesis is displayed on the display 32. The display 32 is configured by an LCD or an organic EL device. Incidentally, in the color tissue image, the luminance value of each pixel is determined by the output light quantity at the end of the calculation. An organization close to the viewpoint is expressed brightly, and an organization far from the viewpoint is expressed darkly.

  The control unit 34 functions as a control unit that controls each component illustrated in FIG. 1 and includes a CPU and an operation program. An operation panel 36 is connected to the control unit 34. The operation panel 36 has various input devices such as a trackball, a switch, and a keyboard.

  FIG. 2 schematically shows a method for generating a tissue image and a depth image. In the illustrated example, a plurality of rays 42 are set in parallel to each other for the three-dimensional region of interest (volume data) 40. XYZ represents an orthogonal coordinate system. The Z direction is the depth direction. An arrow denoted by reference numeral 44 indicates the projection direction from the viewpoint. A plurality of rays 42 are set in parallel to the direction. Note that the viewpoint can be set at an arbitrary position around the volume data.

  Reference numeral 40A denotes a calculation start surface. It is drawn as a plane. Actually, it is possible to freely change the shape or the like of the calculation start surface 40A. In addition to the calculation start surface 40A, a reference surface for defining the depth (distance) may be determined. On each ray 42, the rendering calculation is executed in the depth direction (Z direction in the illustrated example) from the calculation start point on the calculation start surface 40A. Specifically, the output light amount calculation shown in the above equation (1) is repeatedly executed for each voxel existing on the ray 42 (or for each voxel generated by the interpolation calculation). Each time the output light quantity is calculated, the opacity is cumulatively added. When any of the above-described termination conditions is satisfied, the output light amount calculation ends, that is, the rendering calculation ends. As a result, the calculation end point 48 is specified for each ray 42. The calculation end point 48 corresponds to, for example, the fetal surface (or a certain depth in the surface layer). The output light amount at the end of the rendering operation is mapped on the memory as the luminance. A tissue image 52 is formed as a mapping result of a plurality of luminances corresponding to the plurality of rays 42. It is a black and white volume rendering image.

On the other hand, the depth of the calculation end point 48 (distance from the calculation start point to the calculation end point) is mapped on the memory in units of rays. A depth image 50 is formed as a mapping result of a plurality of distances corresponding to the plurality of rays 42. As shown as the gray scale 54, in FIG. 2, the minimum depth to the maximum depth is expressed as a change in luminance. Of course, this is only an example. The tissue image 52 and the depth image 50 have the same two-dimensional pixel array. Based on the tissue image 52 and the depth image 50, a color tissue image is generated.

  FIG. 3 shows a configuration example of the color processing unit 30 shown in FIG. The color processing unit 30 includes two conversion tables 70 and 72 and a weighting synthesizer 74. The first conversion table 70 is a front color conversion table, and is a lookup table that determines a front color (RGB value) Cf as a fetal expression color according to the luminance I. For example, an orange color that can feel warmth is used as the front color. An organization existing on the side close to the viewpoint is represented by such a color. The second conversion table 72 is a back color conversion table, and is a lookup table that determines a back color (RGB value) Cb as a color for expressing the uterine wall according to the luminance I. For example, as the color for expressing the uterine wall, a blue color can be used to increase the perspective by mixing with the color for expressing the fetus, or to visually distinguish the uterine wall from the fetus. The first conversion table 70 and the second conversion table 72 each operate on a pixel basis. A plurality of color pixel values sequentially output from the first conversion table 70 constitute a front color tissue image, and a plurality of color pixel values sequentially output from the second conversion table 72 constitute a back color tissue image.

  The weighting synthesizer 74 has a weighting function (coloring function) 76. The weighting function 76 obtains the combined color C by giving two weights defined from the depth Z to the colors Cf and Cb and adding them together. The weighting function 76 has a depth parameter Zs and a range parameter ΔZ, and the operation of the weighting function 76 is changed by manipulating the values substituted for them.

  The parameter setting unit 78 automatically sets the depth parameter Zs based on the representative depth Zr. Specifically, the representative depth Zr is substituted for the depth parameter Zs. As described above, the boundary between the depth range preferentially expressed by the front color and the depth range preferentially expressed by the back color can be changed in the depth direction by changing the depth parameter Zs. The range parameter ΔZ is manually adjusted as necessary. Note that the range parameter ΔZ may be automatically set based on the representative depth Zr and the like. As a result of the weighted synthesis, a color tissue image composed of a plurality of color pixel values is generated. Note that the color C may be identified immediately from the combination of the luminance I and the depth Z. However, when such an aspect is realized by a look-up table, the configuration becomes considerably complicated. On the other hand, according to the configuration shown in FIG. 3, the configuration can be simplified.

  FIG. 4 shows an example of the weighting function. The weighting function includes a function 80 that defines a weight for Cf and a function 82 that defines a weight for Cb. The horizontal axis indicates the depth Z. The vertical axis represents the weight.

  In the range from 0 to Z1, the weight for Cf is 1.0, and the weight for Cb is 0. In the range from Z1 to Z2, the weight for Cf changes from 1.0 to 0, and the weight for Cb changes from 0 to 1.0. In the range from Z2 to MaxZ, the weight for Cf is 0, and the weight for Cb is 1.0. Here, Zs is a midpoint between Z1 and Z2, and corresponds to the cross point of the two functions 80 and 82 in the illustrated example. ΔZ is a range between Z1 and Z2. A range from 0 to Zs is a depth range where priority is given to color expression by the front color, and a range from Zs to MaxZ is a depth range where priority is given to color expression by the back color.

  When the depth parameter Zs is increased or decreased, the functions 80 and 82 are shifted in the horizontal direction. Therefore, the depth parameter Zs can also be referred to as a shift parameter. When the range parameter ΔZ is increased or decreased, the existence ratio of two colors having a weight of 1.0 in the color tissue image changes. Therefore, the range parameter ΔZ can also be referred to as a contrast parameter.

  When the depth of the voxel at the deepest position included in the face data is set as the depth parameter Zs, the tissue existing on the front side (viewpoint side) from there is represented exclusively by the front color, and from there The organization existing on the side (non-viewpoint side) is represented exclusively by the back color. Therefore, the problem that the face of the fetus is expressed in an unnatural color can be avoided, and the problem that the uterine wall on the back side is expressed in the same color as the fetus can be avoided. That is, the depth parameter Zs can be optimized according to the existence range of the fetal face in the depth direction. For example, a parameter that designates Z2 as the depth parameter may be employed.

  FIG. 5 schematically illustrates the entire volume data processing according to the present embodiment. The volume data 200 includes fetal data (face data) 202 and uterine wall data 204. A plurality of rays 206 are set for the volume data 200, and a rendering operation is executed for each ray. As a result, a tissue image 214 and a depth image 216 are generated. On the other hand, as will be described in detail later, a part search using the template 210 is performed on the volume data 200. Thereby, attention data (for example, data corresponding to the eyes and the surroundings, data corresponding to the mouth and the surroundings) included in the face data is specified. Subsequently, face data as target tissue data including target region data is extracted from the volume data using a connection relationship with the target region data. Further, the deepest voxel 212 included in the face data is specified. The depth at which the voxel 212 exists is set as the representative depth Zr. The representative depth Zr is substituted into the depth parameter Zs. Next, a front color tissue image 218 and a back color tissue image 220 are generated from the tissue image 214. As indicated by reference numeral 222, the images 218 and 220 are weighted and synthesized according to the depth image 216, thereby generating a color tissue image 224. In the weighting synthesis 222, a weighting function having a depth parameter Zs is used.

  FIG. 6 shows a configuration example of the analysis unit 26 shown in FIG. The analysis unit 26 includes a template matching unit 94, a face data extraction unit 96, and a representative depth determination unit 98. The template matching unit 94 is a module for searching attention site data (specific part data) included in the volume data 90 using the template 92. The face data extraction unit 96 extracts the face data including the data using the connection relation with the attention site data in the volume data. The representative depth determination unit 98 identifies the deepest voxel in the extracted face data and defines the depth as the representative depth Zr. Note that it is difficult to extract face data when the face data is in contact with the uterine wall data. In that case, the representative depth of the target tissue data may be determined from the depth of the target region data.

  Next, template matching processing will be described with reference to FIGS. In FIG. 7, volume data 100 is acquired from a three-dimensional space in the mother body. The volume data 100 includes fetal data, amniotic fluid data, uterine wall data, and the like. Surface data (slice data) 102A sequentially cut out from the volume data 100 is sequentially processed. As indicated by reference numeral 103, the surface data 102 may be reduced.

  In the illustrated example, the surface data 102A to be processed is the uterine wall data 106 having a generally high luminance, the amniotic fluid data 108 having a generally low luminance, and the head data (face data) having a generally high luminance. ) 104. The head data 104 includes a circular eyeball portion 116 whose luminance is lower than that of the surrounding area. The number of gradations is, for example, 256. In this case, the luminance changes between 0 and 255.

  The template 110 is a two-dimensional template in the illustrated example, and has a rectangular shape. Specifically, the template 110 includes a circular area 112 and a surrounding area 114. The circular region 112 is a region that simulates an eyeball, and the surrounding region 114 is a region that simulates a tissue around the eyeball. The brightness of the circular area 112 is low, for example, 0. The brightness of the surrounding area 114 is high, for example 100. Each numerical value described in this specification is an example.

  Pattern matching using a template is executed at each position in the individual surface data 102A. Specifically, the degree of similarity (evaluation value) between the surface data 102A and the template 110 is calculated at each scanning position while raster scanning the template 110 with respect to the surface data 102A. The best evaluation value is specified for each surface data 102A, and the evaluation value and the three-dimensional coordinate information of the template 110 at the time when it is obtained are temporarily stored. After the above processing is applied to all the surface data 102A, the best evaluation value is determined from the plurality of stored evaluation values. The best evaluation value corresponds to the right eye or the left eye. The three-dimensional coordinate information corresponding to the best evaluation value (for example, the surrounding area 114 of the template 110 in the best matching state) is used as a starting point when extracting face data.

  In the pattern matching process, a dedicated template is prepared for each region (or region type) of interest. As will be described later, it is also possible to use a template that simultaneously specifies a plurality of parts. Pattern matching is performed while changing the size, brightness, rotation angle, etc. of the template as necessary. It is also possible to use a binarized template. Prior to pattern matching, thinning processing may be performed on the surface data. Since the template 110 shown in FIG. 7 basically has no angle dependency, its rotation is unnecessary. In the above pattern matching, since a two-dimensional template is applied to a two-dimensional image, the processing time can be shortened. Of course, as shown later, a three-dimensional template may be applied to the volume data.

  FIG. 8 shows a template matching process as a flowchart. In S30, the initial value S0 is substituted for the variable S that defines the template size. In S32, the template size is set according to the variable S. In S34, the initial value Z0 is assigned to the variable Z that defines the position in the Z direction (surface data arrangement direction), and the surface data is cut out in the position Z in S36. In S38 and S40, initial values Y0 and X0 are substituted for variables Y and X that define the position in the Y direction and the position in the X direction, respectively. In S42, a template having the above size is set at the position of the coordinates (X, Y) in the surface data at the position of the cut-out Z. In S44, pattern matching is executed, that is, an evaluation value is calculated. The evaluation value is recorded together with the three-dimensional coordinate information.

  If it is determined in S46 that X has not reached Xmax, X is incremented by 1 in S48, and the processes after S42 are repeatedly executed. If it is determined in S46 that X has reached Xmax, S50 is executed. If it is determined in S50 that Y has not reached Ymax, Y is incremented by one in S52, and the processes after S40 are repeatedly executed. If it is determined in S50 that Y has reached Ymax, S54 is executed. If it is determined in S54 that Z has not reached Zmax, Z is incremented by 1 in S56, and the steps after S36 are repeatedly executed. If it is determined in S54 that Z has reached Zmax, S58 is executed. When it is determined in S58 that S has not reached Smax, S is incremented by 1 in S60, and the processes after S32 are repeatedly executed. If it is determined in S58 that S has reached Smax, S62 is executed. In S62, the three-dimensional coordinate information of the template when the best evaluation value is reached is specified. In the illustrated example, S62 is the final process.

  Although not shown in the flowchart shown in FIG. 8, the brightness of the template may be changed stepwise. Or you may change the direction or attitude | position of a template in steps. If there are n degrees of freedom in the movement of the template, the n parameters may be individually varied in stages.

  Next, a second example of template matching processing will be described based on FIGS. In the volume data 120 shown in FIG. 9, the cross section 122 is repeatedly set while changing the position and angle. Thereby, the cross-section data 128 is cut out for each cross-section 122 (see reference numeral 124). In order to reduce the amount of calculation, the volume data 120 is reduced before the section data 128 is cut out, or the cut out section data 128 is reduced (see reference numeral 126).

  FIG. 10 shows one cross-sectional data. The cross-sectional data includes a fetal head image 130. In the illustrated example, the fetal head image 130 corresponds to a horizontal section of the fetal head (a section parallel to the central axis of the head and facing the front). The fetal head image 130 includes three low-luminance portions, specifically, a left-eye portion (low-luminance portion) 142, a right-eye portion (low-luminance portion) 144, and a nasal cavity portion (low-luminance portion) 146. It is. Each of these portions 142, 144, 146 is a substantially circular region. The size of the nasal cavity portion 146 is slightly smaller than the size of the left eye portion 142 and the right eye portion 144. The nasal cavity portion 146 is at a position shifted to the mouth side with respect to the line connecting the left eye portion 142 and the right eye portion 144.

  A tool for simultaneously searching these three parts is a template 132. The template 132 is an image composed of three circular areas 134, 136, 138 and their surrounding areas. Pattern matching is repeatedly executed while the position of the cross section, the angle of the cross section, the position of the template 132, the size of the template 132, and the inclination angle of the template 132 are varied independently. The three-dimensional coordinates, the tilt angle, and the size (magnification) of the template 132 at the time when the best correlation value is obtained are specified. In FIG. 10, a template 140 is shown which is adapted to the low-intensity part corresponding to the two eyes and the low-intensity part corresponding to the nasal cavity (that is, which produced the best correlation value). The size of the template 140 at that time represents the size of the face of the fetus, which is used as size information as necessary. Specifically, the distance between two eyes may be used as size information.

  As illustrated in FIG. 11, the attention site data included in the volume data 148 may be searched using the three-dimensional template 150. When attention site data is detected, face data is extracted by a three-dimensional labeling process or other methods based on the attention site data. The representative depth is specified based on the face data. A depth parameter is automatically set based on the representative depth.

  FIG. 12 shows an operation example of the apparatus shown in FIG.

  In S76, the ultrasonic beam is two-dimensionally scanned and volume data is acquired. In S78, a plurality of attention site data included in the volume data are searched while a plurality of templates corresponding to a plurality of attention regions are used. Based on all or part of them, face data (head data) is extracted using the connection relationship. In S80, a plurality of rays are set for the volume data, and a tissue image and a depth image are generated by executing a rendering operation for each ray. In the process, rendering conditions are optimized according to the region of interest. For example, an opacity curve is selected according to the region of interest, and the smoothing condition is changed. As a result, the image quality of tissue images and the like can be improved.

  In S82, the representative depth is specified based on the face data extracted in S78. In S84, a weighting function is defined based on the representative depth. That is, the depth parameter included in the weighting function is set based on the representative depth. In S86, a front color tissue image and a back color tissue image are generated from the tissue image, and a color tissue image is generated by weighting and synthesizing them.

  In S88, the enlargement ratio of the color tissue image is determined based on the size information, and the center of the zoom range is specified based on the eye position and the like. A display image is generated by applying zoom processing based on the enlargement ratio to the zoom range. In S90, the display image is displayed on the screen.

  As described above, the template matching process can be used for various purposes. Although it is generally difficult to specify face data directly, focus on the part data contained in the face data (relatively easy to specify) and extract it first to extract the face data from the part data. It is relatively easy to extract. When the face data is connected to the uterine wall data, the part data can be used as data representing the face data. According to the above-described embodiment, it is possible to reduce the burden on the user when setting the depth parameter that defines the color processing condition, and it is possible to optimize the depth parameter.

  In the above-described embodiment, one representative depth is specified from the target tissue data and used, but a plurality of representative depths may be specified from the target tissue data and used. .

20 image processing unit, 22 tissue image generation unit, 24 depth image generation unit, 26 analysis unit, 30 color processing unit, 50 tissue image, 52 depth image, 70 first conversion table, 72 second conversion table, 74 weighting Synthesizer, 76 function (weighting function), 78 parameter setting unit, 94 template matching unit, 96 face data extraction unit, 98 representative depth determination unit.

Claims (9)

By analyzing volume data obtained by transmitting and receiving ultrasonic waves to and from the three-dimensional space, an analysis means for obtaining a representative depth for a target tissue in the tissue included in the three-dimensional space;
Based on the volume data, a tissue image that three-dimensionally represents the tissue, and an image generation unit that generates a depth image that represents the depth of the surface of the tissue;
Means for generating a color tissue image by color processing the tissue image based on the depth image, and color processing means for determining the color processing conditions based on the representative depth;
An ultrasonic image processing apparatus comprising: The apparatus of claim 1.
The color processing unit generates the color tissue image by weighting and combining the first color tissue image and the second color tissue image based on the tissue image based on the depth image,
The color processing condition is the weighting composition condition.
An ultrasonic image processing apparatus. The apparatus of claim 2.
The weighting synthesis condition is determined according to a weighting function,
The weighting function includes a depth parameter;
The depth parameter is set based on the representative depth;
An ultrasonic image processing apparatus. The apparatus of claim 1.
The analysis means includes
Means for searching attention site data included in the volume data;
Means for extracting target tissue data as data having a connection relationship with the target region data;
Means for determining the representative depth based on the tissue of interest data;
An ultrasonic image processing apparatus comprising: The apparatus of claim 4.
The means for searching the attention site data searches the attention site data by template matching processing using a two-dimensional or three-dimensional template.
An ultrasonic image processing apparatus. The apparatus of claim 4.
The noted tissue data is data corresponding to the fetal head,
The attention area data is data corresponding to an attention area included in the fetal head.
An ultrasonic image processing apparatus. The apparatus of claim 6.
The representative depth is a depth of a representative voxel included in data corresponding to the fetal head.
An ultrasonic image processing apparatus. The apparatus of claim 7.
The representative voxel is a voxel that is included in data corresponding to the fetal head and is in the deepest position.
An ultrasonic image processing apparatus. A program for executing an ultrasonic image processing method in an ultrasonic image processing apparatus,
The ultrasonic image processing method includes:
A step of obtaining a representative depth for a target tissue in a tissue included in the three-dimensional space by analyzing volume data obtained by transmitting and receiving ultrasonic waves to and from the three-dimensional space;
Generating a tissue image representing the tissue three-dimensionally and a depth image representing the depth of the surface of the tissue by volume rendering on the volume data;
Generating a color tissue image by color processing the tissue image based on the depth image, the step of determining the color processing conditions based on the representative depth;
The program characterized by including.