JP2004350791A - Ultrasonic image processor and three-dimensional data processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To form the three-dimensional blood vessel image of attentional route inside a vascular plexus and to accurately perform the extraction even for a winding blood vessel route especially in an ultrasonic image processor and a three-dimensional data processing method. <P>SOLUTION: A blood vessel cross section extraction processing is executed at respective positions along a blood vessel route, and a reference plane Q<SB>i</SB>and a search plane R<SB>i</SB>parallel to each other are set in each blood vessel cross section extraction processing. A blood vessel cross section S<SB>i</SB>is extracted as a reference cross section on the reference plane Q<SB>i</SB>and the next search direction (search vector) U<SB>i+1</SB>is decided as a direction connecting the centroid G<SB>i</SB>and a temporary centroid K<SB>i</SB>obtained by a prescribed arithmetic operation on the search plane R<SB>i</SB>. Then, the reference plane Q<SB>i+1</SB>or the like is newly set in the next blood vessel cross section extraction processing on the basis of the search direction U<SB>i+1</SB>. In such a manner, the search direction and the reference plane are successively reset along the traveling form of the blood vessel. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an ultrasonic image processing apparatus and a three-dimensional data processing method, and particularly to a three-dimensional image processing of a blood vessel.
[0002]
[Prior art]
2. Description of the Related Art A technique for transmitting and receiving ultrasonic waves to and from a three-dimensional region in a living body and displaying a tissue in a living body as a three-dimensional image based on the three-dimensional data obtained thereby has been put to practical use in medical practice. In a conventional ultrasonic diagnostic apparatus, data corresponding to a blood vessel (blood flow) is extracted from three-dimensional data by utilizing the fact that the echo intensity of blood flow is smaller than the echo intensity of real tissue. A three-dimensional image of a blood vessel is formed based on the extracted data. Also, since a so-called Doppler shift occurs in an echo from a blood flow flowing in a blood vessel, a three-dimensional image of a blood vessel (blood flow) can be formed using Doppler information resulting therefrom.
[0003]
[Patent Document 1]
JP-A-7-178091
[0004]
[Problems to be solved by the invention]
However, it is relatively easy to identify a blood vessel of interest on a three-dimensional image for a part where one or several blood vessels are running in a state that is easy to distinguish from each other. When a vascular system or vascular network (one form of vascular group) is formed that is intricately intertwined spatially by branching out gradually from a thick blood vessel as shown in the figure, it can be expressed as a three-dimensional image. It is very difficult to clearly recognize the running state of each blood vessel portion. This is because even if an attempt is made to observe in detail a certain blood vessel portion existing inside the blood vessel network, other blood vessel portions exist around the portion, which hinders the observation. Therefore, even if an abnormal site exists in the middle of a specific blood vessel route, the abnormal site may not be clearly displayed as an image.
[0005]
Patent Literature 1 discloses a technique for displaying a blood vessel of interest and another blood vessel in a distinguishable manner. According to the technique, when a scan plane of many frames is configured to construct a three-dimensional space, a blood vessel cross section is extracted on a scan plane of a certain frame, and a blood vessel cross section is also extracted on a scan plane of the next frame. This is repeated. Then, a three-dimensional image of the blood vessel is formed by joining the blood vessel cross sections extracted on the scanning plane of each frame.
[0006]
However, with the method disclosed in Patent Document 1, first, it is very difficult to extract a meandering blood vessel. In other words, cross-section extraction is performed on each scanning plane in parallel, and if blood vessels are not orthogonal to each scanning plane, the extraction cannot be performed properly. Secondly, there is no configuration for coping with the bifurcation of a blood vessel and accurately extracting a blood vessel of interest.
[0007]
An object of the present invention is to enable accurate extraction of a meandering blood vessel.
[0008]
Another object of the present invention is to perform blood vessel extraction in response to branching of blood vessels.
[0009]
It is another object of the present invention to provide an image useful for blood vessel diagnosis.
[0010]
[Means for Solving the Problems]
(1) The present invention relates to an ultrasonic image processing apparatus for processing three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body. In the data space, search origin designating means for designating a search origin at a specific site in the blood vessel group, and extracting a blood vessel cross-section while sequentially setting a search vector from the search origin along a path of interest in the blood vessel group. And a search processing means for sequentially executing the search processing means, wherein the search processing means generates a blood vessel cross-section row including a plurality of blood vessel cross sections sequentially extracted along the attention path.
[0011]
According to the above configuration, in the three-dimensional data space, when the search origin is specified by the user or automatically for a specific part in the blood vessel group, the target route including the search origin (the blood vessel route as an extraction target) is included in the target route. A blood vessel cross-section extraction process is sequentially performed along the way, and individual blood vessel cross-sections are extracted as the process proceeds. In that case, the search vector is reset for each blood vessel cross-section extraction process, that is, the search direction is updated, so that it is possible to search for a blood vessel cross-section that is adapted to the traveling state of the route of interest. It is possible to increase. In particular, it is also possible to extract a winding path. Ideally, each search vector is set along or parallel to the center axis of the path of interest, and it is desirable to set search vector setting conditions so as to approach the ideal.
[0012]
According to the present invention, a plurality of blood vessel cross-sections constituting a blood vessel cross-section row are arranged in a non-parallel relationship in response to the direction of the search vector being adaptively set in each blood vessel cross-section extraction process. (However, if the path of interest is perfectly straight, the cross sections of a plurality of blood vessels are aligned in parallel with each other.) At the time of the image processing, the spatial resolution may be improved by applying an interpolation process or the like to the blood vessel section row.
[0013]
The ultrasonic image processing device according to the present invention may be incorporated in the ultrasonic diagnostic device, or may be configured as a computer that processes data output from the ultrasonic diagnostic device. Further, each function or each means can be configured as hardware or can be realized by software.
[0014]
Preferably, the search processing means sequentially sets the search vector according to a traveling mode of the attention route. If each search vector is set according to the traveling mode of the attention route, the extraction accuracy of the blood vessel cross section can be improved.
[0015]
Preferably, the search processing means sequentially sets a reference plane for extracting a blood vessel cross section in the three-dimensional data space based on the search vector set sequentially.
[0016]
The reference plane corresponds to a search area for searching for a blood vessel cross section of the path of interest. It is desirable that such a search area be set as orthogonal to the route of interest as possible. Therefore, the search vector is used as the reference. By the way, in a branch portion or a part where a plurality of blood vessels are running in parallel, a plurality of blood vessel cross sections appear on the reference plane. Further, when the reference plane is configured as a cut surface of the entire vascular network without limiting the size, a large number of blood vessel cross sections appear on the reference plane. In such a case, it is desirable to specify a blood vessel cross section corresponding to the attention path from the connection relationship with the previous blood vessel cross section.
[0017]
Preferably, the search processing means includes a reference plane setting means for setting an i-th reference plane orthogonal to the i-th search vector, and a blood vessel cross-section extraction for extracting an i-th blood vessel cross-section on the i-th reference plane. Means, search plane setting means for setting an i-th search plane at a position parallel to the i-th reference plane and at a unit distance from the i-th search direction, and i-th search plane on the i-th search plane. A blood vessel portion extracting means for extracting an i-th blood vessel portion having an overlapping relationship with the i-th blood vessel cross-section, and determining an (i + 1) -th search vector based on a correlation between the i-th blood vessel cross-section and the i-th blood vessel portion Search vector determining means for performing the search.
[0018]
In the above, i = 1, 2, 3,..., That is, i corresponds to the number of times or the degree of progress of the blood vessel cross-section extraction processing. The i-th reference plane is set as a plane orthogonal to the i-th search vector, and the i-th blood vessel cross section is extracted on the i-th reference plane. On the other hand, the i-th search plane is set at a position parallel to the i-th reference plane and separated by a unit distance in the i-th search direction. Here, the reference plane and the search plane correspond to a search region for a blood vessel cross section and a blood vessel portion, respectively. However, the i-th search plane is positioned as a preliminary plane for preliminary investigation for setting the (i + 1) -th reference plane in a correct direction.
[0019]
That is, on the i-th search plane, the i-th blood vessel portion having an overlapping relationship with the i-th blood vessel cross section is extracted. The i-th blood vessel portion is recognized as a portion in front of the i-th blood vessel cross section and connected thereto. Therefore, the (i + 1) th search vector is determined based on the correlation between the i-th blood vessel cross section and the i-th blood vessel portion. For example, a line connecting a point (centroid, center point, etc.) representing the i-th blood vessel cross section to a point (centroid, center point) representing the i-th blood vessel portion is recognized as being close to the traveling direction of the attention route. , It is desirable to set the direction of the (i + 1) th search vector.
[0020]
In the above configuration, in addition to the reference plane, a plane called a search plane for forward preliminary investigation is used, and the angle of the next reference plane is optimized by using a search result on such a search plane. Can be. The unit distance is desirably constant in each section extraction process, but may be variably set according to various situations. If the unit distance is set to, for example, about one voxel size in the three-dimensional data space, a blood vessel of interest can be extracted with high resolution. The reference plane and the search plane may cover the entire three-dimensional data space, or may be a fixed finite area.
[0021]
Preferably, the blood vessel part extracting means performs a labeling process on the i-th search plane to obtain an i-th labeling image including one or a plurality of closed regions, and the i-th labeling. The i-th blood vessel cross-section is projected on an image to specify a closed area having the largest overlap area with respect to the i-th blood vessel cross-section, and the projected i-th blood vessel in the specified closed area. Projection processing means for extracting a portion overlapping with the cross section as the i-th blood vessel portion. According to this configuration, it is possible to easily recognize the blood vessel portion existing in front of the already extracted blood vessel cross section.
[0022]
Since the closed region having the largest overlap area is selected, the search can be automatically advanced to the thickest main blood vessel path when the front is bifurcated or branched into many. As a modified example, when a plurality of closed regions in a superimposed relationship are confirmed, a labeling image or the like at that time may be displayed, and the user may be left to select a search direction.
[0023]
Preferably, the search vector determining means calculates a center of gravity of the i-th blood vessel section, a means of calculating a center of gravity of the i-th blood vessel section, and calculates the i-th blood vessel section from the centroid of the i-th blood vessel section. Means for determining the (i + 1) th search vector, with the direction passing through the center of gravity of the blood vessel portion as the vector direction. According to this configuration, it is possible to roughly recognize the blood vessel running direction as a direction connecting the center of gravity of the blood vessel section and the center of gravity of the temporarily extracted blood vessel portion, that is, it is possible to set a search vector in that direction.
[0024]
Desirably, the (i + 1) th reference plane corresponds to a plane obtained by rotating the i-th search plane about a predetermined point on the plane as a rotation center. The predetermined point is the center of gravity or the center point of the i-th blood vessel portion.
[0025]
Preferably, a search start direction specifying means for specifying a search start direction is included, and a first search vector is determined based on the search origin and the search start direction. Preferably, the image processing apparatus further includes a reference image forming unit that forms a two-dimensional or three-dimensional reference image based on the three-dimensional data, and the search origin and the search start direction are specified by a user using the reference image. The reference image may be a three-dimensional image representing the three-dimensional space. However, if it is difficult to specify the inside of the three-dimensional space, the reference image may be displayed on a two-dimensional tomographic image obtained by cutting the three-dimensional space at an arbitrary position or a predetermined position. It is desirable to specify the search origin. In this case, an arbitrary cut plane may be set so as to include the center axis of the blood vessel part of interest, and a search origin and a search direction may be designated on an arbitrary tomographic image.
[0026]
The search may be started in both the forward and backward directions from the search origin. In that case, it is desirable to set the search termination condition for each direction.
[0027]
(2) Further, the present invention relates to an ultrasonic image processing apparatus for processing three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body, wherein the ultrasonic image processing apparatus is configured by the three-dimensional data. In the three-dimensional data space, search origin designating means for designating a search origin at a specific site in the blood vessel group, and a search vector is sequentially set from the search origin along a path of interest in the blood vessel group while setting a search vector. Search processing means for sequentially executing the extraction; a storage unit for storing a blood vessel cross-sectional row composed of a plurality of blood vessel cross-sections sequentially extracted along the path of interest; and a target path based on the blood vessel cross-sectional row. And an image forming means for forming an ultrasonic image including:
[0028]
According to the above configuration, the path of interest is extracted as a blood vessel cross-section row, and various ultrasonic images can be configured using the blood vessel cross-section row. The ultrasonic image may be a three-dimensional image of only the attention path, or may be a three-dimensional image in which the attention path has emerged from the background with high luminance or the attention path has emerged from the monochrome background by coloring. Good.
[0029]
Preferably, the ultrasonic image is a three-dimensional image expressing the attention path as a single stroke. By applying the three-dimensional labeling method, it is possible to extract all the areas connected to a certain point, but in such a case, all the blood vessels connected to the blood vessel of interest are similarly extracted, and as a result, observation of the blood vessel of interest May be inhibited. According to the present invention, for example, in a multi-branched blood vessel group, a single path from the search origin toward the trunk and a single path from the search origin toward the main terminal direction can be automatically extracted.
[0030]
Preferably, the ultrasound image is an image representing the three-dimensional space, and is a three-dimensional image in which the attention path is identified and represented. Preferably, a means for giving a larger weighting value to the data of the blood vessel cross-sectional row than other data in order to identify and represent the attention path is included. Thereby, the attention route can be emphasized.
[0031]
(3) The present invention also relates to an ultrasonic image processing apparatus for processing three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body, wherein the ultrasonic image processing device is configured by the three-dimensional data. In the three-dimensional data space, search origin designating means for designating a search origin at a specific site in the blood vessel group, and a search vector is sequentially set from the search origin along a path of interest in the blood vessel group while setting a search vector. It is characterized by including search processing means for sequentially executing extraction, and end determination means for judging the end of the blood vessel section extraction when a predetermined condition is satisfied.
[0032]
Examples of the termination determination condition include that the number of search progresses has reached a set value, that the search route length has reached a set value, that no further search can be performed, and the like.
[0033]
Preferably, an image processing range determining unit that determines a three-dimensional image processing range based on the search origin and the search position at the end determination time point, and performs a three-dimensional labeling process on the blood vessel group within the image processing range. Three-dimensional labeling means.
[0034]
Preferably, the three-dimensional labeling means performs a three-dimensional labeling process forward from the search origin within the image processing range, and a front three-dimensional labeling means for extracting a forward blood vessel branch satisfying a connection relationship from the search origin. And a rear three-dimensional labeling means for performing a three-dimensional labeling process backward from the search origin within the image processing range and extracting a rear blood vessel branch satisfying the connection relationship from the search origin.
[0035]
Preferably, the image forming apparatus further includes an image forming unit that forms a three-dimensional image representing at least one of the anterior vessel branch and the posterior vessel branch.
[0036]
Preferably, the image processing apparatus further includes an image forming unit that forms a three-dimensional image that represents the three-dimensional space and that identifies and expresses at least one of the anterior vessel branch and the posterior vessel branch.
[0037]
(4) The present invention also relates to a method for processing three-dimensional data obtained by transmitting and receiving an ultrasonic wave to and from a three-dimensional space in a living body, wherein the three-dimensional data space constituted by the three-dimensional data is provided. Within, a step of user-designating a search origin at a specific site in a blood vessel group, and sequentially performing a blood vessel cross-section extraction process while sequentially setting a search direction from the search origin to a trunk direction or a terminal direction of the blood vessel group. And thereby generating a blood vessel cross-sectional row composed of a plurality of blood vessel cross-sections.
[0038]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0039]
1 and 2 show examples of images formed by the search processing according to the present invention. Here, FIG. 1 shows a case where search processing is performed in the direction from the end to the trunk in the vascular network, and FIG. 2 shows a case where search processing is performed in the direction from the trunk to the end in the vascular network. .
[0040]
More specifically, FIG. 1A shows a three-dimensional image formed based on three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional region in a living body. ing. In this three-dimensional image, the vascular network 100 as a group of blood vessels constitutes a vascular system branched from the trunk to the end, and a vessel route of interest, that is, a route of interest is represented by reference numeral 102. A search start position 104 is set by the user as a search origin or the like at a specific portion of the attention route 102, and at the same time, a first search direction 106 indicating in which direction the search starts from the search start position 104 is set by the user. Is done.
[0041]
By repeatedly executing a search process described later, a limited extraction image of the attention route can be obtained as shown in FIG. That is, in the limited extraction image of the route, the target route 102 is represented as a three-dimensional image like a one-stroke stroke. Specifically, the route from the search start position 104 to the point that satisfies the predetermined end condition is one point. The book root is shown as a three-dimensional image. In this way, when the search direction is specified in the trunk direction by specifying the terminal portion, even if there is a branch in the middle, only one route is extracted from the specified terminal portion in the direction of the thick trunk and imaged. Becomes possible. Therefore, even when a plurality of branched blood vessels are intertwined with each other and dense, such as in a vascular network in the liver, only a specific blood vessel path is extracted or emphasized and displayed as a three-dimensional image. It is possible to do. Note that such blood vessel extraction can be applied to both arteries and veins. Blood vessels are thought to be essential for tumor tissue to grow. Therefore, the tumor can be killed by identifying and removing all blood vessels extending to the tumor tissue. In an ultrasonic diagnostic apparatus to be described later, all blood vessels extending from a tumor tissue can be displayed, and thus, displaying the blood vessels extending from a tumor can be used for various types of tumor treatment.
[0042]
The ultrasonic diagnostic apparatus of the present embodiment described later has a display mode shown in (C) described below in addition to the display mode shown in (B). That is, (C) shows all the extracted images ahead of the search direction within a certain range set according to the predetermined condition. That is, by performing the three-dimensional labeling process in the search direction 106 with the search start position 104 as a reference, a three-dimensional image as shown in (C) can be formed. In the three-dimensional image, all the blood vessel portions 108 connected to the attention path 102 including the attention path 102 are imaged. In the apparatus according to the present embodiment, the certain range is defined as a search processing result of the attention route 102 shown in FIG. This will be described in detail later.
[0043]
The ultrasonic diagnostic apparatus according to the present embodiment has the display mode shown in (B) and the display mode shown in (C) as described above, and further includes various display variations described later. ing.
[0044]
FIG. 2 shows an example of a case where the search is performed from the trunk to the distal end in the vascular network 100 as described above. (A) shows a three-dimensional image, and a user designates a search start position 104 as a search origin and the like and a first search direction 106 as in the case shown in FIG. Then, the search process is started along the attention path 110, and as a result, as shown in FIG. As shown in (B), the attention path 110 is shown in a one-stroke form as a three-dimensional image. In the present embodiment, as described later, when a blood vessel branches, a thick blood vessel, that is, a main blood vessel is route-selected, and as a result, an image as shown in FIG. As described above, according to such a one-stroke image, it is possible to clearly observe the attention path 110 without being obscured by other blood vessels.
[0045]
(C) shows all the extracted images in the predetermined range in the forward direction of the search direction. As shown in this example, all the connected blood vessel portions 112 including the attention path 110 are extracted, and are imaged. Have been.
[0046]
Note that FIG. 1 shows a search process from the end to the trunk, and FIG. 2 shows a search process from the trunk to the end. In an embodiment described later, these are achieved by the same processing process.
[0047]
Next, the search processing illustrated in FIG. 1 will be described more specifically with reference to FIGS. 3 to 8, and the search processing illustrated in FIG. 2 will be described more specifically with reference to FIGS. 9 to 14.
[0048]
In FIG. 3, a search origin SP0 and coordinates SP1 for designating a search direction are set by a user at a specific site in the vascular network. The upper limit LMAX of the search path length or the degree of progress is determined by the user or as a default setting. Here, the first search direction (search vector) 106 is defined as a direction passing through the coordinate point SP1 from the search origin SP0.
[0049]
When the input by the user as described above is completed, first, as shown in FIG. 4, the first blood vessel section S including the search origin SP0 is obtained. ₁ Is extracted. The specific contents of the blood vessel cross-section extraction process in this case will be described in detail later. When such a blood vessel cross-section is extracted, setting of a reference plane and a search plane and setting of a search vector, which will be described later, are repeatedly performed.
[0050]
As shown in FIG. 5, when the above-described blood vessel cross-section extraction processing is repeatedly performed in the forward direction designated as the search direction, a blood vessel cross-section is extracted at each position, that is, a cross-section row 116 is obtained. The cross-section row 116 is constituted by a plurality of blood vessel cross-sections extracted as described above, and this corresponds to limited extraction data of the attention path. In FIG. 5, the blood vessel cross section first extracted is S ₁ , And the blood vessel section extracted last by satisfying the termination condition is S _m Indicated by That is, in this case, m blood vessel cross sections are extracted along the attention path. In this case, in extracting each blood vessel cross-section, processing is performed so that the connection condition between the blood vessel cross-sections is satisfied. For example, even if there is a branched portion, it is branched like a labeling process. The search process does not proceed to each route.
[0051]
Therefore, the limited extraction image of the attention route shown in FIG. 1 can be constructed as a three-dimensional image by using the cross-sectional row 116 shown in FIG.
[0052]
FIG. 6 shows coordinates SPm representing the blood vessel cross section extracted last. In the present embodiment, a length r connecting the search origin SP0 to the search end coordinate SPm by a linear distance is specified, and is defined as the radius of the image processing range 118. In the present embodiment, the image processing range 118 is defined by a sphere, but of course the present invention is not limited to such a shape. In the present embodiment, a three-dimensional labeling process described below can be executed in such an image processing range 118. Incidentally, although the image processing range 118 is a three-dimensional sphere, the image processing range 118 may be specified as a spherical surface, and the spherical surface may be determined as the limit of the image processing.
[0053]
FIG. 7 illustrates a case where the three-dimensional labeling process is performed from the search origin SP0 in a direction opposite to the search direction. Vessel cross section S which is separated by a unit distance from the reference cross section which is the first blood vessel cross section _-1 All the blood vessel portions 120 that have a connection relationship from the end to the end direction are extracted. That is, reference numeral 120 indicates the rear whole extracted data. By the way, since the echo intensity becomes weaker toward the end, the three-dimensional labeling process ends at that limit.
[0054]
FIG. 8 shows a result in the case where the three-dimensional labeling process is performed forward from the search origin in the search direction. Reference numeral 122 indicates all the extracted data ahead, and in this case, the three-dimensional labeling process is executed only inside the image processing range 118.
[0055]
As described above, according to the method of the present embodiment, when the extraction of the attention route is completed, the image processing range is separately set from the positional relationship between the end point and the search origin at that time, and three-dimensional labeling is performed within the range. By performing the processing, it is possible to form the running states of all the blood vessels centering on the region of interest of the user as separate images. Incidentally, when the three-dimensional labeling result is represented as a three-dimensional image, at least one of the rear full extraction data 120 and the front full extraction data 122 shown in FIGS. 7 and 8 is displayed, and preferably both are combined. Is displayed as something.
[0056]
Next, FIGS. 9 to 14 will be described. FIGS. 9 to 14 basically correspond to the contents shown in FIGS. 3 to 8. When the search is performed in the direction from the trunk to the distal end in the blood vessel network, as shown in FIG. A long LMAX is defined. Then, the search is started in the direction from the search origin SP0 to the coordinate point SP1. That is, as shown in FIG. 10, first, the first blood vessel section S including the search origin ₁ Is extracted, and thereafter, the blood vessel cross-section extraction process is repeatedly executed at each position along the attention path. Then, as shown in FIG. 11, a blood vessel cross-sectional row 116 is obtained.
[0057]
The distance r between the coordinates SPm at the time when the end condition is satisfied and the search origin SP0 is used as a radius defining the image processing range 118, and within the image processing range 118, as shown in FIG. , A three-dimensional labeling process is performed backward from the search origin, and as shown in FIG. 14, a three-dimensional labeling process is performed forward from the search origin. Here, the reference numeral 124 indicates the rear extracted data, and the reference numeral 126 indicates the front extracted data. The extracted data 124 and 126 are desirably integrated, and used as extracted data in the image processing range 118 to construct a three-dimensional image. Therefore, in the case shown in FIGS. 9 to 14, similarly to the case shown in FIGS. 3 to 8, both the three-dimensional image of the attention path and the three-dimensional image of the entire vicinity thereof are formed. Both can be displayed simultaneously or selectively according to the necessity for disease diagnosis or the like.
[0058]
Incidentally, although the three-dimensional labeling process itself may be a known method, in the present embodiment, the search result of the attention route is used as an image processing range, and the three-dimensional labeling process can be performed within the image processing range. It is different from the past.
[0059]
Next, a preferred embodiment of the ultrasonic diagnostic apparatus according to the present embodiment will be described with reference to FIG. This ultrasonic diagnostic apparatus executes the above-described three-dimensional data processing, and specifically includes an image processing unit 17 corresponding to an ultrasonic image processing apparatus. Of course, the image processing unit 17 can be constituted by a computer or the like provided separately from the ultrasonic diagnostic apparatus.
[0060]
The 3D probe 10 is an ultrasonic probe for capturing three-dimensional echo data that is used in contact with a body surface or inserted into a body cavity. The 3D probe 10 has a 2D array vibrator (not shown), and the 2D array vibrator has a large number of vibrating elements arranged in a matrix. The ultrasonic beam can be two-dimensionally electronically scanned by the 2D array transducer. As a result, a three-dimensional echo data acquisition space (three-dimensional space) V is formed. For example, when forming a three-dimensional image of the blood vessel network in the liver, the 3D probe 10 is brought into contact with the surface of the abdominal body, and the three-dimensional space V includes the target blood vessel or the blood vessel network. The posture and contact position of the probe 10 are adjusted.
[0061]
The transmission / reception unit 12 functions as a transmission beam former and a reception beam former. That is, the transmission / reception unit 12 supplies transmission signals to the plurality of vibration elements, and performs phasing addition processing on the plurality of reception signals output from the plurality of vibration elements.
[0062]
The signal processing unit 14 performs predetermined signal processing such as detection and logarithmic compression on the reception signal after the phasing addition output from the transmission / reception unit 12. In the present embodiment, the echo data is a target of image processing. However, it is a matter of course that Doppler information may be extracted for each voxel and the Doppler information may be subjected to three-dimensional image processing.
[0063]
The 3D memory 16 stores echo data captured in a three-dimensional space. In this case, necessary coordinate conversion, interpolation processing, and the like are performed. Therefore, three-dimensional data is stored in the 3D memory 16, and the three-dimensional data forms a three-dimensional data space. The stored three-dimensional data is output to the display processing unit 18 for normal three-dimensional image processing. In the present embodiment, image processing is performed to apply the above-described image processing unique to the present embodiment. It is output to the unit 17.
[0064]
The display processing unit 18 applies a known rendering process to the three-dimensional data, thereby constructing a three-dimensional ultrasound image. In this case, various rendering methods can be used, and for example, a volume rendering method may be used. The display unit 20 displays an image processed by the display processing unit 18.
[0065]
The control unit 24 controls the operation of each component included in the apparatus. The input unit 22 including an operation panel is connected to the control unit 24. The control unit 24 also controls the operation of each component included in the image processing unit 17. Is output.
[0066]
Next, a specific configuration of the image processing unit 17 will be described in detail. The binarization processing unit 26 is a circuit that performs a binarization process on the three-dimensional data stored in the 3D memory 16. That is, a binarization process is performed to distinguish the blood flow data from the real tissue data. Since blood flow data generally has a smaller value than real tissue data, for example, 1 is assigned to such small data, and 0 is assigned to other data. It is possible to construct a data space to which 1 is given only for the blood flow portion in the original data space. The data after the binarization processing is output to the binarization data processing unit 28.
[0067]
The binarized data processing unit 28 includes a search processing unit 30 and a 3D labeling unit 32 in the present embodiment. Of course, in addition to such functions, various functions are provided, which will be described later with reference to FIG. The search processing unit 30 is a processing unit that extracts a blood vessel cross section at each position along the route of interest. The 3D labeling unit 32 is a processing unit that performs a three-dimensional labeling process within an image processing range.
[0068]
To this binarized data processing unit 28, three 3D memories 34, 36, 38 are connected. Of course, the 3D memories 34, 36, 38 may be constituted by a single storage device. The 3D memory 34 stores data corresponding to the blood vessel cross-sectional row extracted by the search processing, that is, limited extraction data of the attention path. The 3D memory 36 stores all extracted data in the forward direction, which is labeling data obtained by performing three-dimensional labeling processing forward from the search origin. The 3D memory 38 stores data obtained by performing the labeling process in the backward direction from the search origin, that is, all the extracted data in the backward direction.
[0069]
The selection / synthesis processing unit 40 appropriately selects or combines the extracted data stored in each of the 3D memories 34, 36, and 38, and outputs the processing result to the display processing unit 18. For example, when forming a limited extraction image of the attention path, only the extraction data output from the 3D memory 34 is output to the display processing unit 18. When the result of the three-dimensional labeling process is used as a three-dimensional image, the extracted data output from each of the 3D memories 36 and 38 is combined and output to the display processing unit 18. In this case, either one of the forwardly extracted data and the backwardly extracted data may be selected.
[0070]
FIG. 16 illustrates an example of a main configuration of the display processing unit 18 illustrated in FIG. Data sequentially read from the 3D memory 16 is input to one input terminal of the multiplier 44. On the other hand, the coefficient w1 or w2 output from the coefficient selector 42 is input to the other input terminal of the multiplier 44.
[0071]
Output data from the selection / synthesis processing unit 40 shown in FIG. 15, that is, extracted data, is input to the coefficient selector 42. Describing the extracted data, values extracted by the above-described blood vessel cross-section extraction processing or three-dimensional labeling processing have a value of 1 and other values have a value of 0. The coefficient selector 42 outputs the coefficient w1 when the value 1 is input as the extraction data, and outputs the coefficient w2 when the value 0 is input as the extraction data. Here, w1 is, for example, 1.0 and w2 is, for example, 0.5. The coefficients w1 and w2 function as weighting values. That is, by giving a large weight to the data extracted by the above-described extraction processing, it becomes possible to express the extracted portion in the three-dimensional image with high brightness. Of course, the portion may be colored, for example, to make it stand out from other real tissues or blood vessels.
[0072]
Therefore, in the multiplier 44, data sequentially output from the 3D memory 16 is weighted, and the weighted data is input to the rendering unit 46. The rendering unit 46 performs three-dimensional image processing based on, for example, the volume rendering method as described above. Various methods can be used to display the extracted blood vessel. The extracted blood vessel part is displayed in a specific color, and the other blood vessel parts are displayed in another color. The real organization may be represented as a black and white image. Alternatively, by providing the output of the binarization processing unit 26 to the multiplier 44, when displaying only a blood vessel as a three-dimensional image, a specific blood vessel part may be expressed with high brightness.
[0073]
By the way, in the multiplier 44, the control unit 24 controls the reading of the 3D memories 16, 34, 36, and 38 in order to correctly assign a corresponding coefficient to certain data.
[0074]
Next, a specific operation example of the device described above with reference to FIG. 17 will be described with reference to FIGS. FIG. 17 is a flowchart showing an operation example of the apparatus.
[0075]
In S101, the coordinate origin SP0 and the coordinate point SP1 are specified by the user, and the parameter LMAX as the maximum number of steps is specified by the user. This maximum number of steps is one of the search termination conditions. Of course, the search end condition can be specified not as the maximum number of steps but as a search path length or the like.
[0076]
When two coordinates are specified in S101, for example, a method as shown in FIGS. 18 and 19 can be used. That is, FIG. 18 shows a three-dimensional image 130 displayed on the display unit. The three-dimensional image 130 is obtained by imaging a three-dimensional region in a living body. A blood vessel network 132 exists in the three-dimensional area, and when it is desired to perform limited extraction on a specific path, that is, a path of interest 134, a cursor or the like is used in the three-dimensional space as shown in this example. An arbitrary cutting plane 136 is set by the user. That is, the arbitrary cut plane 136 is a plane that crosses a part of the attention path 134 that specifies the search origin and the like.
[0077]
FIG. 19 shows an arbitrary tomographic image 138 corresponding to the arbitrary cut plane 136. Here, reference numerals 140, 142, and 144 are cross sections of the blood vessel that appear when an arbitrary cut surface crosses the blood vessel network. Here, the cross section of the blood vessel corresponding to the path of interest is indicated by reference numeral 142, and the user sets the search origin SP0 and the coordinate point SP1 on the blood vessel cross section 142 on such an arbitrary tomographic image 138. , The initial conditions of the search can be specified specifically and easily. Of course, the initial conditions may be set using, for example, a triplane or the three-dimensional image itself without using such an arbitrary tomographic image. However, using such an arbitrary tomographic image has the advantage that the attention route can be specified more easily and accurately.
[0078]
Returning to FIG. 17, in S102, the first search direction U is determined from SP0 and SP1 set as described above. ₁ Is calculated. That is, the search direction U as the first search vector as a direction from SP0 to SP1 ₁ Is required. Also, the first search direction U from the search origin ₁ The coordinates SPA existing in the opposite direction are calculated.
[0079]
This will be described with reference to FIG. As described above, the first search vector U is placed on the line 146 connecting the search origin SP0 and the coordinate point SP1. ₁ Is calculated. Further, the coordinate point SPA is automatically specified as a point on the line 146 that is away from the search origin SP0 by a certain distance in the reverse direction. Here, the fixed distance may be an arbitrary distance, but the unit distance is set to 1 here. By the way, the reference plane Q ₁ And search plane R ₁ Is also 1 as a unit distance. This 1 corresponds to, for example, the size of one voxel data. Note that reference numeral 134 represents a path of interest, that is, a blood vessel of interest. In FIG. 20, 1 is given as a subscript of each mechanism. However, in the general description below, the subscript uses i, and i corresponds to the number of search processes. . Here, the time when i = 1 corresponds to the first search processing.
[0080]
Returning to FIG. 17, in S103, 0 is substituted for the parameter L and the parameter P representing the reference point is set. _i Is substituted for the coordinates of SP0, and 1 is substituted for the parameter i.
[0081]
In S104, the reference plane Q _i Is set. Specifically, as shown in FIG. _i And the search direction, ie, the search vector U _i Is a reference plane Q _i Is defined as This reference plane Q _i The above data is read from the three-dimensional data space, and is stored in a storage area existing in the binarized data processing unit 28.
[0082]
In FIG. 17, in S105, the search plane R _i Is set. That is, as shown in FIG. _i From the search direction U _i A plane separated by a distance 1 in the direction of _i Search plane R as a plane parallel to _i Is set. This search plane R _i Is to roughly grasp the connection relationship of the blood vessel cross sections as preprocessing for actually extracting the blood vessel cross sections, and as a result, appropriately set the next search vector. In this S105, the search plane R is searched from within the three-dimensional data space. _i The above data is read and stored in a storage area existing in the binarized data processing unit 28.
[0083]
Note that the above-described reference plane Q _i And search plane R _i May be defined as a plane over the entire area of the three-dimensional data space, or the search vector U _i It may be defined as a certain finite area centered on the like. According to such a configuration, there is an advantage that the calculation time can be reduced.
[0084]
In S106 in FIG. 17, the reference plane Q extracted as described above _i The above data and search plane R _i A labeling process is performed on each of the above data, and the resulting labeling image (label image) is stored. Here, the labeling process is a two-dimensional labeling process, which is different from the three-dimensional labeling process performed by the 3D labeling unit 32 illustrated in FIG. That is, it corresponds to the labeling processing on each plane, in other words, the blood part extraction processing. Incidentally, FIG. 23 shows the search plane R _i Are shown, and a label as an identification symbol is given to each of the areas A1 and A2. Note that reference numeral 150 represents the reference cross section S. _i Are projected (described later).
[0085]
FIG. 22 shows the result of labeling processing performed on data on each plane. As is well known, in the labeling process, a label, that is, a number for identifying each isolated region, that is, each closed region is assigned. Here, the reference plane Q _i Above, only one region is extracted, while the search plane R _i In the above, two regions A1 and A2 are extracted.
[0086]
In FIG. 17, in S107, the reference plane Q _i From the labeling image of FIG. 5, the region including the reference point Pi, that is, the region connected to the reference point P _i Is extracted as This is shown in FIG. This reference section S _i Is the reference section Q _i It is a blood vessel section of the attention path extracted above. The blood vessel cross section thus extracted, that is, the reference cross section S _i Data is saved. Then, in S108 in FIG. 17, the reference section S _i Is calculated for the center of gravity Gi. It is shown in FIG.
[0087]
Then, at S109 in FIG. _i Is output to the 3D memory 34 shown in FIG. 15 as the searched blood vessel section data.
[0088]
In S110 in FIG. 17, the search plane R _i Of the reference section S in the labeling image _i The largest overlapping part with is extracted. This will be described with reference to FIG. As described above, the search plane R _i At the top of this example, there are two extraction areas A1 and A2, and it is not clear which of them corresponds to the target route to be searched in the future. Therefore, in the present embodiment, the search plane R _i Up reference plane Q _i Upper reference section S _i Is projected, and the connection relationship of the blood vessel cross-section, that is, the running direction of the blood vessel is grasped by recognizing the overlapping portion generated by such projection.
[0089]
Reference numeral 150 in FIG. _i Is shown, where the area 150 extends over both of the two areas A1 and A2, and the larger or larger overlapping portion of the respective areas A1 and A2 is denoted by reference numeral 152. The overlap shown. Here, the overlapping portion 152 is hatched in FIG.
[0090]
In FIG. 17, it is determined in S111 whether such an overlapping portion exists, and if so, it is determined in S112 whether the parameter L has reached MAX, that is, the termination condition is determined, If it is determined that L has not reached LMAX, the steps after S113 are executed. That is, in S113, the center of gravity of the overlapping portion 152 shown in FIG. _i Is calculated as Next, in S114, the reference section S _i Center of gravity G _i From the temporary center of gravity K _i Search direction U as the direction connecting _{i + 1} Is determined.
[0091]
The step S114 will be described in detail with reference to FIG. 21. As described above, the reference plane Q _i To the search plane R _i Is set, and the relationship between the two is determined by the search vector U. _i Defined by On the other hand, the search plane Q _i Above is the reference section S _i Is extracted, its center of gravity G _i Is calculated. Also, the search plane R _i In the above, the overlapping portion is specified by the above-described process, and the center of gravity K _i Is specified. And the center of gravity G _i From the temporary center of gravity K _i Search vector U as the next search direction as the direction passing through _{i + 1} Is determined. Search vector U in that case _{i + 1} Is the temporary center of gravity Ki, which is the reference point P in the next cross-section extraction process. _{i + 1} Is the same as Then, in the next blood vessel section extraction processing, the search vector U _{i + 1} New reference plane Q _{i + 1} Is set, and the reference plane Q _{i + 1} New reference section S _{i + 1} Is extracted.
[0092]
Accordingly, as can be understood from such a series of flows, the search direction is sequentially reset as much as possible along the direction of the central axis according to the traveling state of the attention route 134, and at the same time, The direction of the plane is also sequentially reset so as to be as orthogonal to the center line of the attention path 134 as possible.
[0093]
In performing a series of processes as described above, in S115 shown in FIG. 17, the parameter L is incremented by one, and the parameter P representing the reference point is increased. _i Temporary center of gravity K _i And i is incremented by one. Then, each process after S104 is repeatedly executed.
[0094]
FIG. 24 conceptually shows the result of the blood vessel cross-section extraction processing in a plurality of stages. Here, the direction from the blood vessel portion 140 to 141 is the path of interest, but the blood vessel portion 140 is connected to 141. The other blood vessel portion 142 is also connected to the portion where the blood vessel is located. In FIG. 24, S ₁ ~ S ₅ Represents a blood vessel cross section at each position extracted by the respective blood vessel cross section processing, and R ₁ ~ R ₅ Represents a search plane set in each blood vessel cross-section extraction process. G ₁ ~ G ₅ Represents the center of gravity calculated for the reference section in each blood vessel section extraction process. ₁ ~ K ₅ Indicates the temporary center of gravity calculated on the search plane in each blood vessel cross-section extraction process. As shown in FIG. 24, even if the blood vessel cross section suddenly increases at the junction, (S ₄ Then, the direction of section extraction is optimized along the running direction of the blood vessel, and the processing can proceed without losing sight of the search direction even at such a merging or branching point. As shown in FIG. 22, the identification of the overlapping portion 152 and the provisional center of gravity K _i According to the above calculation, it is possible to extract a blood vessel cross-section along the path of interest even when the confluence portion and the branch portion as described above exist, especially when searching from the trunk to the terminal portion. Even if there is a branch, there is an advantage that the search can be automatically advanced in the direction of the thickest blood vessel portion by specifying the largest overlapping portion on the search plane.
[0095]
Returning to FIG. 17, if it is determined in S111 that there is no overlapping portion, or if it is determined in S112 that the parameter L has reached LMAX, the steps after S116 are executed. In this S116, as shown in FIGS. 6 and 12, the center of gravity Gi of the reference section at the end of the search is specified as the coordinate E at the end of the search. In S117, as shown in FIGS. , The image processing range 118 is determined. In this case, the image processing range 118 may be defined as a sphere region or a sphere surface region.
[0096]
In S118, forward or backward three-dimensional labeling processing as shown in FIGS. 7 and 8 and FIGS. 13 and 14 is performed, and the respective extracted data are stored in the 3D memories 36 and 38 shown in FIG. Is stored.
[0097]
In the above embodiment, in determining the search vector, for example, the center or the center of the region may be used without using the center of gravity. In calculating the temporary center of gravity, the center of gravity of the area A1 including the overlapping portion 152 may be used as the temporary center of gravity instead of obtaining the center of gravity of the overlapping portion 152 shown in FIG.
[0098]
In the above-described embodiment, the three-dimensional labeling process can be executed after the extraction of the cross section of the attention path is completed. However, as a modified example, the image processing range is set independently by a user or the like, and the range is set. The three-dimensional labeling process can also be performed within. However, according to the above-described embodiment, when the two images are compared and observed, since the attention path to be extracted and the entire blood vessel extracted by the three-dimensional labeling process belong to the same region. There is an advantage that the correspondence can be recognized more accurately.
[0099]
【The invention's effect】
As described above, according to the present invention, there is an advantage that blood vessels having various forms can be appropriately extracted. Further, according to the present invention, an image useful for blood vessel diagnosis can be provided.
[Brief description of the drawings]
FIG. 1 is a diagram showing a limited extracted image of a route of interest and an all-extracted image ahead of a search direction within a certain range when a search start position is specified for an end in a blood vessel network.
FIG. 2 is a diagram illustrating a limited extracted image of a target route and an all-extracted image ahead of a search direction within a certain range when a search start position is designated for a trunk in a vascular network.
FIG. 3 is a diagram for describing designation of a coordinate point when a search is performed from a terminal side;
FIG. 4 is a diagram showing a first blood vessel cross section when a search is performed from the distal end side.
FIG. 5 is a diagram showing a cross-sectional row obtained by performing a search from the terminal side.
FIG. 6 is a diagram illustrating an image processing range.
FIG. 7 is a diagram showing all extracted data at the rear obtained by a three-dimensional labeling process.
FIG. 8 is a diagram showing all front extracted data obtained by a three-dimensional labeling process.
FIG. 9 is a diagram showing designation of coordinates when a search is performed from the trunk side.
FIG. 10 is a diagram showing a first blood vessel cross section when a search is performed from the trunk side.
FIG. 11 is a diagram showing a cross-sectional row obtained by performing a search from the trunk side.
FIG. 12 is a diagram illustrating an image processing range.
FIG. 13 is a diagram illustrating all rear extraction data obtained by a three-dimensional labeling process.
FIG. 14 is a diagram showing all front extracted data obtained by a three-dimensional labeling process.
FIG. 15 is a block diagram showing a preferred embodiment of the ultrasonic diagnostic apparatus according to the present embodiment.
16 is a block diagram illustrating a specific configuration example of a display processing unit illustrated in FIG.
17 is a flowchart showing an operation example of the ultrasonic diagnostic apparatus shown in FIG.
FIG. 18 is a diagram showing a three-dimensional image representing a three-dimensional space.
FIG. 19 is a diagram showing an arbitrary tomographic image corresponding to an arbitrary cutting plane set for a three-dimensional space.
FIG. 20 is a diagram for describing search origin and designation of a search direction.
FIG. 21 is a diagram for explaining a positional relationship between a reference plane and a search plane.
FIG. 22 is a diagram showing a state in which a reference section on the reference plane is projected on a search plane.
FIG. 23 is a diagram illustrating a result of a two-dimensional labeling process performed on a search section.
FIG. 24 is a diagram showing a result of each blood vessel cross-section extraction process executed along the attention path.
[Explanation of symbols]
10 3D probe, 12 transmitting / receiving unit, 14 signal processing unit, 16 3D memory, 17 image processing unit, 18 display processing unit, 20 display unit, 24 control unit, 26 binarization processing unit, 28 binarization data processing unit, 30 search processing unit, 32 3D labeling unit, 34, 36, 38 3D memory, 40 selection / synthesis processing unit, 100 vascular network, 104 search start position, 106 search direction, 118 image processing range.

Claims (19)

In an ultrasonic image processing device that processes three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body,
In a three-dimensional data space configured by the three-dimensional data, search origin specifying means for specifying a search origin for a specific site in the blood vessel group,
Search processing means for sequentially executing the extraction of a blood vessel cross-section while sequentially setting a search vector from the search origin along the path of interest in the blood vessel group,
Including
The ultrasound image processing apparatus, wherein the search processing means generates a blood vessel cross-section row including a plurality of blood vessel cross sections sequentially extracted along the attention path. The device of claim 1,
The ultrasonic image processing apparatus, wherein the search processing unit sequentially sets the search vector according to a traveling mode of the attention route. The device according to claim 2,
Ultrasound image processing, wherein the search processing means sequentially sets a reference plane for extracting a blood vessel cross section in the three-dimensional data space based on the sequentially set search vector. apparatus. The device of claim 1,
The search processing means,
reference plane setting means for setting an i-th reference plane orthogonal to the i-th search vector;
A blood vessel section extracting means for extracting an i-th blood vessel cross section on the i-th reference plane; and an i-th search plane at a position parallel to the i-th reference plane and a unit distance away in the i-th search direction. Search plane setting means for setting
A blood vessel portion extracting means for extracting an i-th blood vessel portion having an overlapping relationship with the i-th blood vessel cross section on the i-th search plane;
Search vector determining means for determining an (i + 1) th search vector based on a correlation between the i-th blood vessel cross section and the i-th blood vessel portion;
An ultrasonic image processing apparatus comprising: The device according to claim 4,
The blood vessel portion extraction means,
Two-dimensional labeling means for performing a labeling process on the i-th search plane to obtain an i-th labeling image including one or more closed regions;
Projecting the i-th blood vessel cross-section onto the i-th labeling image to identify a closed region having the largest overlap area with respect to the i-th blood vessel cross-section; Projection processing means for extracting a portion overlapping with the i-th blood vessel cross section as the i-th blood vessel portion;
An ultrasonic image processing apparatus comprising: The device according to claim 4,
The search vector determining means includes:
Means for calculating the center of gravity of the i-th blood vessel section;
Means for calculating the center of gravity of the i-th blood vessel portion;
Means for determining the (i + 1) -th search vector, with a direction passing from the center of gravity of the i-th blood vessel cross-section to the center of gravity of the i-th blood vessel portion as the vector direction
An ultrasonic image processing apparatus comprising: The device according to claim 4,
An ultrasonic image processing apparatus, wherein the (i + 1) th reference plane corresponds to a plane obtained by rotating the i-th search plane about a predetermined point on the plane as a rotation center. The device of claim 1,
A search start direction specifying means for specifying a search start direction;
An ultrasonic image processing apparatus, wherein an initial search vector is determined based on the search origin and the search start direction. The device according to claim 8,
Including reference image forming means for forming a two-dimensional or three-dimensional reference image based on the three-dimensional data,
The ultrasound image processing apparatus, wherein the search origin and the search start direction are designated by a user using the reference image. In an ultrasonic image processing device that processes three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body,
In a three-dimensional data space configured by the three-dimensional data, search origin specifying means for specifying a search origin for a specific site in the blood vessel group,
Search processing means for sequentially executing the extraction of a blood vessel cross-section while sequentially setting a search vector from the search origin along the path of interest in the blood vessel group,
A storage unit that stores a blood vessel cross-sectional row including a plurality of blood vessel cross-sections sequentially extracted along the noted route,
Image forming means for forming an ultrasonic image including a path of interest based on the blood vessel cross-sectional row,
An ultrasonic image processing apparatus comprising: The device according to claim 10,
The ultrasonic image processing apparatus according to claim 1, wherein the ultrasonic image is a three-dimensional image expressing the attention path as a single stroke. The device according to claim 10,
The ultrasound image processing apparatus, wherein the ultrasound image is an image representing the three-dimensional space, and is a three-dimensional image in which the attention path is identified and represented. The apparatus according to claim 11,
An ultrasonic image processing apparatus, comprising: means for giving a larger weighting value to the data of the blood vessel cross-sectional row than other data in order to identify and express the path of interest. In an ultrasonic image processing device that processes three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body,
In a three-dimensional data space configured by the three-dimensional data, search origin specifying means for specifying a search origin for a specific site in the blood vessel group,
Search processing means for sequentially executing the extraction of the blood vessel cross-section while sequentially resetting the search vector from the search origin along the path of interest in the blood vessel group,
End determination means for determining the end of the blood vessel cross-section extraction when a predetermined condition is satisfied,
An ultrasonic image processing apparatus comprising: The apparatus according to claim 14,
Image processing range determining means for determining a three-dimensional image processing range based on the search origin and the search position at the end determination time,
Three-dimensional labeling means for performing three-dimensional labeling processing on the blood vessel group within the image processing range,
An ultrasonic image processing apparatus comprising: The apparatus of claim 15,
The three-dimensional labeling means,
Forward three-dimensional labeling means for performing three-dimensional labeling processing forward from the search origin within the image processing range, and extracting forward blood vessel branches satisfying the connection relationship from the search origin,
A rear three-dimensional labeling unit that performs three-dimensional labeling processing backward from the search origin within the image processing range, and extracts a rear blood vessel branch satisfying a connection relationship from the search origin,
An ultrasonic image processing apparatus comprising: The apparatus of claim 16,
An ultrasonic image processing apparatus comprising: an image forming unit that forms a three-dimensional image representing at least one of the anterior vessel branch and the posterior vessel branch. The apparatus of claim 16,
An ultrasonic image processing apparatus, comprising: an image forming unit that forms an image representing the three-dimensional space and that identifies and represents at least one of the anterior vessel branch and the posterior vessel branch. In a method of processing three-dimensional data obtained by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body,
In the three-dimensional data space configured by the three-dimensional data, a step of allowing the user to specify a search origin at a specific site in the blood vessel group,
A step of sequentially performing a blood vessel cross-section extraction process while sequentially setting a search direction from the search origin to a trunk direction or a distal direction of the blood vessel group, thereby generating a blood vessel cross-sectional row including a plurality of blood vessel cross-sections. When,
A three-dimensional data processing method comprising:

Images