JP4653599B2 - Ultrasonic diagnostic equipment - Google Patents

Description

  The present invention relates to an ultrasonic diagnostic apparatus, and more particularly to a technique for removing noise contained in a target region.

  By using the ultrasonic diagnostic apparatus, various diagnostic information relating to a target region such as the inside of the heart (heart chamber) can be acquired. A typical example of diagnostic information is the volume (volume) of the heart chamber. When measuring the volume of the heart chamber with the ultrasonic diagnostic apparatus, it is necessary to transmit and receive ultrasonic waves in a space including the heart, and to separate the heart chamber and the heart chamber from the echo data obtained thereby. Since there is a relatively large difference in echo level between the myocardium and the heart chamber, for example, only the heart chamber portion can be extracted by separation processing based on binarization processing or the like.

  However, a noise region that is a relatively large lump that cannot be removed by the normal noise removal process may remain in the heart chamber extracted by the separation process even after the noise removal process or the like.

  Against this background, Patent Document 1 proposes an epoch-making technique for removing an isolated region (noise region) remaining in a heart chamber or the like by using a labeling process or the like.

JP 2004-267484 A

  In the above-mentioned Patent Document 1, an isolated region existing in the heart chamber is extracted using a labeling process, and a technique for determining whether the region is a noise region according to the volume value of the extracted isolated region, A method for determining whether or not the user is in a noise area from the image of the area is described.

  The present invention has been made in the process of improving the technique described in the above-mentioned patent document, and its object is to remove a noise area existing in the target area accurately and with a relatively easy process. It is to provide.

  In order to achieve the above object, an ultrasonic diagnostic apparatus according to a preferred embodiment of the present invention transmits and receives an ultrasonic wave in a diagnostic region including a target region to acquire an echo signal, and a diagnosis formed from the echo signal In an ultrasonic diagnostic apparatus that performs data processing on data in a region, target region extraction means for extracting a target region from a diagnostic region based on an echo level, and a plurality of isolated regions other than the target region in the diagnostic region A labeling processing unit that performs a labeling process to separate areas and assigns a labeling number to each isolated area, and based on the labeling number, noise is detected from the isolated areas that are surrounded by the target area. Noise area extracting means for extracting as an area, noise area removing means for removing the noise area by taking the extracted noise area into the target area, and Characterized in that it has a.

  In the above configuration, the target region is, for example, a heart chamber in the heart. The diagnosis area may be a three-dimensional area or a two-dimensional area. In the above configuration, the isolated region is accurately extracted and separated by the labeling process. Since the noise region is extracted based on the labeling number, for example, it is possible to omit the trouble of calculating the volume of the noise region or the like, or the operation of the user confirming the image and specifying the noise region.

  In a further preferred aspect of the ultrasonic diagnostic apparatus, the ultrasonic diagnostic apparatus further comprises background processing means for setting a background edge at an outer edge of the diagnostic area, and the labeling processing means surrounds the target area and is connected to the background edge. An isolated region is used as a background region, and the noise region extracting unit extracts an isolated region other than the background region from the plurality of isolated regions as the noise region.

  According to the above configuration, even in a situation where the background portion surrounding the target region is divided into a plurality of isolated regions, the background portion is divided by applying the labeling process after providing the background edge on the outer edge of the diagnostic region. A plurality of isolated regions are connected by a background edge and extracted as a single isolated region. Since the background area is extracted as a single isolated area, the background area can be easily specified, and the isolated area other than the background area can be easily specified.

  In order to achieve the above object, an ultrasonic diagnostic apparatus according to a preferred aspect of the present invention includes a plurality of voxels that constitutes the diagnostic region by transmitting and receiving ultrasonic waves in the diagnostic region including the target region. A transmitting / receiving means for acquiring data; a target area extracting means for extracting a target area voxel corresponding to the target area based on voxel data from the plurality of voxels; and a plurality of voxels other than the target area voxel. Based on the labeling number, a labeling processing means for separating the isolated voxel group which is a group of a plurality of spatially connected voxels into a plurality of isolated voxel groups in units and assigning a labeling number to each of the plurality of isolated voxel groups Thus, an isolated voxel group surrounded by the target area voxel is extracted as a noise voxel group from a plurality of isolated voxel groups Noise voxel group extraction means, and noise voxel group removal means for removing the noise voxel group by taking the voxels belonging to the extracted noise voxel group into the target area voxel, and constitutes the outer edge of the diagnostic area By setting a plurality of voxels as background voxels, the labeling processing means surrounds the target area voxel, and a group of isolated voxels connected by the background voxel is used as a background area, and the noise voxel group extraction means includes: An isolated voxel group other than a background region is extracted as the noise voxel group from the plurality of isolated voxel groups.

  In the ultrasonic diagnostic apparatus of a more desirable aspect, the labeling processing unit assigns a labeling number to each of a plurality of isolated voxel groups in order from the isolated voxel group of the background region, and the noise voxel group extracting unit includes a plurality of noise voxel group extracting units. It is characterized in that an isolated voxel group in the background region and other isolated voxel groups are discriminated from a comparison between a labeling number assigned to each of the isolated voxel groups and a threshold of the labeling number. Further, in the ultrasonic diagnostic apparatus of a more desirable mode, the diagnostic region is a region in a three-dimensional space, and volume calculation means for calculating the volume of the target region from the number of target region voxels into which the voxels belonging to the noise voxel group have been captured It further has these.

  According to the present invention, it is possible to accurately and relatively easily remove a noise region existing in a target region. For example, it is possible to omit the trouble of calculating the volume of the noise region and the like, and the operation of the user confirming the image and specifying the noise region.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described with reference to the drawings.

  FIG. 1 shows a preferred embodiment of an ultrasonic diagnostic apparatus according to the present invention, and FIG. 1 is a block diagram showing the overall configuration thereof.

  The 3D probe 10 is an ultrasonic probe for acquiring three-dimensional echo data, and the 3D probe 10 is used in contact with the patient's body surface or inserted into a body cavity. The 3D probe 10 scans ultrasonic waves in a three-dimensional space by mechanically scanning a 1D array transducer that forms a two-dimensional scanning surface by electronic scanning. The 3D probe 10 may be one that scans ultrasonic waves in a three-dimensional space by electronically scanning a 2D array transducer in which transducers are two-dimensionally arranged.

  The transmission / reception unit 12 controls the 3D probe 10 to transmit and receive ultrasonic waves in a three-dimensional space including the heart that is the target tissue. That is, the transmission / reception unit 12 functions as a transmission beamformer and a reception beamformer, acquires voxel values (echo values) for each voxel of a plurality of voxels constituting a three-dimensional space, and sends them to the three-dimensional data memory (1) 14. Remember.

  Although the diagnosis target of the ultrasonic diagnostic apparatus according to the present invention is not limited to the heart, the heart will be described below as a target tissue. In general, in the ultrasound diagnosis of the heart, an ultrasound image is obtained in which the heart cavity portion is greatly projected and the heart muscle portion surrounds the heart cavity portion. That is, most of the formed ultrasound image is occupied by the myocardial part and the heart cavity part. Therefore, in the following description, it is assumed that ultrasonic waves are transmitted and received within a three-dimensional diagnostic region composed of a myocardial part and a cardiac cavity part.

  In the three-dimensional data memory (1) 14, each voxel value is recorded at an address corresponding to the coordinate value in the three-dimensional space. The coordinates in the three-dimensional space may be coordinate values of the rθφ polar coordinate system suitable for the sector scanning method of the ultrasonic beam, or may be coordinate values of an xyz orthogonal coordinate system suitable for the rectangular parallelepiped shape.

  The target region extraction unit 16 extracts a heart cavity portion that is a target region from the diagnosis region. In general, the heart cavity has a lower echo level than the myocardium. For this reason, the target region extraction unit 16 can extract the heart chamber by binarization processing. That is, by setting the threshold value to a level smaller than the level corresponding to the myocardial part and larger than the level corresponding to the cardiac cavity part, the myocardial part and the cardiac cavity part can be roughly classified based on the threshold value. .

  The target region extraction unit 16 is not limited to a simple binarization process. For example, the target region extraction unit 16 uses a technique described in Patent Document 1 (Japanese Patent Application Laid-Open No. 2004-267484) described above, and a myocardial part and a heart chamber. Parts may be separated. That is, for example, a window composed of 9 3 × 3 voxels is provided, and each voxel constituting the window is binarized to be divided into voxels greater than or equal to the threshold value and voxels less than the threshold value. Based on the voxel classification result, it may be determined whether the target voxel located at the center of the window is a voxel corresponding to the myocardium or a voxel corresponding to the heart chamber.

  The binarization processing unit 18 assigns the voxel value “1” to the voxels of the heart chamber and assigns the voxel value “0” to the other voxels based on the extraction result of the heart chamber in the target region extraction unit 16. Then, binarized data (binarized image) in which the heart chamber and other portions are separated is formed. If the target region extraction unit 16 has already performed the process of assigning the voxel value “1” to the voxel of the heart chamber and assigning the voxel value “0” to the other voxels, the binarization processing unit 18. May be omitted. That is, the binarization processing unit 18 may be incorporated in the target area extraction unit 16. The binarized data is stored in the three-dimensional data memory (2) 20.

  The noise region removal block 30 removes the noise region from the inside of the heart chamber that is the target region based on the binarized data. Therefore, the principle of noise region removal in this embodiment will be described with reference to FIGS.

  FIG. 2 is a diagram for explaining the principle of removing a noise region in the present embodiment. In FIG. 2, two-dimensional image data will be described as an example.

  FIG. 2A shows a binarized image from which the heart chamber is extracted. That is, a binarized image in which a voxel value “1” is assigned to a voxel in the heart chamber in the diagnosis region 100 and a voxel value “0” is assigned to other voxels is shown. Note that the portion of the voxel value “0” is represented by a black dot. Incidentally, an image image corresponding to the binarized image of FIG. 2A is FIG. 2A ′, and the heart chamber 110 and other parts are displayed as images of the diagnostic region 100. The heart chamber 110 is surrounded by a background region 130 mainly composed of myocardium. Further, a noise region 120 exists in the heart chamber 110. In the present embodiment, the noise region 120 is removed by processing described below.

  FIG. 2B is an inverted image obtained by inverting the binarized image of FIG. That is, the image is obtained by inverting the voxel value “0” in FIG. 2A to “1” and the voxel value “1” in FIG. 2A to “0”. Incidentally, the image corresponding to the binarized image of FIG. 2B is FIG. 2B ′, and the heart chamber 110 and the noise region are displayed as the images of the diagnostic region 100 as in FIG. 2A ′. 120 and a background area 130 are displayed.

  In the present embodiment, a labeling process is performed on the reverse image in FIG. That is, the voxel value “1” in FIG. 2B is separated into a plurality of isolated voxel groups in units of isolated voxel groups that are a group of a plurality of spatially connected voxels. Give each a labeling number. For example, scanning is started from the upper left of the reverse image in FIG. 2B, and numbers are assigned in ascending order from 1 to a plurality of voxel clusters having the same voxel value that are spatially connected.

  FIG. 2C shows a state after the labeling process. As a result of starting scanning from the upper left of the reverse image, the label “1” is assigned to the isolated voxel group that is the background region surrounding the heart chamber, and the portion surrounded by the heart chamber inside the heart chamber Labels “2” and “3” are assigned to the isolated voxel group.

  In this embodiment, based on the assigned labeling number, an isolated voxel group that is surrounded by the heart chamber is extracted as a noise region (noise voxel group) from among a plurality of isolated voxel groups. At this time, the background region and the noise region are discriminated from a comparison between the labeling number assigned to each of the plurality of isolated voxel groups and the threshold of the labeling number. That is, in the state of FIG. 2C, an isolated voxel group having a labeling number of 2 or more is extracted as a noise region (noise voxel group).

  As a result, only a group of isolated voxels surrounded by the heart chamber that is the target region is extracted as a noise region, and noise is acquired by capturing voxels that belong to the noise region (noise voxel group) into the voxels of the heart chamber that is the target region. The region is removed.

  FIG. 2D shows a state where the noise region is removed. That is, the voxels with the labeling numbers “2” and “3” in FIG. 2C are taken into the voxels of the heart chamber, the voxels of the heart chamber are set to the voxel value “1”, and the other voxels are set to the voxel values. The binarized image set to “0” (illustrated by a black dot) is the image after removing the noise region in FIG. Incidentally, the image corresponding to the binarized image of FIG. 2D is FIG. 2D ′, and the heart chamber 110 and the background region 130 surrounding the heart chamber 110 are displayed as the image of the diagnostic region 100. ing. Then, the noise region (reference numeral 120 in FIG. 2 (a ′)) existing in the heart chamber 110 is removed.

  In the present embodiment, the principle of noise region removal described with reference to FIG. 2 is used. Furthermore, in the present embodiment, the background processing described below is performed in consideration of the case where the background region is separated into a plurality of isolated voxel groups.

  FIG. 3 is a diagram for explaining the background processing in the present embodiment. FIG. 3A shows a binarized image from which the heart chamber is extracted. That is, a binarized image in which a voxel value “1” is assigned to a voxel in the heart chamber in the diagnosis region 100 and a voxel value “0” is assigned to other voxels is shown. Note that the portion of the voxel value “0” is represented by a black dot.

  FIG. 3B is an inverted image obtained by inverting the binarized image of FIG. That is, the image is obtained by inverting the voxel value “0” in FIG. 3A to “1” and the voxel value “1” in FIG. 3A to “0”.

  Then, the labeling process described with reference to FIG. 2 is performed on the reverse image shown in FIG. That is, scanning is started from the upper left of the reverse image in FIG. 3B, and numbers are assigned in ascending order from 1 to each of a plurality of spatially connected voxel clusters. The result is shown in FIG.

  FIG. 3B ′ shows a state after the labeling process. As a result of starting scanning from the upper left of the inverted image, the label “1” is assigned to the isolated voxel group in the background portion located at the upper left of the image. However, since the background area is separated into a plurality of background portions (isolated voxel groups), a plurality of labeling numbers are assigned to the isolated voxel group corresponding to the background area. That is, the labels “1”, “2”, “3”, “6”, and “7” are isolated voxel groups as background regions, and the labels “4” and “5” are the portions of the heart chamber surrounded by the heart chamber. It is an isolated voxel group.

  As shown in FIG. 3B ', when the background area is separated into a plurality of background portions (isolated voxel groups), a large number of labels are attached to the background area. In this case, the noise region cannot be extracted simply by comparing the labeling number with its threshold value. Therefore, in the present embodiment, the following background processing is executed.

  FIG. 3C shows an image subjected to background processing. That is, a plurality of voxels constituting the outer edge of the diagnosis region 100 (black belt portion in the figure) are set as background voxels, and the voxel value is forcibly set to “1”. Then, a labeling process is performed on the background-processed image.

  FIG. 3D shows the result of labeling processing performed on the image shown in FIG. Since the voxel value of the outer edge of the diagnostic region 100 is set to “1” by the background processing, the background portion that has been divided into a plurality is connected by the outer edge voxels and extracted as a group of isolated voxels, The label “1” is attached. The labels “2” and “3” are assigned to the isolated voxel group in the portion surrounded by the heart chamber inside the heart chamber. As described above, as a result of background processing, a background portion that has been divided into a plurality of portions can be extracted as a group of isolated voxels.

  Then, as described with reference to FIG. 2, an isolated voxel group having a labeling number of 2 or more is extracted as a noise region (noise voxel group). As a result, only a group of isolated voxels surrounded by the heart chamber that is the target region is extracted as a noise region, and noise is acquired by capturing voxels that belong to the noise region (noise voxel group) into the voxels of the heart chamber that is the target region. The region is removed.

  FIG. 3E shows a state where the noise region is removed. That is, the voxels with the labeling numbers “2” and “3” in FIG. 3D are taken into the voxels of the heart chamber in the binarized image of FIG. 3A, and the noise shown in FIG. A binarized image after region removal is formed.

  In FIG. 2 and FIG. 3, the noise area removal processing and background processing have been described using two-dimensional image data as an example. These processes can also be applied to three-dimensional image data.

  FIG. 4 is a conceptual diagram for explaining processing on three-dimensional image data. FIG. 4A shows a binarized image from which the heart chamber 110 is extracted. That is, a binarized image in which the voxel value “1” is assigned to the voxel of the heart chamber 110 in the three-dimensional cube-shaped diagnostic region 100 and the voxel value “0” is assigned to the other voxels is shown.

  FIG. 4B shows a state in which the labeling process is performed on the inverted image after the inversion process is performed on the binarized image in FIG. The labeling process in the case of 3D is, for example, starting from one of the vertices of the cube of the diagnostic region 100, and numbering in ascending order from 1 for each of a plurality of voxel clusters having the same voxel value spatially connected. The process of attaching is performed. As a result, the label “1” is given to the isolated voxel group that is the background region surrounding the heart chamber, and the labels “2” and “2” are assigned to the isolated voxel group of the portion surrounded by the heart chamber inside the heart chamber. 3 "is given.

  Then, an isolated voxel group having a labeling number of 2 or more is extracted as a noise region (noise voxel group), and the voxel belonging to the noise region is taken into the voxel of the heart chamber as the target region, thereby removing the noise region.

  FIG. 4C shows a state where the noise region is removed. That is, the voxels with the labeling numbers “2” and “3” in FIG. 4B are taken into the voxels of the heart chamber, the voxels of the heart chamber are set to the voxel value “1”, and the other voxels are set to the voxel values. The binarized image set to “0” is the image after removing the noise region in FIG.

  When background processing is performed on the three-dimensional diagnosis region 100, the six outer surfaces of the cubic diagnosis region 100 are used as outer edges of the diagnosis region 100, and the voxel values of the voxels at the six outer surface positions are calculated. What is necessary is just to set to "1" compulsorily. Incidentally, even when the diagnosis region 100 is not in a cubic shape, background processing may be performed on the voxels of the outer surface portion constituting the diagnosis region 100. As a result, even when the diagnosis region 100 is three-dimensional, a plurality of divided background portions can be connected by the outer edge voxels and extracted as a group of isolated voxels.

  In the present embodiment, the principle described above is used. Returning to FIG. 1, the operation of the ultrasonic diagnostic apparatus of this embodiment will be described.

  The noise region removal block 30 removes the noise region from the inside of the heart chamber, which is the target region, based on the binarized data stored in the three-dimensional data memory (2) 20. First, in the inversion processing unit 32, inversion processing of binarized data (binarized image) is performed.

  Next, the background value processing unit 34 executes background processing for setting a plurality of voxels constituting the outer edge of the diagnostic region as background voxels. That is, a plurality of voxels constituting the outer edge of the diagnosis region are set as background voxels, and the voxel value is forcibly set to “1” (see FIG. 3C).

  Then, the labeling processing unit 36 separates a plurality of voxels other than the voxel of the heart chamber that is the target region into a plurality of isolated voxel groups, and assigns a labeling number to each of the plurality of isolated voxel groups (FIG. 3D). )reference).

  Further, based on the labeling number, the noise region determination unit 38 extracts, as a noise voxel group, an isolated voxel group that is surrounded by a heart chamber voxel that is the target region from among a plurality of isolated voxel groups. The region removal unit 39 removes the noise voxel group by taking the extracted voxel belonging to the noise voxel group into the cardiac chamber voxel (see FIG. 3E). The image data from which the noise area has been removed is stored in the three-dimensional data memory (3) 40.

  The display image forming unit 42 forms a display image based on the image data stored in the three-dimensional data memory (3) 40. That is, a display image from which the noise region in the heart chamber is removed is formed. As the display image, for example, an image using a volume rendering method described in detail in JP-A-10-33538 may be formed. Moreover, you may form arbitrary cross-sectional images in a three-dimensional diagnostic area | region. Note that the display image forming unit 42 may form a display image based on the image data stored in the three-dimensional data memory (1) 14 and before the noise area removal. The display image formed by the display image forming unit 42 is displayed on the monitor 50.

  The volume calculation unit 44 calculates the volume of the heart chamber based on the image data stored in the three-dimensional data memory (3) 40. The volume calculation unit 44 calculates the volume of the heart chamber by multiplying the number of voxels of the heart chamber voxel into which the voxel belonging to the noise voxel group is taken by the volume per voxel. The calculated volume value is displayed on the monitor 50.

  The preferred embodiment of the present invention has been described above. In the above-described embodiment, since the noise region is extracted based on the labeling number, for example, the volume of the noise region is calculated in order to extract the noise region. It is possible to omit the trouble and the operation of the user confirming the image and specifying the noise region. Further, since the noise region inside the heart chamber is specified by the labeling process, the contour shape of the heart chamber is not adversely affected. And since the noise area | region inside a heart chamber is removed, exact volume calculation is attained.

  The above-described embodiments are merely examples in all respects, and do not limit the scope of the present invention.

1 is a block diagram showing an overall configuration of an ultrasonic diagnostic apparatus according to the present invention. It is a figure for demonstrating the removal principle of the noise area | region in this embodiment. It is a figure for demonstrating the background process in this embodiment. It is a conceptual diagram for demonstrating the process with respect to three-dimensional image data.

Explanation of symbols

  32 inversion processing unit, 34 background value processing unit, 36 labeling processing unit, 38 noise region determination unit, 39 noise region removal unit.

Claims (5)

In an ultrasonic diagnostic apparatus that transmits and receives an ultrasonic wave in a diagnostic region including a target region to acquire an echo signal, and performs data processing on data in the diagnostic region formed from the echo signal.
Target area extraction means for extracting the target area from the diagnostic area based on the echo level;
A labeling processing unit that performs a labeling process for separating a region other than the target region in the diagnostic region into a plurality of isolated regions, and assigns a labeling number to each isolated region;
Based on the labeling number, a noise area extracting means for extracting an isolated area surrounded by the target area as a noise area from a plurality of isolated areas;
Noise area removing means for removing the noise area by taking the extracted noise area into the target area;
Having
An ultrasonic diagnostic apparatus. The ultrasonic diagnostic apparatus according to claim 1,
A background processing means for setting a background edge at the outer edge of the diagnostic region;
The labeling processing means surrounds the target area, and uses a lump of isolated areas connected by the background edge as a background area,
The noise area extracting means extracts an isolated area other than a background area as the noise area from the plurality of isolated areas.
An ultrasonic diagnostic apparatus. Transmitting / receiving means for transmitting / receiving ultrasonic waves in a diagnostic region including a target region, and acquiring voxel data of a plurality of voxels constituting the diagnostic region;
Target area extraction means for extracting a target area voxel corresponding to the target area from the plurality of voxels based on voxel data;
A plurality of voxels other than the target region voxel are separated into a plurality of isolated voxel groups in units of isolated voxel groups that are a group of a plurality of spatially connected voxels, and a labeling number is assigned to each of the plurality of isolated voxel groups Labeling processing means,
Noise voxel group extraction means for extracting, as a noise voxel group, an isolated voxel group that is surrounded by the target region voxel from a plurality of isolated voxel groups based on the labeling number;
Noise voxel group removing means for removing the noise voxel group by taking the extracted voxel belonging to the noise voxel group into the target area voxel;
Have
By setting a plurality of voxels constituting the outer edge of the diagnostic region as background voxels,
The labeling processing unit surrounds the target area voxel, and a group of isolated voxels connected by the background voxel is used as a background area,
The noise voxel group extraction means extracts an isolated voxel group other than a background region as the noise voxel group from the plurality of isolated voxel groups.
An ultrasonic diagnostic apparatus. The ultrasonic diagnostic apparatus according to claim 3.
The labeling processing unit assigns a labeling number in order from the isolated voxel group of the background region to each of a plurality of isolated voxel groups,
The noise voxel group extraction means discriminates an isolated voxel group in the background region from other isolated voxel groups based on a comparison between a labeling number assigned to each of the plurality of isolated voxel groups and a threshold of the labeling number.
An ultrasonic diagnostic apparatus. The ultrasonic diagnostic apparatus according to claim 4,
The diagnostic region is a three-dimensional space region;
Volume calculation means for calculating the volume of the target area from the number of voxels of the target area voxel in which the voxels belonging to the noise voxel group are captured
An ultrasonic diagnostic apparatus.

Images