JP4538260B2 - Specific area extraction method and apparatus - Google Patents

Description

  The present invention relates to a specific region extraction method and apparatus for extracting individual organs from an image obtained by a medical image diagnostic apparatus such as an X-ray CT apparatus or an MRI apparatus.

  In recent years, an image obtained by a medical image diagnostic apparatus such as an X-ray CT apparatus or an MRI apparatus has been actively used not only for diagnosis but also for treatment. Recently, research on automatically extracting a specific organ from a display image has been actively conducted. In particular, a specific region extraction process using a so-called region expansion (region growing) method is effective for extracting a specific region such as the liver. FIG. 1 is a diagram illustrating an example of a region expansion method. In this region glowing method, a pixel 10 as a starting point is specified, and whether or not surrounding pixels are pixels that satisfy a predetermined threshold condition with respect to the density of the pixel 10 is determined. In this case, the pixel is expanded as an extraction region.

However, since this region expansion method determines whether or not to expand in units of pixels, as shown in FIG. 2, if there is even a narrow gap for one pixel, it is not expected. However, it has a problem that it protrudes outside and expands outside the target area (so-called expansion overflow or expansion leakage). FIG. 2 is a diagram showing an example in the case where the region expansion method protrudes outside the target region. As is apparent from the figure, there is a problem that when the image is finely squeezed and the pixel in that portion satisfies a predetermined threshold condition, the protruding region 20 appears outside the target region. Therefore, as shown in Patent Document 1, when an overhang area 20 as shown in FIG. 2 exists, an area in which the overhang area 20 is removed by an artificial process has been proposed.
JP-A-5-317313

  However, the method described in Patent Document 1 cannot delete the protruding area unless it is manually operated on the display image using a pointing device such as a mouse, and the work takes a lot of time. Have the problem of being cumbersome.

  An object of the present invention is to provide a specific region extraction method and apparatus capable of extracting a specific organ region and the like with a simple operation while minimizing expansion overflow or expansion leakage.

  The first feature of the specific region extraction method according to the present invention is that, in the specific region extraction method for extracting a specific region based on a medical image, the determination as to whether or not to expand an already extracted region is described above. Using a circle or sphere centered on the already-extracted pixel point or a cone centered on the already-extracted pixel point, the pixel included in the circle, the sphere, or the cone, the circumference, the outer circumference of the sphere, or Determining whether or not a pixel value of a pixel included in a band-shaped region along the outer periphery of the cone or a calculation value using the pixel value satisfies a predetermined threshold value; and Includes a step of expanding the extracted area.

  When extracting a specific organ region from a medical image, instead of expanding in pixel units, a region defined by a figure such as a circle, a sphere, or a cone set in advance as a sense region, The region is expanded by determining whether the pixel value of the pixel included in the sense region or the operation value using the pixel value satisfies a predetermined threshold value. Therefore, even if a narrow gap that is smaller than the sense area is present in the medical image, if the gap is smaller than the sense area, expansion overflow or expansion leakage that protrudes from the sense area is reduced. A specific organ region or the like can be extracted by operation. The circle includes an ellipse.

  The second feature of the specific region extraction method according to the present invention is the specific region extraction method according to the first feature, wherein the step of displaying the state during the expansion of the step of expanding the already extracted region on the display means as needed. It is to include. This is to allow the state of the expansion process to be monitored from the outside.

  A third feature of the specific region extraction method according to the present invention is the specific region extraction method according to the first or second feature, wherein the radius of the circle and the sphere and the apex angle of the cone are variable. It is to have done. In this case, the sense region can be appropriately changed.

  A fourth feature of the specific region extraction method according to the present invention is the specific region extraction method according to the first, second, or third feature, wherein the extension process according to the determination as to whether or not the condition is satisfied is completed. A step of further setting a band-like area having a predetermined pixel width around the already-extracted area at a time point, and whether a pixel value of the band-like area or an operation value using the pixel value satisfies a predetermined threshold value And a step of expanding the band-like region as the extracted region when the condition is satisfied. In this method, a strip-shaped region is further set around the extracted region extracted using the sense region, and the extracted region is finally expanded using the pixel value of the strip-shaped region.

  A fifth feature of the specific region extraction method according to the present invention is a specific region extraction method for extracting a specific region based on a medical image, wherein a first position that designates an arbitrary position on the medical image as a first point is specified. A second step of rotating a moving radius around the first point to obtain a direction in which a concentration gradient on the moving radius is maximized, and setting the first point a predetermined distance in the direction A third step in which the second point is designated as the second point, and the second and third steps are repeatedly executed with the second point as the first point. And a fifth step of extracting the specific region by connecting a plurality of points obtained by repetition of the fourth step by interpolation processing. In this method, the moving direction and the moving distance are determined in accordance with the concentration gradient on the moving radius by using the moving radius having a variable length, and are sequentially moved.

  A sixth feature of the specific area extracting apparatus according to the present invention is that, in the specific area extracting apparatus that extracts a specific area based on a medical image, the determination as to whether or not to expand an already extracted area is performed. Using a circle or sphere centered on the already-extracted pixel point or a cone centered on the already-extracted pixel point, the pixel included in the circle, the sphere, or the cone, the circumference, the outer circumference of the sphere, or First determination means for determining whether or not a pixel value of a pixel included in a band-shaped region along the outer periphery of the cone, or an operation value using the pixel value satisfies a predetermined threshold value; In the case where the determination condition of the first determination means is satisfied, a first expansion means for expanding the already extracted region is provided. This is an invention of an apparatus corresponding to the specific region extraction method described in the first feature.

  A seventh feature of the specific area extracting apparatus according to the present invention is that, in the specific area extracting apparatus described in the sixth feature, a state in which the already extracted area is being expanded by the first extending means is displayed as needed. The display means to be provided. This is an invention of an apparatus corresponding to the specific area extracting method described in the second feature.

  An eighth feature of the specific area extracting apparatus according to the present invention is that the specific area extracting apparatus according to the first or second feature is configured such that the radius of the circle and the sphere and the apex angle of the cone are variable. It is to have done. This is an invention of an apparatus corresponding to the specific region extraction method described in the third feature.

  According to a ninth feature of the specific area extracting apparatus according to the present invention, in the specific area extracting device described in the first, second, or third characteristic, the expansion process according to the determination as to whether or not the condition is satisfied is completed. A band-shaped area setting means for further setting a band-shaped area having a predetermined pixel width around the already extracted area at a time point, and a pixel value of the band-shaped area or a calculation value using the pixel value satisfies a predetermined threshold condition A second determination unit that determines whether or not the second determination unit satisfies the condition; and a second expansion unit that expands the band-like region as the already-extracted region when the determination condition of the second determination unit is satisfied. is there. This is an invention of an apparatus corresponding to the specific area extracting method described in the fourth feature.

  According to a tenth feature of the specific area extracting apparatus of the present invention, in the specific area extracting apparatus that extracts a specific area based on a medical image, a first position that designates an arbitrary position on the medical image as a first point is specified. Position specifying means, direction determining means for obtaining a direction in which the concentration gradient on the radial radius is maximized by rotating the radial radius around the first point, and determining the first point in the predetermined direction. A second position specifying means for moving the distance and specifying the moved point as a second point; and the direction determining means and the second position specifying means with the second point as the first point. It is provided with repetitive execution means for repetitively executing processing, and area extraction means for extracting the specific area by connecting a plurality of points obtained by repetitive processing of the repetitive execution means by interpolation processing. This is an invention of an apparatus corresponding to the specific area extracting method described in the fifth feature.

  According to the region expansion method and apparatus of the present invention, there is an effect that it is possible to extract a specific organ or the like by a simple operation while minimizing expansion overflow or expansion leakage.

  Hereinafter, preferred embodiments of a specific area extracting method and apparatus according to the present invention will be described with reference to the accompanying drawings. FIG. 3 is a block diagram showing a hardware configuration of the entire specific area extracting apparatus to which the present invention is applied. This specific area extraction apparatus uses, for example, an X-ray CT image obtained by imaging a specific organ, for example, a heart, as a processing target image. This specific area extracting apparatus stores a central processing unit (CPU) 30 that controls the operation of each component, a main memory 31 that stores a control program for the entire apparatus, and a plurality of tomographic image data and programs. Operating a magnetic disk 32, a display memory 33 for temporarily storing image data for display, a CRT display 34 as a display device for displaying an image based on the image data from the display memory 33, and a soft switch on the screen Mouse 35 and its controller 36, a keyboard 37 having keys and switches for setting various parameters, a speaker 38, and a common bus 39 for connecting the above components.

  In this embodiment, a case where only the magnetic disk 32 is connected as a storage device other than the main memory 31 is shown, but in addition to this, a floppy disk drive, a hard disk drive, a CD-ROM drive, a magneto-optical disk ( MO) drive, ZIP drive, PD drive, DVD drive, etc. may be connected. Furthermore, it is possible to connect to various communication networks 3a such as a LAN (local area network), the Internet, and a telephone line via a communication interface so that image data can be exchanged with other computers or CT apparatuses 3b. It may be. Further, the exchange of image data may be performed by connecting a specific area extracting apparatus capable of collecting a tomographic image of a subject such as an X-ray CT apparatus or an MRI apparatus to the communication network 3a such as the LAN.

  Hereinafter, an example of the operation of the specific area extracting apparatus in FIG. 3 will be described with reference to the drawings. FIG. 4 is a diagram illustrating an example of a main flow executed by the specific area extracting apparatus. FIG. 5 is a flowchart showing details of the large grain region expansion processing of FIG. 6-13 is a figure which shows typically the mode of operation | movement of a specific area extraction apparatus. 3 operates according to the main flow of FIG. Hereinafter, details of the main flow will be described in the order of steps. This main flow is roughly divided into a large-grain area expansion process and a process of filling the periphery. The CPU 30 executes the large-area expansion process, and then fills the area that could not be filled by the large-area expansion process by executing a process of filling the periphery, thereby extracting a specific area. .

Next, details of the large-area expansion process in FIG. 5 will be described in the order of steps.
[Step S51]
Before executing the expansion process, it is determined whether or not expansion is possible. For this purpose, a determination area (sense area) and an expansion condition are set. Here, the determination region, that is, the sense region is a region surrounded by a circle having a radius R1 ≦ r ≦ R2 with the already extracted point as the center. Here, the expansion condition refers to what percentage (number) of pixels satisfying the upper limit, lower limit, density gradient, etc. of the CT value in the sense region. These numerical values are set by acquiring from the vicinity of the mouse position when the mouse is clicked or by reading what is recorded in advance on the magnetic disk.

[Step S52]
In this step, pixels around the extracted area are obtained. Then, the non-contact number register N is cleared to “0”.
[Step S53]
In this step, one of the peripheral pixels of the already extracted area is designated. For example, in the case of FIG. 6, the pixel 62 having the smallest x coordinate is designated from the peripheral pixels in the extracted region 61. However, the present invention is not limited to this, and the maximum x coordinate and the minimum or maximum y coordinate may be specified.
[Step S54]
In this step, the specified pixel is set as the center, and the radius R is moved in the range of R1 <R <R2, and it is checked whether or not various threshold values meet the expansion condition. When the expansion condition is met, it means that the non-target area has not been touched, and when the expansion condition is not met, the non-target area has been touched. Here, the threshold is a quantitative attribute relating to a pixel such as CT value, density, density gradient, the absolute value of the difference between adjacent pixels, and standard deviation.

[Step S55]
It is determined whether or not the radius R moved in the range of R1 <R <R2 has touched a non-target region (non-target region). That is, it is checked whether or not the extended condition is not satisfied. If contacted (yes), the process proceeds to step S56, and if not contacted (no), the process proceeds to step S57.
[Step S56]
Since it is determined that the contact has been made in the previous step S55, here, a circle having a radius (C × R) centering on the designated pixel is filled, and the filled area is made a part of the extraction area. Here, C is a constant satisfying 0 ≦ C <1.

[Step S57]
Since it is determined that the contact has not occurred in the previous step S55, here, a circle having a radius (C × R2) centering on the designated pixel is filled, and the filled area is set as a part of the extraction area. Here, C is a constant satisfying 0 ≦ C <1.
[Step S58]
“1” is added to the non-contact number register N.
[Step S59]
It is determined whether or not the above-described steps S54 to S58 have been executed for all the peripheral pixels. If not (no), the process proceeds to step S5A. If (Yes), the process proceeds to step S5B. move on. Note that the processing may not be performed for all the peripheral pixels, but may be performed for pixels by thinning out several pixels from the peripheral pixels.
[Step S5A]
One other pixel of the peripheral pixel is designated, and the process returns to step S54.
[Step S5B]
It is determined whether or not the value of the non-contact number register N is “0”. If “0” (yes), the process is terminated, and if the value is other than “0”, the process returns to step S52.

  For example, as shown in FIG. 6, when the extracted area 61 exists, pixels around the extracted area 61 are obtained, and the non-contact number register N is cleared to “0”. As a peripheral pixel of the already extracted region 61, a pixel 62 having the smallest x coordinate is set as a designated pixel. With the designated pixel 62 as the center, the radius R is moved within the range of the solid line radius R1 to the dotted line radius R2, and the expansion condition is examined. In the case of FIG. 6, since the radius R <b> 2 does not contact the non-target region, a circle having a radius (C × R <b> 2) centered on the designated pixel 62 is filled, and the filled region is a part of the extracted region 61. Here, the radius (C × R2) is a value approximated to the radius R1. Therefore, in FIG. 6, the circle with the radius R1 is filled. The above processing is sequentially executed for other peripheral pixels of the already extracted region 61 as shown in FIG. FIG. 7 shows a state in which the above-described processing is executed while the first designated pixel 62 sequentially moves in the clockwise direction with respect to the peripheral pixels of the already extracted region 61. FIG. 8 shows the result of the above-described processing in steps S54 to S58 performed on all the peripheral pixels in the already extracted area 61, and the extracted result exists as a new already extracted area 81. FIG. 6 is a diagram illustrating a state in which pixels around the designated area are designated as designated pixels 82. The already extracted area 81 in this case is an area larger than the already extracted area 61 of FIG. 6 by a radius (C × R2).

  As shown in FIG. 8, since a new extracted area 81 is formed, pixels around the extracted area 81 are obtained, and the non-contact number register N is cleared to “0”. As a peripheral pixel of the already extracted area 81, a pixel 82 having the smallest x coordinate is a designated pixel. With this designated pixel 82 as the center, the radius R is moved within the range of the solid line radius R1 to the dotted line radius R2, and the expansion condition is examined. In the case of FIG. 8, as in the case of FIG. 6, the radius R2 does not contact the non-target region, so that a circle with a radius (C × R2) centered on the designated pixel 82 is filled, and the filled region is the extraction region 81. Part of By sequentially executing the same processing for other peripheral pixels of the already extracted area 81 in FIG. 8, the already extracted area 81 in FIG. 8 becomes the new already extracted area 91 in FIG.

  As shown in FIG. 9, the same process as described above is executed for the new extracted area 91. In the case of the already extracted region 91, as shown in the figure, when the expansion condition is examined by moving the radius R in the range of the solid line radius R1 to the dotted line radius R2 with the designated pixel 9n as the center, an arbitrary radius R Touches the non-target area. Therefore, a circle having a radius (C × R) centering on the designated pixel 9n is filled, and the filled area is made a part of the extracted area 91. By sequentially executing the same processing for other peripheral pixels of the already extracted area 91 of FIG. 9, the already extracted area 91 of FIG. 9 becomes like the new already extracted area 101 of FIG. By repeating the same process for the already extracted area 101, a new already extracted area 111 as shown in FIG. 11 is extracted, and an already extracted area 121 as shown in FIG. 12 is finally extracted.

When the process of FIG. 5 is completed, an area 121 as shown in FIG. 12 is finally extracted. For this area 121, a process of filling the periphery of FIG. 4 is executed. There are two methods for filling the periphery. One is a method of filling the periphery with a curve and a method of filling with a circle (sphere) having a smaller radius than the above-described processing. FIG. 13 is a diagram illustrating an example of a method of filling the periphery with a curve. As is apparent from the figure, the periphery of the finally extracted region 121 is expanded while increasing the threshold value line by line. The number of lines in the case of thickening may be obtained from the following equation using the above-described radius R2 in addition to one line.
R2−C × R2 + α (α is an arbitrary value determined empirically)
The region 123 obtained in this way is the final extraction region.

In the above-described embodiment, the case where the radius R is moved in the range from the solid line radius R1 to the dotted line radius R2 to examine the expansion condition has been described. However, the radius R is from the center (radius R = 0) to the radius R2. The expansion condition may be examined by moving within the range of. In this case, when the radius R at the time of contact is larger than the radius R1, a circle having a radius (C × R) centering on the designated pixel is filled, the filled area is made a part of the extraction area, and conversely the radius at the time of contact If R is less than or equal to radius R1, nothing may be done. As a result, similar processing can be executed.
In the above embodiment, the two-dimensional image has been described. However, in the case of a three-dimensional image, similar processing may be performed using a sphere instead of a circle as shown in FIG.

  FIGS. 15 to 18 are diagrams illustrating an example in which a specific region is extracted from a tomographic image adjacent to the already extracted region 123 extracted from one tomographic image. FIG. 15 is a diagram showing an already extracted region 123 extracted on the tomographic image 155 by the above-described processing. In FIG. 15A, tomographic images 152, 154 and 156 are shown before and after the tomographic image 155. Then, a tomographic image (interpolated image) 150 orthogonal to the tomographic image 155 is acquired. The already-extracted region 123 existing on the cross section orthogonal to the interpolation image 150 and the tomographic image 155 becomes the start point (region) 15 a of the interpolation image 150. FIG. 15B is a diagram showing a state of the start point (region) 15 a on the interpolation image 150.

  FIG. 16 is a diagram illustrating an example of expansion in the slice thickness direction. FIG. 16 shows an example of processing when extracting an extended region of the adjacent tomographic image 154 based on the start point (region) 15a on the tomographic image 155 extracted by the processing of FIG. An arbitrary pixel at the start point (area) 15a is designated as the designated pixel. As in the case described above, this designation may designate a pixel having the smallest or largest y coordinate, or may designate the middle point of the start point (region) 15a as the designated pixel. When the designated pixel of the start point (area) 15a is determined, the expansion condition is examined in the same manner as described above by moving the radius R within the range of R1 <R <R2 as the sense area with the designated pixel as the center. This expansion condition is performed for a tomographic image 154 included in a circle 161 having a radius R, as shown in FIG. Then, it is determined whether or not the radius R moved in the range of R1 <R <R2 is in contact with a non-target region (non-target region), and the filling range is controlled depending on whether or not the contact is made. Extract the region. As described above, by executing the same processing as the above-described processing for the start point (region) 15b, the extended region 16a as shown in FIG. 16C is extracted. The expanded area 16a indicates the result of executing the process of filling the periphery of FIG.

  FIG. 17 is a diagram showing a modification of FIG. In FIG. 16, the case where the region extraction processing is performed based on the circle 161 having the radius R as the sense region has been described. However, in this embodiment, an elliptical shape 171 is used instead of the circle 161 as the shape of the sense region. It is a thing. In this case as well, as described above, an arbitrary pixel at the start point (area) 15a is designated as the designated pixel. When the designated pixel of the start point (area) 15a is determined, the extension condition is examined in the same manner as described above by moving the size of the ellipse as the sense area in the range from the ellipse 171 to the ellipse 172. The area is extracted by controlling the filling range. In this way, by executing the same processing as the above-described processing for the start point (region) 15b, the extended region 17a as shown in FIG. 17C is extracted.

  FIG. 18 is a diagram showing a modification of FIG. FIG. 18 illustrates a case where region extraction processing is performed using a cone having an angle Θ having a designated pixel as a vertex as a sense region. Also in this case, as described above, an arbitrary pixel Pn-1 at the start point (region) 15a is designated as the designated pixel. When the designated pixel Pn-1 of the start point (area) 15a is determined, a cone having an angle Θ having the designated pixel Pn-1 as a vertex is set as a sense region, and the size of the angle Θ of the cone is 0 ≦ Θ <. By moving within the range of θ, the expansion condition is examined in the same manner as described above, and the area is extracted by controlling the filling range. Here, θ is set to about π / 4. In this way, by executing the same processing as the above-described processing for the start point (region) 15b, the extended region 18a as shown in FIG. 18C is extracted. In addition to the circle, ellipse, and cone, the shape of the sense region may be a rectangle or other shapes.

  FIG. 19 is a diagram illustrating an example of a magnetic disk device storing expansion conditions. As shown in the figure, the radius R1 and its expansion condition may be stored for each organ type, and the condition may be set for each target organ. For example, for the liver, the radius is 10 pixels, the expansion condition is A, for the brain, the radius is 8 pixels, the expansion condition is B, for the large intestine, the radius is 9 pixels, and the expansion condition is C. . These extended conditions shall use what combined the above-mentioned various conditions suitably.

  20 and 21 are diagrams showing a method called a snake that extracts a region using a stretchable curve. First, in this snake method, when the region 200 is extracted, the stretchable curve 201 is contracted and stopped at the boundary of the region 200, so that the region 200 can be extracted. However, actually, as shown in FIG. 21, noises 202 and 203 exist in the tomographic image. Therefore, the curve 201 is caught by the noises 202 and 203 and stopped when the contraction occurs, as shown in FIG. 21. There was a problem that accurate extraction was not possible.

  Therefore, in this embodiment, the moving direction and the moving distance are determined according to the concentration gradient on the moving radius using the moving radius having a variable length, and the moving radius is sequentially moved. FIG. 22 is a flowchart showing an example of a specific area extraction method for extracting a specific area according to the concentration gradient on the radius vector.

Details of the specific area extraction processing of FIG. 22 will be described in the order of steps.
[Step S221]
The radial length is set before executing the extraction process. The radial length is set to a predetermined length for each target organ. Then, the first point P as the start point is designated with a mouse or the like. As shown in FIG. 23, when the point 231 clicked first does not exist on the contour, the moving radius 232 is rotated around the point 231 clicked with the mouse, and the density in the direction along the moving radius is increased. The maximum point of the gradient is searched, and this is set as the start point (P point) 233.
[Step S222]
The coordinates of the point P designated as the start point 233 are recorded on the memory.

[Step S223]
In this step, a density gradient image is created for the tomographic image. For example, a difference in density between adjacent pixel points may be taken to obtain an absolute value and used as a density gradient image. Or you may make a wide area filter act.
[Step S224]
In this step, the radius is rotated around the point P, and an angle Θ that maximizes the sum of the concentration gradients on the radius of rotation is obtained. The length of the moving radius here is variable between Lmin and Lmax. The angle Θ is an angle with respect to the x coordinate. That is, as shown in FIG. 24, the moving radius 242 is rotated around the P point 241 to obtain the angle Θ that maximizes the sum of the concentration gradients on the moving radius.

[Step S225]
In this step, the point P is moved in the direction of the angle Θ obtained in the previous step by a distance corresponding to the radial length. In FIG. 24, the P point 241 is moved by a length obtained by multiplying the radial length by a coefficient (0.7) to obtain a new P point 243. In this step, the radius vector length is made variable, and the radius vector is determined so as to maximize the value obtained by dividing the concentration gradient on the radius vector by the radius vector. Since the radius vector showing the maximum appears in two places before and after the point P, the traveling direction is limited to, for example, clockwise, and the radius matching the direction is defined as the radius length at the point P. decide. Further, the direction Θ2 in which the distance between the coordinates of each P point, the coordinates (X1, Y1) of the P point 241 and the coordinates (X3, Y3) of the P point 244 is maximum may be adopted. Furthermore, as shown in FIG. 24, the contour changes rapidly at the P point 244 portion, and thus the radial length is small at such a portion. In this way, when the contour irregularity is intense and changes rapidly, the contour can be changed by changing the radial length by finding a location where the standard deviation of the concentration gradient is large beforehand or by finding it at any time. It can be extracted accurately.
[Step S226]
In this step, the coordinates after the movement in the previous step are recorded in the memory.

[Step S227]
It is determined whether or not the point P has returned to the first starting point, that is, whether or not the area has been circled. If the circle has been circled (yes), the process proceeds to the next step S228. The process returns to S224.
[Step S228]
Spline interpolation is performed between the coordinates of each P point recorded in the memory, and a curve is connected. The curve drawn in this way becomes the extraction region.

  FIG. 25 is a diagram illustrating a special case in the specific area extraction processing of FIG. That is, as shown in FIG. 25, when the P point 250 exists at a place where the two regions 251 and 252 are in contact with each other, there are two local maximum directions in step S225. Need to be processed to determine which direction is A or B. Here, the shorter one of the distance to the center of gravity 253 of the point that has been reached is selected. Other than this, the shorter distance to the start point 254 may be selected, or the shorter distance from the center of gravity 253 and the total distance to the start point 254 may be selected.

It is a figure which shows an example of the area | region expansion method. It is a figure which shows an example at the time of projecting out of the target area | region in the area | region expansion method. It is a block diagram which shows the hardware constitutions of the whole specific area extraction apparatus with which this invention is applied. It is a figure which shows an example of the main flow which a specific area extraction apparatus performs. It is a flowchart figure which shows the detail of the large grain area | region expansion process of FIG. It is the 1st figure which shows typically the mode of operation of a specific field extraction device. It is the 2nd figure which shows typically the mode of operation | movement of a specific area extraction apparatus. It is a 3rd figure which shows typically the mode of operation | movement of a specific area extraction apparatus. It is a 4th figure which shows typically the mode of operation | movement of a specific area extraction apparatus. It is the 5th figure which shows typically the mode of operation of a specific field extraction device. It is the 6th figure which shows typically the mode of operation of a specific field extraction device. It is a 7th figure which shows typically the mode of operation | movement of a specific area extraction apparatus. It is the 8th figure which shows typically the mode of operation | movement of a specific area extraction apparatus. It is a figure which shows the example in the case of performing a specific area extraction process using a sphere in the case of a three-dimensional image. It is a figure which shows an example in the case of extracting a specific area | region with respect to the tomographic image adjacent to it using the already extracted area | region extracted in one tomographic image, It is extracted on a tomographic image by specific area extraction processing It is a figure which shows the already extracted area | region. It is a figure which shows an example in the case of extending in a slice thickness direction. It is a figure which shows the modification of FIG. It is a figure which shows another modification of FIG. It is a figure which shows an example of the magnetic disk apparatus which stored the expansion conditions. It is a figure which shows the method called a snake which extracts an area | region using an elastic curve. It is a figure explaining the problem of the snake method of FIG. It is a flowchart figure which shows an example of the specific area extraction method which extracts a specific area according to the density gradient on a moving radius. It is a figure which shows typically operation | movement of step S221 of FIG. It is a figure which shows typically operation | movement of step S224 of FIG. 22, and step S225. It is a figure which shows the special case in the specific area extraction process of FIG.

Explanation of symbols

10 ... Pixel 20 ... Extrusion area 30 ... Central processing unit (CPU)
31 ... main memory 32 ... magnetic disk 33 ... display memory 34 ... CRT display 35 ... mouse 35a ... cursor 36 ... controller 37 ... keyboard 38 ... speaker 39 ... common bus 3a ... communication network 3b ... other computers or CT devices 61, 81 , 91, 101, 111, 121, 123... Already extracted regions 62, 82, 9n.

Claims (4)

In a specific area extraction method for extracting a specific area based on a medical image,
A determining step sensing area is whether predetermined extension satisfies the conditions comprising a succession plurality of pixels already from the pixels of the extracted pre-extraction region,
Wherein when the extended condition is satisfied viewed contains a expansion step of expanding the already extracted region based on the sensing area,
The specific area extraction method according to claim 1, wherein the expansion condition is that a predetermined ratio of pixels among the plurality of pixels in the sense area satisfies a predetermined threshold condition . In the specific area extraction method according to claim 1,
The determination step, the repetition and changing the range of the sense region, the range of the modified sense region and the determination of whether the expansion conditions are satisfied, while the extended condition is satisfied in the range of the sense region Continue the change to a predetermined size,
The expansion step is to add a region smaller than the satisfying expansion condition region obtained as a result of the determination and change ranges repeated in cell Nsu region already extraction region setting a new pre-extraction region A specific region extraction method characterized by the above. In a specific area extraction device that extracts a specific area based on a medical image,
A first determination means for determining whether extended condition is satisfied or not the sense area consisting of contiguous plural pixels already from the pixels of the extracted pre-extracted region is predetermined,
A first extension means for extending the extracted area based on the sense area when the extension condition is satisfied ,
The specific region extraction apparatus , wherein the expansion condition satisfies a predetermined threshold condition for a predetermined proportion of pixels among the plurality of pixels in the sense region . In the specific area extraction device according to claim 3 ,
It said first determining means, said a change in the scope of the sense region, the range of the modified sense region repeated and determines whether the expansion conditions are satisfied, while the extended condition is satisfied the sense region Continue to change the range to a predetermined size,
It said first expansion means, the range of changes and results area smaller than the area that satisfies the resulting said expanded condition in addition to the already extracted regions new already extraction region determination was repeated in degrees Nsu region A specific area extracting apparatus characterized by setting the value.

Images