JP5777053B2 - Image processing device - Google Patents

Description

  Embodiments described herein relate generally to an image processing apparatus.

  When treating cancer tissue in the brain, it is important to treat the nerve tissue adjacent to the cancer tissue while preserving it. Therefore, it is desired to develop a technique that can display the positional relationship between a tumor and a nerve as easily as possible and can be used as guidance for making a treatment plan.

  An object of the present invention is to provide an image processing apparatus capable of improving image diagnostic ability.

The image processing apparatus according to the present embodiment includes a storage unit that stores a three-dimensional neural image related to the neural tissue of the brain and a three-dimensional tumor image related to the tumor of the brain, and a composite image of the neural image and the tumor image. A determination unit that determines an index value relating to an anatomical running state of a nerve region derived from the neural image with respect to a tumor region derived from the tumor image in the composite image; A color image generator that generates a two-dimensional information image based on the composite image, in which an information value corresponding to the index value is assigned to the tumor region included in the composite image, and displays the generated information image And the display unit calculates a distance from each of a plurality of pixels included in the tumor region to the nerve region as the index value .

1 is a diagram illustrating a configuration of an image processing apparatus according to an embodiment. The figure for demonstrating the process of the nerve parameter | index value determination part which concerns on Example 1 of this embodiment. The figure for demonstrating the process of the nerve parameter | index value determination part which concerns on Example 2 of this embodiment. The figure which shows an example of the nerve area | region which has entered the local area | region of FIG. The figure which shows the positional relationship of a tumor and a nerve bundle. The figure which shows the structure of the nerve parameter | index value determination part which concerns on Example 3 of this embodiment. The figure which shows the typical flow of the image process which concerns on Example 3 performed under control of the control part which concerns on Example 3 of this embodiment. The figure which shows the tumor area | region and nerve area | region in a color image which are displayed by the display part in step SA8 of FIG. FIG. 8 is another view showing a tumor region and a nerve region in a color image displayed by the display unit in step SA8 in FIG. The figure which shows the structure of the nerve parameter | index value determination part which concerns on the modification of this embodiment. The figure which shows the typical flow of the image process which concerns on the modification performed under control of the control part which concerns on the modification of this embodiment. It is a figure for demonstrating the process of the arrival point specific | specification part in FIG.11 SB2, and is a figure which shows the tumor area | region RT, the nerve area | region RN, and the brain area | region surface SB in a multiplex composite image. FIG. 12 is a diagram showing an example of a multiple color image displayed by the display unit in step SB4 of FIG. The figure which shows the simulated brain image (Broadman's brain map) displayed by the display part which concerns on the modification of this embodiment. The figure which shows the example of a display of the Broadman's brain map with which the arrival point of the nerve area | region emphasized displayed by the display part which concerns on a modification.

  The image processing apparatus according to this embodiment will be described below with reference to the drawings.

  FIG. 1 is a diagram illustrating a configuration of an image processing apparatus 1 according to the present embodiment. As shown in FIG. 1, the image processing apparatus 1 includes a storage unit 2, an extraction unit 3, a composite image generation unit 4, a nerve index value determination unit 5, a display image generation unit 6, a display unit 7, an operation unit 8, and a control. Part 9 is provided.

  The storage unit 2 stores data of a three-dimensional functional image (hereinafter referred to as a nerve image) related to the neural tissue of the subject's brain. A nerve image is generated by applying a fiber tracking technique (also called tractography) to a diffusion weighted image generated by imaging the brain of a subject with a magnetic resonance imaging apparatus. The voxel value of each voxel of the neural image corresponds to an index value related to diffusion. As index values relating to diffusion, for example, ADC (apparent diffusion coefficient), FA (fractional anisotropy), RA (relative anisotropy), VR (volume ratio), and the like are used. The fiber tracking technique is a method of tracking a voxel in the direction indicated by the eigenvector corresponding to the maximum eigenvalue of the diffusion tensor. In addition, the storage unit 2 stores data of a three-dimensional morphological image (hereinafter referred to as a brain image) relating to the subject's brain. A brain image is generated by imaging the brain of a subject with a medical image diagnostic apparatus such as a magnetic resonance imaging apparatus, an X-ray computed tomography apparatus, an X-ray diagnostic apparatus, or a nuclear medicine diagnostic apparatus. The brain image and the nerve image are images related to the same subject and are captured at substantially the same time. It is assumed that a tumor has occurred in the brain of the subject. The tumor may be a malignant tumor such as cancer or a benign tumor. The storage unit 2 stores an image processing program to be described later.

  Here, it is assumed that the three-dimensional image is composed of a plurality of pixels (voxels) defined by orthogonal three-dimensional coordinates of X, Y, and Z. That is, the three-dimensional image is a so-called volume image. It is assumed that a two-dimensional image is composed of a plurality of pixels (pixels) defined by orthogonal two-dimensional coordinates of X and Y.

The extraction unit 3 extracts the tumor region from the brain image based on the voxel values that the tumor region can have. The extracted morphological image relating to the tumor region will be referred to as a tumor image. Specifically, the extraction unit 3 extracts a tumor region from the brain image by performing image processing such as threshold processing and region extraction processing on the brain image. Tumor image data relating to the tumor region is stored in the storage unit 2. The extraction unit 3 extracts the brain region surface from the brain image based on the voxel values that the brain region surface may have. The extracted morphological image related to the surface of the brain region is called a brain surface image. The brain surface image data is stored in the storage unit 2.

  The composite image generation unit 4 generates a composite image of a nerve image and a tumor image by image processing. When the nerve image and the tumor image are generated by the magnetic resonance imaging apparatus, the nerve image and the tumor image are mechanically aligned at the imaging stage. In this case, alignment by image processing between the nerve image and the tumor image is unnecessary. On the other hand, when the tumor image is generated by a medical image diagnostic apparatus other than the magnetic resonance imaging apparatus, the mechanical alignment between the nerve image and the tumor image is not performed. In this case, it is necessary to align the nerve image and the tumor image by image processing. For example, the composite image generation unit 4 performs alignment using at least one anatomically identical feature point set for each of the nerve image and the tumor image. After the alignment, the composite image generation unit 4 combines the aligned neural image and the tumor image to generate a composite image. The composite image generation unit 4 can also generate a composite image (hereinafter referred to as a multiple composite image) of a nerve image, a tumor image, and a brain surface image by a similar method.

  The nerve index value determination unit 5 represents an index value that represents the anatomical running state of the nerve region with respect to the tumor region based on the geometric arrangement relationship between the tumor region and the nerve region included in the composite image or the multiple composite image. (Hereinafter referred to as a nerve index value). Specifically, the nerve index value is interrupted in the tumor area, which is the distance between the tumor area and the nerve area, the amount of the nerve area included in the tumor area, and the total nerve area entering the tumor area. It is defined as the percentage of nerve area that is present.

  The display image generation unit 6 combines a two-dimensional image (hereinafter referred to as a color image) in which a color value corresponding to the determined nerve index value is assigned to a tumor region included in the composite image. Occurs based on. Specifically, the display image generation unit 6 assigns RGB (red, green, blue) color values corresponding to the nerve index values to the tumor region included in the composite image. At this time, the display image generation unit 6 may assign a specific color value to a nerve region included in the composite image. Next, the display image generation unit 6 performs three-dimensional image processing on the synthesized image to which the color value is assigned, and generates a two-dimensional color image. Examples of 3D image processing include volume rendering, surface rendering, pixel value projection processing, MPR (multi-planar reconstruction), and the like. As the pixel value projection process, typically, a maximum intensity projection (MIP) method may be employed. The color value assignment may be performed after the three-dimensional image processing instead of the three-dimensional image processing. Further, the display image generation unit 6 performs a similar three-dimensional image processing to generate a two-dimensional image (hereinafter referred to as a multiple color image) in which a color value corresponding to a nerve index value is assigned to a tumor region based on a multiple composite image. Can also be generated. The processing target of the three-dimensional image processing is not limited to a composite image or a multiple composite image. For example, the display image generation unit 6 may perform three-dimensional image processing on a nerve image, a tumor image, or a brain image.

  The display unit 7 displays various display images such as the generated color image and multiple color image on a display device. Examples of the display device include a CRT display, a liquid crystal display, an organic EL display, and a plasma display.

  The operation unit 8 receives various commands and information inputs from the user via the input device. Examples of the input device include a pointing device such as a mouse and a trackball, a selection device such as a switch button, or a keyboard.

  The control unit 9 functions as the center of the image processing apparatus 1. The control unit 9 reads out the image processing program stored in the storage unit 2, and in accordance with an image processing program for color display of a tumor or the like, each unit 2, 3, 4, 5, 6, 7, 8 is controlled.

  Next, details of the image processing apparatus 1 according to the present embodiment will be described. The operation of the image processing apparatus 1 is divided into a plurality of embodiments according to the type of nerve index value. Each example will be described below.

[Example 1]
The nerve index value according to the first embodiment is set to the distance between the tumor region and the nerve region in the composite image or the multiple composite image (hereinafter referred to as a tumor-nerve distance). The nerve index value determining unit 5 according to the first embodiment calculates the tumor-nerve distance based on the geometrical arrangement relationship between the tumor region and the nerve region. In the following description, it is assumed that the tumor-nerve distance calculation is performed using a composite image.

  FIG. 2 is a diagram for explaining processing of the nerve index value determination unit 5 according to the first embodiment, which is performed under the control of the control unit 9. As illustrated in FIG. 2, it is assumed that the composite image includes a tumor region RT and a plurality of nerve regions RN. Each nerve region RN corresponds to one nerve bundle in the brain. First, the nerve index value determination unit 5 specifies a plurality of voxels VO included in the surface, inside, or the whole of the tumor region RT. It is arbitrarily set by the user via the operation unit 8 whether the voxel to be identified is a front voxel, an internal voxel, or an entire voxel. The identified voxel will be referred to as a processing target voxel VO. When a plurality of processing target voxels VO are specified, the nerve index value determination unit 5 performs the following processing for each processing target voxel VO. First, the nerve index value determination unit 5 calculates the perpendicular PE of each processing target voxel VO. When the perpendicular PE is calculated, the nerve index value determination unit 5 identifies the voxel VN of the nerve region RN that is present at the closest position to the processing target voxel VO on the perpendicular PE that goes from the processing target voxel VO to the outside of the tumor region RT. To do. Then, the nerve index value determination unit 5 calculates the distance (tumor-nerve distance) between the voxel RN of the specified nerve region and the processing target voxel VO. When the tumor-nerve distance is calculated for each processing target voxel VO in this way, the nerve index value determination unit 5 ends the calculation process of the nerve index value. Data of the coordinates of each processing target voxel VO and the distance between the tumor and the nerve is supplied to the display image generation unit 6.

  Using the coordinates of each processing target voxel and the tumor-nerve distance data from the nerve index value determination unit 5, the display image generation unit 6 uses each processing target voxel as a color value corresponding to the tumor-nerve distance. Is assigned. Specifically, the display image generation unit 6 holds a color table in which a tumor-nerve distance is associated with a color value. Using this color table, the display image generation unit 6 specifies a color value corresponding to the tumor-nerve distance for each processing target voxel and assigns the specified color value to the processing target voxel. For example, a darker red color is assigned as the tumor area is approached, and another color such as a dark blue color moving away from the tumor area may be assigned. Then, the display image generation unit 6 performs the above-described three-dimensional image processing on the composite image to which the color value is assigned, and generates a two-dimensional color image. The generated color image is displayed on the display unit 7. At this time, in order to present the correspondence between the tumor-nerve distance and the color on the screen to the user, a color bar indicating the correspondence between the tumor-nerve distance and the color value is displayed together with the color image. good.

  Each pixel of the tumor region included in the color image is displayed in a color corresponding to the tumor-nerve distance. Therefore, the user can easily grasp the running state of the nerve near the tumor, specifically the distance between the tumor and the nerve, by identifying the color of the pixel in the tumor region.

  Note that the processing in the first embodiment is performed on the composite image. However, the present embodiment is not limited to this. For example, the above-described processing may be performed on the multiple composite image. Thereby, the positional relationship among the tumor region, the nerve region, and the brain surface can be clearly grasped with one image.

  The displayed image is not limited to a color image or a multiple color image. For example, the display unit 7 may display a color image or a multiple color image superimposed on an MPR image or volume rendering image of a three-dimensional brain image.

[Example 2]
The nerve index value according to the second embodiment is the amount of the nerve region included in the tumor region in the composite image or the multiple composite image. The nerve index value determination unit 5 according to the second embodiment calculates the amount of the nerve region in the tumor region based on the geometric arrangement relationship between the tumor region and the nerve region. In the following description, it is assumed that the amount of nerve area is calculated using a composite image.

  FIG. 3 is a diagram for explaining processing of the nerve index value determination unit 5 according to the second embodiment, which is performed under the control of the control unit 9. As shown in FIG. 3, a nerve region RN may enter the tumor region RT. Hereinafter, a method for calculating the amount of the nerve region RN entering the tumor region RT in the tumor region RT will be described in detail. First, the nerve index value determination unit 5 sets a plurality of local regions RL in the tumor region RT. The local region RL includes a plurality of voxels. The size and shape of each local region RL can be arbitrarily set by the user or the like via the operation unit 8. The plurality of local regions RL may be set so as to be distributed over the entire tumor region RT, or may be set so as to be distributed locally in a specific region. When a plurality of local regions RL are specified, the nerve index value determination unit 5 performs the following processing for each local region. The nerve index value determination unit 5 calculates the amount of the nerve region RN included in each local region RL. The amount of the nerve region RN is defined by the number of voxels constituting the nerve region RN or the number of nerve regions RN. The number of voxels or the number can be arbitrarily set by the user via the operation unit 8. Note that the nerve image has a spatial resolution capable of recognizing the nerve region RN in units of nerve bundles one by one in image processing.

  FIG. 4 is a diagram illustrating an example of the nerve region RN entering the local region RL. FIG. 4A shows a state in which three nerve regions RN exist in the local region RL, and FIG. 4B shows one nerve region RN in the local region RL. The situation when it exists is shown. For example, the nerve index value determination unit 5 searches for voxels in the local region RL using the voxel value range that the nerve region RN may have as an integration condition, and sequentially integrates voxels that satisfy the integration condition. Then, the nerve index value determination unit 5 counts the number of integrated voxels. Alternatively, the nerve index value determination unit 5 regards each integrated voxel region as one nerve region RN, and counts the number of voxel regions as the number of nerve regions RN. In this way, the nerve index value determination unit 5 can calculate the amount of the nerve region RN for each local region RL. When the amount of the nerve region RN is calculated for each local region RL, the nerve index value determining unit 5 ends the nerve index value calculation process. Data on the coordinates of each local region RL and the number of voxels in the nerve region RN or the number of nerve regions is supplied to the display image generation unit 6.

  Using the data of the coordinates of each local region and the number of voxels in the nerve region or the number of nerve regions from the nerve index value determination unit 5, the display image generation unit 6 displays voxels in the nerve region in each local region. A color value corresponding to the number of nodes or the number of nerve regions is assigned. Specifically, the display image generation unit 6 holds a color table in which the number of voxels in the nerve region or the number of nerve regions is associated with the color value. Using this color table, the display image generating unit 6 identifies a color value corresponding to the number of voxels in the nerve region or the number of nerve regions for each local region, and the identified color value is stored in the local region. Assign to voxels. For example, a deep red color is assigned as the number of voxels in the nerve region increases, and another color such as a deep blue color is assigned as the number decreases. Further, it is preferable that dark red is assigned as the number of nerve regions increases, and other colors such as dark blue are assigned as the number of nerve regions decreases. Then, the display image generation unit 6 performs the above-described three-dimensional image processing on the composite image to which the color value is assigned, and generates a two-dimensional color image. The generated color image is displayed on the display unit 8.

  Each pixel of the tumor area included in the color image is displayed in a color corresponding to the amount of the nerve area existing in the vicinity thereof. For example, a portion where the nerve region in the tumor region is not included and a portion where the nerve region in the tumor region is included are displayed in different colors. Therefore, the user can easily grasp the running state of nerves in the tumor, specifically the amount of nerves in the tumor, by observing the color of the pixels in the tumor region.

  Note that the processing in the second embodiment is performed on the composite image. However, Example 2 is not limited to this. For example, the above-described processing may be performed on the multiple composite image. Thereby, the positional relationship among the tumor region, the nerve region, and the brain surface can be clearly grasped with one image.

  The displayed image is not limited to a color image or a multiple color image. For example, the display unit 7 may display a color image or a multiple color image superimposed on an MPR image or volume rendering image of a three-dimensional brain image.

[Example 3]
The nerve index value according to the third embodiment is a ratio of discontinuous nerve regions in the tumor region in all the nerve regions entering the tumor region included in the composite image or the multiple composite image. The nerve index value determination unit 5 according to the third embodiment determines a ratio of discontinuous nerve regions in all the nerve regions in the tumor region based on the geometric arrangement relationship between the tumor region and the nerve region. In the following description, it is assumed that the ratio is determined using a composite image.

  First, the background of the third embodiment will be described. FIG. 5 is a diagram illustrating a positional relationship between a tumor and a nerve bundle. As shown in FIG. 5, the positional relationship between the tumor and the nerve bundle is roughly divided into four types. That is, when the nerve bundle is close to the tumor, that is, when it has not entered the tumor ((a) of FIG. 5), when the nerve bundle has entered the tumor and is not interrupted within the tumor (( b)), when the nerve bundle has entered the edema around the tumor (FIG. 5 (c)), when the nerve bundle has entered the tumor and is interrupted within the tumor (FIG. 5 (d)) There is.

  When the tumor is malignant, the structure of the tumor cell is often destroyed. Therefore, the nerve bundle passing through the malignant tumor is difficult to visualize. When the tumor is benign, the structure of the tumor cells is often not broken. Therefore, the nerve bundle is easily visualized as compared with the case of a malignant tumor. In other words, if the nerve bundle connected to the tumor is interrupted inside the tumor, it can be said that the tumor is suspected of being malignant. That is, the ratio of the discontinuous nerve region to the entire nerve region entering the tumor region functions as one index for measuring the malignancy of the tumor. Therefore, the image processing apparatus 1 according to the third embodiment determines this ratio, generates a color image to which a color value corresponding to the determined ratio is assigned, and displays the generated color image.

  FIG. 6 is a diagram illustrating the configuration of the nerve index value determination unit 5 according to the third embodiment. As illustrated in FIG. 6, the nerve index value determination unit 5 includes an inclusion relationship determination unit 51, a break determination unit 52, a ratio setting unit 53, and a ratio calculation unit 54.

  The inclusion relationship determination unit 51 determines whether a nerve region has entered a tumor region in the composite image based on the voxel value. When the discontinuity determination unit 52 determines that the nerve region has entered the tumor region by the inclusion relationship determination unit 51, whether or not the nerve region is connected in the tumor region, that is, whether or not the nerve region is interrupted. Is determined based on the continuity of the pixel values. When it is determined by the inclusion relationship determination unit 51 that the nerve region has not entered the tumor region, the ratio setting unit 53 sets a default value to that effect as the nerve index value. When it is determined by the break determination unit 52 that the nerve region is connected in the tumor region, the ratio setting unit 53 sets a default value to that effect as the nerve index value. When the interruption determination unit 52 determines that the nerve region is interrupted in the tumor region, the ratio calculation unit 54 occupies the ratio of the nerve region that is interrupted in the tumor region in all the nerve regions that enter the tumor region. Calculate

  FIG. 7 is a diagram illustrating a typical flow of image processing according to the third embodiment, which is performed under the control of the control unit 9. For example, the control unit 9 receives an image processing start request from the user via the operation unit 8 and starts image processing according to the third embodiment.

  First, the control unit 9 causes the inclusion relationship determination unit 51 of the nerve index value determination unit 5 to perform a determination process (step SA1). In step SA1, the inclusion relationship determination unit 51 determines whether or not a nerve region has entered the tumor region based on the voxel values of voxels around the nerve region. First, the inclusion relationship determination unit 51 identifies a tumor region and a nerve region in the composite image. When the tumor region and the nerve region are specified, the inclusion relationship determination unit 51 determines whether the tumor region includes a nerve region based on the positional relationship between the tumor region and the nerve region. For example, when the vicinity of the nerve region is surrounded by voxels constituting the tumor region, the inclusion relationship determination unit 51 determines that the nerve region has entered the tumor region. On the other hand, when the vicinity of the nerve region is not surrounded by the voxels constituting the tumor region, the inclusion relationship determination unit 51 determines that the nerve region has not entered the tumor region.

  When it is determined in step SA1 that no nerve region has entered the tumor region (step SA1: NO), the control unit 9 causes the ratio setting unit 53 to perform setting processing (step SA2). In step SA2, the ratio setting unit 53 sets a predetermined value corresponding to the fact that the nerve region does not enter the tumor region as the nerve index value. The predetermined value is set to a value that does not overlap with a numerical range that the ratio calculated in step SA5 can have, for example, when the ratio is calculated as a percentage, greater than 0 and less than or equal to 100. The default value in step SA2 can be set to an arbitrary value within the above-described limit range by the user via the operation unit 8, for example. The default value in step SA2 may be set to −1, for example.

  When it is determined in step SA1 that the nerve region has entered the tumor region (step SA1: YES), the control unit 9 causes the interruption determination unit 52 of the nerve index value determination unit 5 to perform determination processing (step SA3). . In step SA3, the discontinuity determination unit 52 determines whether the nerve region is discontinuous in the tumor region based on the continuity of the pixel values. Specifically, the discontinuity determination unit 52 first tracks the nerve region in the composite image toward the tumor region, and specifies a portion where the nerve region is discontinuous (that is, the end point of the nerve region). Next, the discontinuity determination unit 52 determines whether or not the discontinuous portion is included in the tumor region. For example, the discontinuity determination unit 52 determines that the nerve region is discontinuous in the tumor region when the vicinity of the discontinuous portion is surrounded by the tumor region. On the other hand, the discontinuity determination unit 52 determines that the nerve region is not discontinuous in the tumor region when the vicinity of the discontinuous portion is not surrounded by the tumor region. This process is performed for all nerve regions in the tumor region until a discontinuous nerve region is identified.

  When it is determined in step SA3 that there is no interrupted nerve region (step SA3: NO), the control unit 9 causes the ratio setting unit 53 to perform setting processing (step SA4). In step SA4, the ratio setting unit 53 sets a predetermined value corresponding to the fact that a discontinuous nerve region is not included in the tumor region as the nerve index value. The predetermined value is set to a value that does not overlap with a numerical range that the ratio calculated in step SA5 may have, for example, a numerical range that is greater than 0 and less than or equal to 100. The default value set in step SA4 may be the same as or different from the default value set in step SA2. The default value in step SA4 can be set to an arbitrary value within the above-described limit range by the user via the operation unit 8, for example. The default value in step SA4 may be set to 0, for example.

  When it is determined in step SA3 that there is a discontinuous nerve region (step SA3: YES), the control unit 9 causes the ratio calculation unit 54 of the nerve index value determination unit 5 to perform calculation processing (step SA5). In step SA5, the ratio calculation unit 54 calculates the ratio of the nerve region that is interrupted in the tumor region to the total nerve region that has entered the tumor region. Specifically, the ratio calculation unit 54 first counts the number of all nerve regions entering the tumor region. Next, the ratio calculation unit 54 determines whether or not each nerve region is interrupted in the tumor region, and counts the number of nerve regions determined to be interrupted. Then, the ratio calculation unit 54 calculates the ratio of the number of nerve regions interrupted in the tumor region to the number of nerve regions entering the tumor region. The ratio may be simply calculated as a ratio or may be calculated as a percentage. This ratio is treated as a nerve index value.

  When step SA5 is performed, the control unit 9 causes the display image generation unit 6 to perform allocation processing (step SA6). In step SA6, the display image generating unit 6 assigns a color value corresponding to the default value set in steps SA2 and SA4 and the ratio calculated in step SA5 to the tumor region. For example, as the ratio increases, a dark red color value is assigned, and as the ratio decreases, a color value of another color such as dark blue is assigned. Typically, a single color value corresponding to the connection state of nerve regions in the tumor region is assigned to the tumor region.

  When step SA6 is performed, the control unit 9 causes the display image generation unit 6 to perform generation processing (step SA7). In step SA7, the display image generating unit 6 performs three-dimensional image processing on the three-dimensional composite image assigned with color values generated in step SA6 to generate a two-dimensional color image.

  When step SA7 is performed, the control unit 9 causes the display unit 7 to perform display processing (step SA8). In step SA8, the display unit 7 displays the two-dimensional color image generated in step SA7.

  FIG. 8 is a diagram showing the tumor region RT and the nerve region RN in the color image displayed by the display unit 7. As shown in FIG. 8, each pixel of the tumor region RT included in the color image is displayed in a color corresponding to the ratio of the interrupted nerve region in the entire nerve region entering the tumor region RN. Tumor areas with a high proportion correspond to tumors that contain many discontinuous nerves. Conversely, a tumor area with a small proportion corresponds to a tumor that does not contain much disrupted nerves. For example, when all the nerve regions entering the tumor region are not interrupted ((a) in FIG. 8), that is, when the ratio is 0%, the tumor region is displayed in dark blue. For example, when the ratio is 30% (FIG. 8B), the tumor area is displayed in light blue. For example, when the ratio is 60% (FIG. 8C), the tumor area is displayed in light red. When all nerve regions entering the tumor region are interrupted, that is, when the ratio is 100% ((d) in FIG. 8), the tumor region is displayed in dark red. Note that when it is determined in step SA1 that the nerve region does not enter the tumor region, this substantially corresponds to a rate of 0%. Therefore, when all the nerve areas that have entered the tumor area are not interrupted (ratio 0%) and the nerve area does not pass through the tumor area, for example, the tumor area is dark blue or other It is good to display with a color (for example, green). By observing the color of each pixel in such a tumor region, the user can easily grasp the running state of the nerve bundle in the tumor, specifically the proportion of the nerve region that is broken in the tumor. be able to. Thereby, the user can grasp the malignancy of the tumor.

  Note that it is not necessary to assign the same color value to all voxels constituting a single tumor region. FIG. 9 is another diagram showing the tumor region RT and the nerve region RN in the color image displayed by the display unit 7. As shown in FIG. 9, for example, the display image generation unit 6 may assign a color value to that effect to a region RD in the tumor region RT where the discontinuous portions of the nerve region RN are distributed more than others. good. Thereby, the display part 7 can display the area | region RD in the tumor area | region RT by the color different from another area | region. In other words, the display unit 7 can highlight the region RD with many discontinuities from the entire tumor region RT with a color. Therefore, the user can easily grasp the region RD having many discontinuous portions from the difference in color.

  Moreover, although it was calculated that the ratio of the discontinuous nerve area | region occupied in the all nerve area | regions which entered the tumor area | region was calculated, Example 3 is not limited to this. In this case, when it is determined in step SA3 that the nerve region is interrupted in the tumor region, the ratio setting unit may set a default value to that effect as the nerve index value. This default value is set to a color different from the default value in step SA4. Thereby, the user can discriminate | determine by a color whether the tumor is malignant by identifying the color of a tumor area | region.

  The displayed image is not limited to a color image or a multiple color image. For example, the display unit 7 may display a color image or a multiple color image superimposed on an MPR image or volume rendering image of a three-dimensional brain image.

[Modification]
If the tumor is malignant, it is necessary to remove the entire tumor. In this case, there is a possibility that a disorder corresponding to the nerve bundle removed together may occur in the subject. Accordingly, it is clinically very important to present which region of the brain the nerve bundle connected to the malignant tumor is connected to, in other words, the range of influence of the nerve bundle on the brain.

  The image processing apparatus 1 according to the modification emphasizes the arrival point in the brain of the nerve connected to the tumor when displaying the multiple color image. The image processing apparatus 1 according to the modification will be described below. In the following description, components having substantially the same functions as those of the third embodiment are denoted by the same reference numerals, and redundant description is provided only when necessary.

  FIG. 10 is a diagram illustrating a configuration of the nerve index value determining unit 5 according to the modification. As illustrated in FIG. 5, the nerve index value determination unit 5 includes an inclusion relationship determination unit 51, a break determination unit 52, a ratio setting unit 53, a ratio calculation unit 54, and a reaching point specifying unit 55.

  When the inclusion relationship determination unit 51 determines that the nerve region has entered the tumor region, the arrival point specifying unit 55 specifies the arrival point on the brain region of the nerve region that has entered the tumor region in the multiple composite image. To do.

  FIG. 11 is a diagram illustrating a typical flow of image processing according to the modification performed under the control of the control unit 9 according to the modification. For example, the control unit 9 receives an image processing start request from the user via the operation unit 8 and starts image processing according to the modification.

  First, the control unit 9 causes the inclusion relationship determination unit 51 of the nerve index value determination unit 5 to perform a determination process (step SB1). In step SB1, the inclusion relationship determination unit 51 determines whether a nerve region has entered a tumor region included in the composite image or the multiple composite image. Since the process of step SB1 is the same as that of step SA1, description here is abbreviate | omitted.

  When it is determined in step SB1 that the nerve region has not entered the tumor region (step SB1: NO), the control unit 9 ends the image processing according to the modification. In this case, for example, the control unit 9 may perform processing in the order of image processing steps SA2 → SA6 → SA7 → SA8 according to the third embodiment.

  When it is determined in step SB1 that the nerve region has entered the tumor region (step SB1: YES), the control unit 9 causes the arrival point specifying unit 55 to perform a specifying process (step SB2). In step SB2, the reaching point specifying unit 55 specifies a reaching point on the surface of the brain region of the nerve region in the multiple composite image. Hereinafter, the process of step SB2 will be described with reference to FIG. FIG. 12 is a diagram illustrating a tumor region RT, a nerve region RN, and a brain region surface SB in a multiple composite image. First, the reaching point specifying unit 55 specifies the nerve region RN and the brain region surface SB based on the voxel values in the multiple synthesized image. Next, the reaching point specifying unit 55 tracks the nerve region RN toward the brain region surface SB, and specifies the intersection PR between the nerve region RN and the brain region surface SB. This intersection PR is set as the arrival point PR. More specifically, the reaching point PR is set to a voxel constituting the adjacent nerve region RN and a voxel constituting the brain region surface SB. The coordinate data of the reaching point PR in the multiple composite image is supplied to the display image generating unit 6.

  When step SB2 is performed, the display image generating unit 6 generates a two-dimensional multiple color image in which the arrival point specified in step SB2 is emphasized with a color based on the multiple composite image. Specifically, the display image generation unit 6 first assigns a specific color value indicating the arrival point to the voxel set as the arrival point. The specific color value is set to a value that can distinguish the arrival point from the surface of the brain region by color. Note that the voxel to which the color value is assigned is not strictly limited to the intersection between the nerve region and the brain region surface. For example, a color value may be assigned to a region having a predetermined size including the intersection. Thereby, the visibility of the arrival point can be improved even when the intersection is small. When the color value is assigned, the display image generating unit 6 performs three-dimensional image processing on the multiple composite image to which the color value is assigned, and generates a two-dimensional multiple color image in which the arrival point is emphasized by color.

  When step SB3 is performed, the control unit 9 causes the display unit 7 to perform display processing (step SB4). In step SB4, the display unit 7 displays the multiple color image generated in step S3.

  FIG. 13 is a diagram illustrating an example of the multiple color image IMC displayed by the display unit 7 in step SB4. As shown in FIG. 13, when displaying the multi-color image IMC, the display unit 7 displays a region RR having a predetermined size including the arrival point on the brain region surface SB of the nerve region RN entering the tumor region RT. Emphasize with color. Note that only the reaching point may be emphasized by color. By observing this multiple color image, the user can easily grasp the arrival point on the surface of the brain region of the nerve region. In addition, a user with abundant medical knowledge can predict a disorder that may occur in a subject by grasping the position of the arrival point on the surface of the brain region and extracting a nerve bundle connected to a malignant tumor. .

  When displaying multiple color images, the display unit 7 may display images that simulate the brain (hereinafter referred to as simulated brain images) side by side. Hereinafter, the display processing of the simulated brain image will be described.

  For example, the storage unit 2 stores a plurality of simulated brain image data respectively corresponding to a plurality of cross-sectional positions or viewpoint directions. As the simulated brain image, for example, a Broadman brain map is adopted. The Broadman brain map is a simulated brain image showing a standard brain form, as shown in FIG. Broadman's brain map is assigned brain function identification numbers for each area according to brain function, tissue structure, and the like.

  When displaying the multi-color image in which the arrival point is emphasized, the display unit 7 stores the broadman's brain map in the cross-sectional position or the viewpoint direction corresponding to the anatomical position of the nerve region entering the tumor region. Are displayed side by side in a multiple color image in which the arrival points are emphasized. For example, the cross-sectional position corresponding to the position of the reaching point specified by the reaching point specifying unit 55 or the Broadman brain map in the viewpoint direction is read by the display unit 7. In this way, the Broadman brain map and the multiple color image with the arrival point emphasized are displayed side-by-side, so that the user can estimate the position of the arrival point on the Broadman brain map and thereby become a malignant tumor. It is possible to easily know the range of influence of the connected nerves, and thus it is possible to predict a disorder that can occur in the subject by removing all the nerve bundles connected to the malignant tumor.

  When displaying the Broadman brain map, the display unit 7 may emphasize the arrival point of the nerve region in the Broadman brain map in order to clarify the position of the reaching point in the Broadman brain map. FIG. 15 is a diagram illustrating a display example of Broadman's brain map in which the arrival points of the nerve region are emphasized. As shown in FIG. 15, the Broadman brain map may be displayed in the application window WA. In the application window WA, the Broadman brain map ITB related to the cross-sectional position or viewpoint direction corresponding to the arrival point and the Broadman brain map IOB related to other cross-sectional positions or viewpoint directions are displayed. In FIG. 15, Broadman's brain map ITB relates to the inner surface of the brain, and Broadman's brain map IOB relates to the outer surface of the brain. On the Broadman's brain map ITB, a region RS having a predetermined size and shape including a reaching point of a nerve region that has entered the tumor region is displayed. The area RS is set as follows, for example.

  For example, the display image generation unit 6 performs a fitting process on the Broadman brain map ITB and the multiple color image in which the arrival points are specified, thereby calculating the arrival point of the nerve region in the Broadman brain map ITB. Next, the display unit 7 sets a region RS having a predetermined size including the calculated arrival point in the Broadman brain map ITB. The display unit 7 highlights the set region RS with a color or the like when displaying the Broadman brain map ITB. The size and shape of the region RS and the display color of the region RS can be arbitrarily set by the user via the operation unit 8. The display unit 7 may highlight only the calculated arrival point on the Broadman brain map with a color or the like when displaying the Broadman brain map ITB.

  In the above description, when the reaching point is emphasized, an area having a predetermined size and shape including the reaching point is also emphasized. In order to more clearly show the influence range of the nerve connected to the malignant tumor, the display unit 7 displays the influence range of the nerve corresponding to the nerve area entering the tumor area in the Broadman brain map or the multiple color image. You may emphasize it. A specific example of this process will be described below.

  The range of influence of nerves is determined medically according to the type of nerve. The type of nerve can be specified, for example, according to the anatomical position of the nerve. The storage unit 2 stores a simulated brain image (for example, Broadman's brain map) in which one of a plurality of sections is emphasized by color. A simulated image in which one section is emphasized is prepared for all sections. A set of simulated images in which one section is emphasized is prepared for a plurality of cross sections and line-of-sight directions. Each simulated brain image is managed by an identification number. For example, the display unit 7 includes a table in which the identification number of the simulated brain image is associated with the type of nerve bundle that affects the range of influence emphasized in the simulated brain image.

  When the reaching point is specified by the reaching point specifying unit 55, the display unit 7 specifies the type of nerve bundle corresponding to the nerve region based on the anatomical position of the nerve region reaching the reaching point. . Next, the display unit 7 identifies the identification number of the simulated brain image from the type of nerve bundle identified using the above table, and stores the simulated brain image associated with the identified identification number. Read from 2. The display unit 7 displays the read simulated image. In the displayed simulated brain image, the influence range of the nerve bundle reaching the reaching point is highlighted in color. Therefore, the display unit 7 can more clearly show the influence range of nerves connected to the malignant tumor.

  In addition, according to the present embodiment, it is possible to know the correspondence between the position of the tumor and the influence range of the nerve connected to the tumor. For example, the storage unit 2 stores, in association with each position of the tumor region to be processed, simulated image data in which the influence range corresponding to the nerve region connected to the tumor region is emphasized by color or the like. Thereby, the memory | storage part 2 can accumulate | store the correspondence of the position of the tumor, and the influence range of the nerve connected to the tumor as a database. The accumulated data can be read by the user via the operation unit 8 and displayed on the display unit 7.

  This completes the description of the first embodiment, the second embodiment, the third embodiment, and the modification.

  According to the embodiment, the image processing apparatus 1 according to the present embodiment includes at least the composite image generation unit 4, the nerve index value determination unit 5, the display image generation unit 6, and the display unit 7. The composite image generation unit 4 generates a composite image of the nerve image and the tumor image. The nerve index value determination unit 5 determines a nerve index value that represents the anatomical running state of the nerve region derived from the nerve image with respect to the tumor region derived from the tumor image. The display image generation unit 6 generates a color image in which a color value corresponding to the nerve index value is assigned to the tumor region based on the synthesized image. The display unit 7 displays a color image. Therefore, the user can grasp the anatomical running state of the nerve bundle with respect to the tumor by the color difference by observing the color image. More specifically, it is possible to grasp the positional relationship between the tumor and the nerve bundle, the state of the nerve inside the tumor, the malignancy of the tumor, and the risk in treating the tumor. As described above, the image processing apparatus 1 can provide a lot of useful information such as planning of a treatment policy, determination of whether or not treatment is possible, and informed consent. Therefore, the image diagnosis capability is improved by the image processing apparatus 1.

  Thus, according to the present embodiment, improvement in image diagnostic ability is realized.

  In the above description, the information corresponding to the nerve index value is a color value. However, the present embodiment is not limited to this. For example, the information corresponding to the nerve index value may be a gray value. In this case, for example, the display image generation unit 6 generates a two-dimensional gray image in which a gray value corresponding to the nerve index value is assigned to a tumor region included in the composite image based on the composite image. At this time, the display image generation unit 6 may assign a specific gray value to the nerve region included in the composite image. Next, the display image generation unit 6 performs three-dimensional image processing on the synthesized image to which the gray value is assigned, and generates a two-dimensional gray image. Examples of 3D image processing include volume rendering, surface rendering, pixel value projection processing, MPR (multi-planar reconstruction), and the like. As the pixel value projection process, typically, a maximum intensity projection (MIP) method may be employed. The gray value assignment may be performed after the three-dimensional image processing instead of the three-dimensional image processing. Further, the display image generation unit 6 may generate a two-dimensional multiple gray image in which a gray value corresponding to a nerve index value is assigned to a tumor region based on a multiple composite image by the same three-dimensional image processing. it can. The display unit 7 displays various display images such as the generated gray image and multiple gray image on the display device.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 1 ... Image processing apparatus, 2 ... Memory | storage part, 3 ... Extraction part, 4 ... Composite image generation part, 5 ... Neural index value determination part, 6 ... Display image generation part, 7 ... Display part, 8 ... Operation part, 9 ... Control unit

Claims (10)

A storage unit for storing a three-dimensional neural image relating to the neural tissue of the brain and a three-dimensional tumor image relating to the brain tumor;
A composite image generator for generating a composite image of the neural image and the tumor image;
A determination unit for determining an index value related to an anatomical running state of a nerve region derived from the nerve image with respect to a tumor region derived from the tumor image in the composite image;
An information image generating unit that generates, based on the composite image, a two-dimensional information image in which information corresponding to the determined index value is assigned to the tumor region included in the composite image;
A display unit for displaying the generated information image;
Comprising
The determining unit, as the index value, images processing device distance you calculate from each of a plurality of pixels included in the tumor area to said neural area. A storage unit for storing a three-dimensional neural image relating to the neural tissue of the brain and a three-dimensional tumor image relating to the brain tumor;
A composite image generator for generating a composite image of the neural image and the tumor image;
A determination unit for determining an index value related to an anatomical running state of a nerve region derived from the nerve image with respect to a tumor region derived from the tumor image in the composite image;
An information image generating unit that generates, based on the composite image, a two-dimensional information image in which information corresponding to the determined index value is assigned to the tumor region included in the composite image;
A display unit for displaying the generated information image;
Comprising
The determining unit, as the index value, the each of the plurality of local regions of the tumor area, the images processing it counts the number or the number of voxels constituting the nerve region of the nerve region of the local region apparatus. A storage unit for storing a three-dimensional neural image relating to the neural tissue of the brain and a three-dimensional tumor image relating to the brain tumor;
A composite image generator for generating a composite image of the neural image and the tumor image;
A determination unit for determining an index value related to an anatomical running state of a nerve region derived from the nerve image with respect to a tumor region derived from the tumor image in the composite image;
An information image generating unit that generates, based on the composite image, a two-dimensional information image in which information corresponding to the determined index value is assigned to the tumor region included in the composite image;
A display unit for displaying the generated information image;
Comprising
The determination unit
A determination unit that determines whether or not the nerve region is connected in the tumor region based on continuity of pixel values;
When it is determined that the nerve region is connected, the determination unit sets a first predetermined value indicating that the nerve region is connected to the index value, and determines that the nerve region is not connected If it is, the setting unit for setting a second predetermined value indicating that the neural region is not connected to the index value, images processing device Ru comprising a. The image generation unit generates, as the information image, a two-dimensional color image in which a color value corresponding to the determined index value is assigned to the tumor region included in the composite image based on the composite image. The image processing apparatus according to any one of claims 1 to 3 . The storage unit further stores a three-dimensional brain image representing the morphology of the brain,
The synthesis unit generates a multiple synthesized image of the brain image, the tumor image, and the nerve image,
The information image generation unit generates a two-dimensional multiple information image in which information corresponding to the index value is allocated to the tumor region included in the multiple composite image,
The display unit displays the generated multiple information image;
The image processing apparatus according to claim 1. The determination unit determines whether or not the tumor region and the nerve region are in contact with each other, and if it is determined that the tumor region and the nerve region are in contact with each other, A specifying unit that specifies a reaching point of the nerve region of the brain region surface included, and
The display unit emphasizes the identified arrival point when displaying the multiple information image;
The image processing apparatus according to claim 5 . The display unit, when displaying the multiple information image, medically determined according to the type of nerve corresponding to the nerve region, the influence range of the nerve in the surface of the brain region including the reaching point The image processing apparatus according to claim 6, wherein the image processing apparatus is emphasized. The image processing apparatus according to claim 6 , wherein the display unit displays a Broadman brain map corresponding to an anatomical position of the nerve region in the brain region. The display unit, when displaying the Broadman brain map, is determined medically according to the type of nerve corresponding to the nerve region, and the nerve on the Broadman brain map including the reaching point The image processing apparatus according to claim 8 , wherein an influence range of the image is emphasized.   The image processing apparatus according to claim 1, wherein the neural image is generated by applying a fiber tracking technique to a diffusion weighted image generated by imaging the brain with a magnetic resonance imaging apparatus.

Images