JP2012085833A - Image processing system for three-dimensional medical image data, image processing method for the same, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology for easily designating a three-dimensional interest region in medical image volume data on a two-dimensional image.SOLUTION: Designation of the three-dimensional interest region by a user is supported in an image display of three-dimensional medical image data in a three-dimensional medical image processing system. A candidate object to be displayed in the three-dimensional medical image data is selected according to user input, and visualization processing of the three-dimensional medical image data is performed to display an image that specifies the selected candidate object. An object for designating the three-dimensional interest region is selected according to user selection input to the displayed candidate object image. The user inputs interest region designation to the selected object.

Description

  The present invention relates to an image processing system for 3D medical image data, an image processing method thereof, and a program, and more particularly to a technique for supporting designation of a 3D region of interest on a 2D image generated from medical image volume data.

  In a diagnosis using a medical image inspection apparatus typified by an X-ray CT (X-ray Computed Tomography) apparatus or an MRI (Magnetic Resonance Imaging) apparatus, a doctor uses three-dimensional medical image data (hereinafter referred to as “medical use”). In some cases, a visualized image obtained from “image volume data” or “volume data” is used as a diagnostic aid.

  The visualized image is obtained by performing visualization processing on the medical image volume data using a plurality of medical image processing algorithms. As a medical image processing algorithm used here, there are a plurality of methods according to the purpose of diagnosis, such as a method of extracting and displaying only a necessary tissue from input volume data and a method of highlighting and displaying an edge.

  In recent years, the number of tomographic images (hereinafter also referred to as “slice images”) that can be photographed at once by the medical image capturing apparatus is increasing at an accelerating rate due to technological advancement. The amount of is also a huge size.

  Therefore, in diagnosis using a medical image, in addition to interpretation for each slice image, a reconstructed two-dimensional image is read using a three-dimensional reconstruction method such as Volume Rendering (VR) or Maximum Intensity Projection (MIP). It has become mainstream. However, in order to specify the 3D position of an object in a 2D image of the object having 3D information, it is necessary to handle information in the depth direction on the 2D plane in addition to the position information in the 2D plane. is there.

  The interpreting doctor reports the interpretation result to the attending doctor as a report, but it is difficult to express the three-dimensional position on the two-dimensional image, and this is the reason why the information exchange between the interpreting doctor and the attending doctor is smooth. Often not.

  In the past, electronic medical records (reports) and 3D medical image data were handled separately, but in recent years, electronic medical records and a medical image server PACS (Picture Archiving and Communication System) are used as one system. The number of cases is increasing, and a method for easily selecting a three-dimensional position on a two-dimensional image is required. In response to these requirements, the following techniques have been proposed as prior art.

  Patent Document 1 is a technique for obtaining coordinates on a specified one slice image on the image and a slice image number to which the coordinates belong when a medical slice image is stacked and an image of a three-dimensional object is displayed. From the coordinates of a point on the image of the three-dimensional object displayed at an arbitrary viewpoint and the coordinates of an arbitrary point on the trajectory of the point obtained by rotation, the slice image number and coordinates are processed by the computer. Is a technique for calculating

  Patent Document 2 is an image processing method for determining the three-dimensional coordinates of a point based on a plurality of two-dimensional windows. In Patent Literature 3, when a two-dimensional image obtained by projecting three-dimensional data onto a plane is displayed, a three-dimensional positional relationship between the attention point in the two-dimensional image and the surrounding region of interest is arbitrarily determined. This technology aims to enable display from the direction.

JP-A-6-274599 JP-A-8-212390 JP-A-9-81786

  As described above, as a method for designating a point or region represented by three-dimensional coordinates on a two-dimensional plane for a three-dimensional image projected on a two-dimensional plane, a plurality of projections projected from a plurality of viewpoints A method for designating an image using an image has been proposed. However, a living body to be medically diagnosed has a complicated structure composed of many tissues and organs. In medical diagnosis, a three-dimensional region of interest is specified more quickly and easily using a two-dimensional image. There is a need for technology that can do this.

  One aspect of the present invention is an image processing system for three-dimensional medical image data including an input acquisition unit that acquires user input, a data storage device, an image processing unit, and a display device. The data storage device stores the three-dimensional medical image data. The image processing unit visualizes the three-dimensional medical image data to generate one or a plurality of candidate object image data that clearly indicate one or a plurality of candidate objects, and the one or a plurality of candidates Object image data is stored in the data storage device. The display device displays an image of the one or more candidate object image data stored in the data storage device. The image processing unit is configured to display the one or the plurality of candidate object image data according to a user selection input for the displayed image. An object for designating a three-dimensional region of interest is selected from a plurality of candidate objects. The input acquisition unit receives a designation input of the three-dimensional region of interest from the user for the display image of the selected object.

  According to the present invention, a three-dimensional region of interest in medical image volume data can be easily designated on a two-dimensional image.

1 is a system configuration diagram schematically illustrating a configuration of a medical image processing system according to an embodiment. In this embodiment, it is a flowchart which shows an example of the three-dimensional region-of-interest designation | designated process performed with an image processing apparatus. It is a block diagram which shows typically a partial structure of the medical image processing system which concerns on this embodiment. In this embodiment, it is a figure explaining the two-dimensional designation | designated position and determination area | region used for a three-dimensional region-of-interest designation | designated process. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. It is a figure which shows an example of the screen display in the step which designates an object in this embodiment. In this embodiment, it is a figure which shows the other example of the screen display in the step which designates an object. In this embodiment, it is a figure explaining the method of extracting an object using the change of the brightness | luminance of a voxel. In this embodiment, it is a figure explaining the method of estimating an object using distribution of the brightness | luminance of a voxel. It is an example of the object brightness | luminance correspondence table | surface which shows the response | compatibility of a brightness | luminance and an object which the object estimation process using the brightness distribution of the voxel of this embodiment refers.

  Embodiments of the present invention will be described below in detail with reference to the drawings. For clarity of explanation, the following description and drawings are omitted and simplified as appropriate. Moreover, in each drawing, the same code | symbol is attached | subjected to the same element and the duplication description is abbreviate | omitted as needed for clarification of description.

  The medical image processing system of this embodiment is a user (mainly a doctor) three-dimensional region of interest (hereinafter also referred to as a region of interest) for three-dimensional medical image data (hereinafter also referred to as medical image volume data or volume data). Support the designation of The region of interest is a region where the user desires to display a detailed image on the object represented by the volume data.

  The medical image processing system generates two-dimensional images of one or a plurality of objects from volume data in response to user input and displays them. An object represents a unit in a living body, such as a tissue, organ, or area in an organ, in a medical image processing system. An object in medical image volume data is a group of voxels handled in the image processing system, and preferably matches an actual tissue or organ in a living body, but may not match.

  In the present embodiment, the user selects at least one object image from one or more displayed object images. The user can specify a region of interest by input to the image of the selected object. As described above, the user can designate the region of interest simply and quickly by automatically displaying the image of the object according to the user input for the user to designate the region of interest.

  FIG. 1 is a block diagram schematically showing the configuration of the medical image processing system of the present embodiment. As shown in FIG. 1, the medical image processing apparatus system 1 includes medical image volume data and medical image imaging such as an external storage device 10 including a nonvolatile storage medium that stores data related thereto, an X-ray CT apparatus, an MRI apparatus, and the like. An image processing device 11 that performs image processing on the volume data captured by the device, a display device 12 that displays the image processed image, an input device 13 that inputs instructions for starting each process and operations on the screen Is provided.

  Typical examples of the input device 13 are a keyboard and a pointer device. The image processing apparatus 11 may be connected to the display device 12 and the input device 13 via a network and an input / output computer connected to the network. The user accesses the image processing apparatus 11 via the input / output computer. The external storage device 10 can be connected to the image processing device 11 directly or via a network. Typically, the external storage device 10 is a hard disk drive including a magnetic disk or a flash memory device including a semiconductor nonvolatile storage medium.

  The image processing apparatus 11 includes a CPU (Central Processing Unit) 14 that is a processor that executes data arithmetic processing, and an internal memory 15 that stores data (including a program to be executed) processed by the CPU 14. The CPU 14 functions as an image processing algorithm execution unit 141, an input operation acquisition unit 142, and a designated area calculation unit 143 by operating according to a program. Further, in the operation as the image processing algorithm execution unit 141, the CPU 14 functions as a visualization process execution unit 144 and an object process execution unit 145. At least part of the processing of the CPU 14 may be performed by a GPU (Graphics Processing Unit). In particular, faster processing can be realized by the GPU performing part of the processing of the image processing algorithm execution unit 141.

  In the image processing algorithm execution unit 141, the visualization process execution unit 144 executes a visualization process on the volume data and displays it on the display device. The object processing execution unit 145 performs processing on an object and determination related thereto. The input operation acquisition unit 142 acquires an input from the input device 13. The designated area calculation unit 143 calculates an area designated by the user in the volume data based on the user input obtained from the input operation acquisition unit 142. Details of these functional units will be described later.

  The CPU 14 operates according to a command selected from a program stored in each storage area in the internal memory 15. Therefore, the description using the function unit realized by the operation of the CPU 14 as the subject may be the description using the program as the subject. Part or all of the program may be realized by dedicated hardware. The program can be installed in the image processing system 1 by a program distribution server or a computer-readable medium, and can be stored in a storage area of the external storage device 10.

  The internal memory 20, which is an example of a data storage device, is typically a RAM (Random Access Memory) and includes a main memory and a cache memory. The internal memory 20 stores volume data and related information (data) in addition to programs. Specifically, the data acquired by the input operation acquisition unit 142 and the data generated by the image processing algorithm execution unit 141 and the designated area calculation unit 143 are stored in those storage areas. Volume data is loaded from the external storage device 10 into the internal memory 15. Data transfer between programs is performed via the internal memory 15.

  Hereinafter, a process for designating a three-dimensional region of interest in volume data in the medical image processing system 1 of the present embodiment will be described. The medical image processing system 1 displays a two-dimensional visualization image of volume data, and displays an image of one or a plurality of objects according to a user input for the display.

  When the object is defined in the volume data, the medical image processing system 1 displays an image of the defined object. When the object is not defined in the volume data, the medical image processing system 1 estimates the object from the volume data and displays an image of the object.

  The medical image processing system 1 displays images of a plurality of objects sequentially or collectively. The user selects one object image and designates a region of interest in the image of the selected object. The medical image processing system 1 displays the designated region of interest.

  The overall flow of processing for designating a three-dimensional region of interest on a two-dimensional image will be described with reference to the flowchart of FIG. In the description of this process, FIG. 3 is referred to as appropriate. FIG. 3 shows a partial configuration of the image processing system 1 shown in FIG. 1, and schematically shows a data flow in this processing in the partial configuration.

  As shown in FIG. 2, the image processing apparatus 11 starts the three-dimensional region designation process in response to an input from the input device 13 or an instruction from the external system (the input device 13 via) (S10). As shown in FIG. 3, the volume data 151 is transferred from the external storage device 10 to the internal memory 15.

  Next, the visualization process execution unit 144 (image processing algorithm execution unit 141) acquires the volume data 151 stored in the internal memory 15, executes the visualization process on the volume data, and performs the initial visualization image. Data 152 is generated (S11). As shown in FIG. 3, the generated initial visualized image data 152 is stored in the internal memory 15. Thereafter, the initial visualized image data 152 is transferred to the display device 12, and the display device 12 displays the initial visualized image (S12).

  The visualization process performed by the visualization process execution unit 144 in step 12 is a process of generating a two-dimensional visualization image (data) from the volume data 151. For this visualization processing, for example, a three-dimensional visualization method and a cross-section generation method can be used.

  An example of a three-dimensional visualization method is volume rendering in which transparency is set from the voxel values, the light of the voxels on each line of sight is added and the light of the voxels is added, and the object is displayed in three dimensions (Volume Rendering: VR), MIP (Maximum Intensity Projection) that projects the maximum voxel value of voxels on each line of sight.

  Examples of the cross-section generation method include Multi Planar Reformat (MPR) for displaying a cross section of a specified cross section such as Axial, Sagittal, Coronal, etc. and a cross section of an arbitrary cross section. ) Etc.

  The visualization processing execution unit 144 displays an initial visualization image using the display device 12. The target range of the visualization process (S11) for generating the initial visualized image data 152 and the visualization method to be used are set in advance or can be specified by the user via the input device 13. Also good.

  The user designates a position in the two-dimensional initial visualized image displayed on the display device 12 by the input device 13. The input operation acquisition unit 142 receives the designated two-dimensional position designated for the initial visualized image displayed on the display device 12 from the input device 13. For example, when the user uses a mouse as the input device 13, two-dimensional coordinates designated on the display screen by designation with a cursor or clicking are the designated two-dimensional positions. This two-dimensional position is a coordinate in the visualized image.

  The input operation acquisition unit 142 acquires the two-dimensional coordinates of the designated two-dimensional position and the position of the currently displayed initial visualized image as one-dimensional coordinates, and transfers the information to the designated region calculation unit 143. The position of the initial visualized image is the slice position when the initial visualized image is a cross-sectional image, and the position of the projection plane when the initial visualized image is a projected image such as a VR image or MIP image. When the one-dimensional coordinates of the initial visualized image are specified values, only the two-dimensional coordinates of the designated two-dimensional position need be sent to the designated area calculation unit 143.

  This one-dimensional coordinate is, for example, the coordinate of the center position of the initial visualized image on a specific coordinate axis in the volume data 151. As described above, the one-dimensional coordinates (visualized image position) are set in advance or determined by the user's designation. The one-dimensional coordinates may be determined based on any point on the initial visualization image, or may be determined based on the designated coordinates.

  The designated area calculation unit 143 calculates three-dimensional coordinates in the volume data 151 from the two-dimensional coordinates of the designated two-dimensional position and the one-dimensional coordinates of the initial visualized image (S13). Furthermore, the designated area calculation unit 143 transfers the calculated three-dimensional coordinates as the designated position (initial position) to the object processing execution unit 145 of the image processing algorithm execution unit 141.

  In the example in which the visualized image is a planar cross-sectional image or projection image, the designated area calculation unit 143 includes the coordinates of the specific coordinates at the center position of the planar image, the normal direction vector in the volume data of the planar image, and the designated two-dimensional position. The coordinates in the volume data can be calculated from the two-dimensional coordinates. When the image is a part of a spherical surface, for example, the coordinates can be calculated using the coordinates in the volume data of the curvature center of the image.

  The designated area calculation unit 143 determines a determination area from the acquired initial position, and stores data 153 specifying the determination area in the internal memory 15 as illustrated in FIG. 3. Further, as shown in FIG. 2, the object processing execution unit 145 determines whether or not an object defined in the volume data 151 exists in the area indicated by the determination area specifying data 153 (S14). ). When at least a part of the object is in the determination area, the object processing execution unit 145 determines that the object exists in the determination area.

  As described above, an object is a unit of a living body that is an object, and is an organ or tissue such as skin, bone, blood vessel, stomach, liver, brain, or the like. The determination area is a line including the initial position, a surface having a predetermined outer shape, or a three-dimensional area having a predetermined outer shape. Details of the determination area and the method for determining the presence / absence of a defined object will be described later.

  When there is no object defined in the volume data 151 in the determination area, the object process execution unit 145 executes an object estimation process (S15). In the object estimation process, an object in the determination area in the volume data 151 is estimated. Details of the object estimation process will be described later with reference to FIGS.

  As illustrated in FIG. 3, the object processing execution unit 145 stores data 154 for specifying an object (candidate object) in the determination area in the internal memory 15. Although not shown in the flowchart of FIG. 2, if neither the object defined in the volume data 151 nor the estimated object exists in the determination area, the user selects the area of interest without selecting an object. Specify (S18).

  Next, as illustrated in FIG. 2, the visualization processing execution unit 144 displays a candidate object image (candidate object image) (S <b> 16). Specifically, as illustrated in FIG. 3, the visualization processing execution unit 144 generates candidate object image data 155 from the volume data 151 and the candidate object specifying data 154 and stores it in the internal memory 15. The candidate object image data 155 is sent to the display device 12.

  The display device 12 displays an image of the object represented by the object image data. When there are a plurality of candidate objects, the candidate object images are sequentially or collectively displayed. Details of the candidate object image display method will be described later with reference to FIGS. 5A to 5F and 6.

  The user visually recognizes the object displayed on the image display device 12, determines whether or not to use the object for designating the region of interest, and inputs the determination through the input device 13 (S17). If the user determines not to use all the displayed images of candidate objects (N in S17), the process proceeds to a region of interest designation step (S19).

  The user decides (inputs) that the displayed object image is to be used for designating the region of interest (Y in S17), and selects the object image to be used by the input device 13 (S18). Typically, selection of a displayed candidate object image by the user is an input for determining the use of the object image.

  When the display device 12 displays the target object as a candidate object, the user inputs information to be determined by the input device 13. The input operation acquisition unit 142 transfers information that the candidate object has been selected as the selection object to the designated area calculation unit 143. The designated area calculation unit 143 defines the candidate object as a selected object.

  Thereafter, the process proceeds to a region of interest designation step (S20). The visualization process execution unit 144 displays an image of the selected object on the display device 12. In a configuration in which images of a plurality of candidate objects are displayed on the screen, typically, an enlarged image of one selected object is newly displayed. In the configuration in which the images of the candidate objects are sequentially displayed, the selected image is used as it is or a newly selected image is redisplayed.

  The user performs input for designating a region of interest in the displayed image of the object. Details of the method of specifying the region of interest in the object image will be described later. The visualization processing execution unit 144 displays an image of the region of interest designated by the user using the input device 13 on the display device 12.

  Specifically, the designated region calculation unit 143 converts the input data acquired from the input operation acquisition unit 142 and the coordinate data of the target object acquired from the object processing execution unit 145 into volume data 151 indicating the specified region of interest. Calculate the coordinate data. The visualization process execution unit 144 acquires the calculated coordinate data from the designated area calculation unit 143.

  The visualization processing execution unit 144 generates image data 156 of the designated region of interest from the coordinate data acquired from the designated region calculation unit 143 and the volume data 151, and stores it in the internal memory 15. The region-of-interest image data 156 is transferred to the display device 12, and the display device 12 displays the image represented by the region-of-interest image data.

  If the region of interest desired by the user is displayed (Y in S20), the region of interest designating process ends (S21). When the desired region of interest is not displayed (N in S20), the user performs an input indicating that the region of interest designation process is not completed by the input device 13. The process re-executes the initial position specifying step (S13) and the subsequent steps. Depending on the design, the process may start from a visualization process step (S11) or a visualization display step (S12).

  Further, when a plurality of regions of interest are designated for one volume data, the process is once ended (S21) and then started from the initial position specifying step (S13). Similarly to the above, the processing may be started from the visualization processing step (S11) or the visualization display step (S12). In this case, for example, the visualization processing execution unit 144 performs coloring on the region of interest that has already been specified so that the user can visually recognize that the region has already been specified as the region of interest. It is preferable to do.

  Also, in a preferred configuration, whether a voxel included in a region that has already been specified as a region of interest can be included in a newly specified region of interest, or whether one voxel belongs to only one region of interest Can be set. The designated region calculation unit 143 may be configured to perform different processing on voxels that already belong to any region of interest and voxels that do not.

<Judgment area>
Details of some steps in the flowchart shown in FIG. 2 will be described below. First, the determination step of Step 14 will be specifically described. Step 14 identifies an object that overlaps (at least a part of) the determination area (at least a part of). The object to be determined is an object defined in the volume data 151.

  A preferable example of the determination area will be described with reference to FIG. The left diagram in FIG. 4 is a diagram schematically showing an image in which the volume data of the head is visualized and displayed by VR, and is an initial visualized image in which the user designates a two-dimensional position. The projection plane of the VR image is located in front of the head. In this example, skin, skull, and brain objects are shown from the outside of the head.

In the left view of FIG. 4, the X axis extends from left to right and the Y axis extends from top to bottom as viewed in the figure. Further, the Z-axis extends from the front of the figure toward the back. The direction of the Z axis coincides with the normal direction of the initial visualized image (projection plane thereof). The two-dimensional position 21 designated by the user is indicated by x. Furthermore, the determination area 23 is an area surrounded by a circle and includes the designated position 21.

  The right figure of FIG. 4 is a diagram schematically showing an image obtained by visualizing and displaying the volume data of the same head on the projection plane with the X-axis direction as a normal line. The Z axis extends from left to right as viewed in the figure, and the Y axis extends from top to bottom. Further, the X axis extends from the back of the figure toward the front. In this example, the Z coordinate of the designated position 21 is the frontmost position (Z = 0) in the volume data 151. A straight line extending from the designated position 21 along the Z axis is a determination straight line 22. As can be understood from FIG. 4, the determination region 23 is a region in a cylinder (columnar region) that includes the determination straight line 22 and extends along the Z axis.

  The object processing execution unit 145 specifies an object whose determination area 23 includes at least a part thereof. The object is defined in the volume data 151. An object is composed of one or a plurality of voxels in volume data, and the voxels in the object are defined to belong to the object.

  One technique for defining objects in the volume data 151 uses a mask. The mask is data that covers an area associated with the mask, identifies the area, and distinguishes it from other areas. The mask may be stored as DICOM (Digital Imaging and Communication in Medicine) tag information, and the mask information may be stored in a format different from the volume data.

  In the preferred example shown in FIG. 4, the determination area 23 is a cylindrical solid area centered on a determination straight line 22 extending from the designated position 21 along the Z axis. The object processing execution unit 145 determines whether each voxel included in the determination area 23 is defined as belonging to any object. When it is defined that the voxel in the determination area 23 belongs to the object, the object processing execution unit 145 determines that the object is a candidate object to be displayed.

  The determination region typically extends from one outer end of the object to the other outer end unless specified by the user. The determination area may be a three-dimensional area as in this example, a line made of continuous voxels, or a surface having a predetermined shape. For example, in the example of FIG. 4, the object process execution unit 145 can use the determination line 22 as a determination region. When the voxel on the determination line 22 belongs to the object, the object processing execution unit 145 determines that the object is a candidate object to be displayed.

  The line is one of the preferable shapes of the determination region. The determination area, which is a line, can include only the part desired by the user. The line is preferably a straight line so that the relationship between the determination region and the object can be easily understood and from the viewpoint of simplicity of processing.

  On the other hand, when the determination area is a line, the displayed candidate object changes due to a shift of several voxels at the designated position. Therefore, in a more preferable configuration, the determination area is a three-dimensional area having a predetermined outer shape. Thereby, it is possible to appropriately display candidate objects in an area desired by the user without being affected by a slight shift of the designated position.

  From the viewpoint of easy understanding of the relationship between the judgment area and the object and the simplicity of processing, the solid area is a solid that extends in the straight line direction including the user-specified position and has a constant cross-sectional shape with the straight line as a normal line. It is preferable. A typical cross-sectional shape is a circle as in the determination region 23 of FIG.

  In the example of FIG. 4, the direction of the determination straight line 22 is the normal direction of the initial visualized image (projection plane thereof). The direction in which the determination area extends from the two-dimensional position designated by the user is preferably this direction from the viewpoint of easy understanding of the relationship between the determination area and the object and the ease of processing. Similarly, from the point, the direction of the determination straight line may coincide with the line-of-sight direction of the designated position in the initial visualization image. When the initialized visualized image is a cross section, the line-of-sight direction at the designated position is the normal direction. Depending on the design, the direction of the determination straight line may be different from these directions.

  In addition, the shape of the determination area (including the outer shape and the extending direction thereof) may be set in advance or may be specified by the user. In the configuration that can be designated by the user, the designated area calculation unit 143 determines the shape of the determination area according to the user input from the input device 13. For example, the user selects one of a straight line and a cylinder. When a straight line is selected, the direction is specified. When a cylinder is selected, the direction and the size of the cross section are specified.

<Select object>
Next, the display of a candidate object image that overlaps the determination area and the selection of an object for specifying a region of interest by the user (S16 to S18) will be specifically described. First, according to the example shown in FIG. 4, an example of displaying a candidate object image in the determination area 23 and selecting an object by the user will be described.

  5A to 5F show different screen displays of the display device 12. In this example, the display device 12 sequentially displays the images of FIGS. 5A to 5F on the screen. 5A to 5F, the screen of the display device 12 displays images of candidate objects in the determination area 23. In this example, the skin 31, the skull 32, and the brain 33 are objects in the determination region 23. The skin 31 is on the outermost side, the skull 32 is inside, and the brain 33 is inside the skull.

  5A to 5F, the left figure is an image with respect to the projection plane set in front of the head, and the right figure is an image with respect to the projection plane set beside the head. In this example, the left diagram and the right diagram are two-dimensional images having three-dimensional information, which are images formed using a VR or MIP technique. The left diagrams of FIGS. 5A to 5F show candidate objects in an emphasized manner. The right figure displays candidate positions 24 that indicate candidate objects. These points are the same as in FIG. Therefore, the X, Y, and Z axes in FIG. 4 are applied to FIGS. 5A to 5F.

  The display device 12 displays the images of FIGS. 5A to 5F in order on the screen according to the user input. When the user selects an object in any of the images, the image processing apparatus 11 proceeds to a region of interest designation step (S19 in FIG. 2) without displaying the subsequent images. The image processing apparatus 11 may accept the user's selection after displaying all the images.

  The display device 12 first displays the image of FIG. 5A. FIG. 5A shows an image in which skin 31 that is the outermost object is selected as a candidate object. The visualization process execution unit 144 acquires information about the object 31 from the object process execution unit 145, and generates image data of the two images 51 and 52 in FIG. 5A.

  Specifically, the object processing execution unit 145 identifies the object 31 closest to the designated position 21 on the straight line (determination line 22) extended from the designated position 21 in the Z direction. When the object is defined in the volume data 151, the object processing execution unit 145 is obtained from the information about the object attached to the volume data 151, and from the information about the object estimated in the volume data 151 when the object is not defined. Can specify the object closest to the designated position 21.

  The visualization processing execution unit 144 acquires information related to the object 31 from the object processing execution unit 145, and generates image data of the two images 51 and 52 shown in FIG. 5A. The image data of the right figure 52 includes the data of the visualization image of the head and the data for displaying the candidate position 24 in the image. The candidate position 24 is an intersection position between the determination line 23 and the skin 31 that is a candidate object, and is a position closest to the designated position 21. As shown in FIG. 5A, the right diagram 52 includes a graphic showing the candidate position 24 in addition to the visualized image of the head.

  The visualization process execution unit 144 specifies the candidate object 31 so that the candidate object 31 can be clearly seen in the two images 51 and 52. In the example of FIG. 5A, the outline of the candidate object (skin) 31 is schematically displayed as a solid line, and the outlines of other objects are displayed as dotted lines.

  In a preferred configuration, the visualization process execution unit 144 performs different visualization processes on the voxels belonging to the candidate object 31 and other voxels in the generation of the image data 51 and 52. For example, the visualization processing execution unit 144 visualizes voxels belonging to the candidate object 31 with VR, and visualizes other voxels with MIP to generate image data 51 and 52 (solid lines in FIG. 5A are VR, dotted lines). Is MIP).

  In this way, by generating image data by applying different visualization methods to the candidate object and the other parts, the visualized image displays the candidate object together with the other parts, and also displays the candidate object from the other parts. It can be clearly distinguished and displayed.

  The visualization process execution unit 144 may distinguish and display candidate objects by other methods. For example, the visualization process execution unit 144 displays only the voxels that belong to the candidate object 31, and does not display other parts (dotted line parts in FIG. 5A). In another example, the visualization processing execution unit 144 applies the same visualization method to all voxels, but sets a color or transparency different from other voxels for voxels belonging to the candidate object. Candidate objects can be displayed separately from other parts.

  From the viewpoint of ease of visual recognition by the user, it is preferable to use the same visualization process for the generation of the image data in the left diagram 51 and the generation of the image data in the right diagram 52. However, the visualization process execution unit 144 may use different visualization processes in the data generation of the two images. For example, the visualization process execution unit 144 may show a cross-sectional view in FIG. Alternatively, in the generation of the image data in the left figure 51, different visualization methods are applied to the candidate object and the other parts, and in the generation of the image data in the right figure 52, the same visualization method is applied to all the voxels. Thus, the candidate object may be set with a different transparency from other parts.

  When the first candidate object 31 shown in FIG. 5A is not selected, the user instructs the input device 13 to display the next candidate object. For example, using the mouse wheel or the arrow keys of the numeric keypad as the input device 13, the user inputs an instruction to advance the candidate position 24 in the positive direction of the Z axis on the determination line 22.

  The input operation acquisition unit 142 notifies the object processing execution unit 145 that the candidate position 24 has advanced one step in the positive Z-axis direction on the determination line 22. The object processing execution unit 145 determines the next object 32 on the determination line 22 as a candidate object. The object 32 is a skull.

  The visualization process execution unit 144 acquires information related to the skull object 32 from the object process execution unit 145, and generates data of two images illustrated in FIG. 5B. The generation method of the image data is the same as the generation of the image data of the images 51 and 52 shown in FIG. 5A. The images 53 and 54 in FIG. 5B display the candidate object 32 while distinguishing it from other parts, and also display points indicating the candidate position 24 corresponding to the candidate object 32.

  When the candidate object 32 is not selected, the user inputs an instruction to advance the candidate position 24 one step in the positive direction of the Z axis from the input device 13. Similar to the generation of the image data illustrated in FIG. 5B, the visualization processing execution unit 144 generates the image data of the image illustrated in FIG. 5C from the information on the object 33 obtained from the object processing execution unit 145. The next candidate object 33 is the brain.

  As described above, when the user advances the candidate position 24 step by step with the input device 13, the visualization processing execution unit 144 displays the screen display 1, 2, 3, 4, 5 as shown in FIGS. Change the image. When the display device 12 displays the target object as a candidate object, the user inputs information for selecting the candidate object displayed from the input device 13. The input operation acquisition unit 142 transfers information that the current candidate object has been selected as the selected object to the designated area calculation unit 143. The designated area calculation unit 143 defines the current candidate object being displayed as a selected object.

  In the above example, images displaying the same object are the same image. That is, FIGS. 5A and 5F show the same image, FIGS. 5B and 5E show the same image, and FIGS. 5C and 5D show the same image. Unlike this, the visualization processing execution unit 144 may display images of different projection planes of the same object in accordance with the candidate positions 24 in the left diagrams of FIGS. 5A to 5F.

  In a preferred configuration, the visualization processing execution unit 144 displays two images on one screen as shown in FIGS. 5A to 5F. By simultaneously displaying images from different directions, the user can recognize the candidate object more accurately. Depending on the design, the visualization processing execution unit 144 may sequentially display only one type of image. For example, in the above example, only one of the left diagram and the right diagram is sequentially displayed.

  Preferably, in a configuration in which one type of image is displayed, the visualization processing execution unit 144 displays an image in the same direction as the initial visualization image in which the user specifies the initial position. When the one type of image is an image on a projection plane using VR or MIP, an image in the same direction has the same projection plane and line of sight. In the above example, the left image in the above example corresponds to this image. Thereby, the user can visually recognize the candidate object in relation to the initial visualized image.

  In a configuration in which different images are displayed at the same time, it is preferable that one of the images clearly indicates the position of the candidate object in the determination area. In the above example, the image of the candidate position 24 on the determination line 22 in the right figure clearly shows this position. The plane of this image is preferably parallel to the determination area as shown in the right figure of the above example, but it need not be parallel. The other image is preferably in the same direction as the initial visualized image.

  The visualization process execution unit 144 may display the image of the candidate object by a method different from the preferred examples shown in FIGS. 5A to 5F. In another preferable configuration, the visualization processing execution unit 144 displays a plurality of candidate object images collectively on the screen. FIG. 6 displays an image of the object in the head volume data according to the example of FIG. The display screen shown in FIG. 6 shows an example in which all candidate objects in the determination area 23 are collectively displayed.

  The screen in FIG. 6 shows an initial visualization image 61 and candidate object images 62 to 64. A part of each of the skin, bone, and brain objects 31 to 33 is a voxel in the determination region 23, and the candidate object images 62, 63, and 64 display the skin 31, the skull 32, and the brain 33, respectively. Yes. The user can designate a candidate object image to be selected from the displayed candidate object images using the input device 13.

  The visualization processing execution unit 144 receives information on all objects to which the voxels in the determination area 23 belong from the object processing execution 145, and displays those images. It is preferable to attach numbers and alphabets to the displayed images of each object, as this example shows 1: skin, 2: bone, 3: brain.

  The visualization process execution unit 144 preferably displays all the candidate object images at once, but may display an image set of some of the plurality of object objects on the screen. Good. The visualization process execution unit 144 sequentially displays different image sets according to user input. The visualization processing execution unit 144 preferably displays the initial visualization image 61, but may omit it. The visualization process execution unit 144 may display the images 52 shown in FIGS. 5A to 5F together.

<Object estimation>
Hereinafter, the object estimation (S15) and the estimated object image display method (S16) shown in the flowchart of FIG. 2 will be described in detail. A preferred estimation method uses voxel brightness in the volume data. Hereinafter, a method for estimating an object using a change in luminance and a method for estimating an object using a luminance distribution will be described.

  First, an example of a method for estimating an object using a change in luminance will be described with reference to FIG. In many cases, the luminance value of a voxel shows a significant change at the boundary of the object. That is, the difference in luminance value between adjacent voxels becomes significantly large at the boundary of the object. Therefore, the object processing execution unit 145 can estimate the object by calculating the change in the voxel luminance value.

  7, the upper diagram is a model diagram of the head, the user-specified position 21, the determination line 22, and the determination region 23 as seen from the Z-axis direction and the X-axis direction, as in FIG. The middle diagram in FIG. 7 is a model diagram of a graph showing the luminance V of the voxel on the determination line 22 in the upper diagram. The horizontal axis represents coordinates on the Z axis, and the vertical axis represents luminance V. In each coordinate value on the Z axis, the luminance V may be an average value of luminance values of voxels forming a plane perpendicular to the Z axis in the determination region 23.

The lower diagram in FIG. 7 is a model diagram of a graph showing a difference in luminance value between adjacent voxels. In the graph, the horizontal axis indicates the Z-axis coordinate, and the vertical axis indicates the luminance difference V _dif . When the luminance average value of the surface perpendicular to the Z axis is used, the luminance difference V _dif is a difference in luminance average value between adjacent surfaces. A position where V _dif is large is a position where V shows a large change. The object processing execution unit 145 sets the position where the luminance difference V _dif exceeds the specified threshold value VD _th as Z _s . In this example, the object processing execution unit 145 detects six positions Z _{0 to} Z ₅ . These positions correspond to the six candidate positions shown in FIGS. 5A to 5F.

  The object processing execution unit 145 performs the same processing for all the pixels in the initial visualized image, and obtains the coordinate information of the object in the head volume data. In this example, coordinate information of objects corresponding to skin, skull, and brain is obtained. The visualization processing execution unit 144 acquires coordinate information of these objects from the object processing execution unit 145, generates image data that is substantially the same as the images shown in FIGS. 5A to 5F and FIG. 6, and responds to user input. Display these images.

  In a preferred configuration, the object processing execution unit 145 calculates coordinate data indicating the entire object as described above. In another configuration, the object processing execution unit 145 performs the above processing only for voxels in the determination area 22. This reduces the processing amount for object estimation.

  Specifically, the object processing execution unit 145 calculates a luminance change for pixels in the determination area 23 in the initial visualized image. The object processing execution unit 145 estimates an object in an area included in the determination area 23 and generates coordinate data of each object. The visualization process execution unit 144 generates display image data from the object information acquired from the object process execution unit 145 and the volume data.

  The visualization process execution unit 144 generates the image data of the right diagram or the left diagram in FIGS. 5A to 5F and displays the image. At this time, the visualization processing execution unit 144 may simultaneously display the left and right diagrams in FIGS. 5A to 5F. The visualization process execution unit 144 highlights and displays the candidate object in the determination area 22 and distinguishes the candidate object part from other parts. It is preferable that the visualization process execution unit 144 also displays the candidate position 24 together.

  In another configuration, the object processing execution unit 145 calculates a luminance change only for voxels on the determination line 22 and estimates an object only on the determination line 22. As described with reference to FIG. 7, the object processing execution unit 145 detects the boundary position of the object on the determination line 22. The visualization process execution unit 144 displays the candidate position 24 in FIGS. 5A to 5F. The display of the candidate position 24 indicates a candidate object, and the visualization processing execution unit 144 displays the candidate object by distinguishing it from other parts by displaying the candidate position 24.

  When the user selects one candidate position via the input device 13, the designated area calculation unit 143 defines the selected candidate position as the designated position for the area of interest designation (S19). The image processing algorithm execution unit 141 executes a region of interest designation step (S19) based on this position.

  Next, with reference to FIG. 8, a method for estimating an object using the distribution of luminance of voxels (S15) will be described. The upper diagram in FIG. 8 is a model diagram in which the head, the user-specified position 21, the determination line 22, and the determination region 23 are viewed from the Z-axis direction and the X-axis direction, as in the upper diagram of FIG. The middle diagram in FIG. 8 is a model diagram of a graph showing the luminance V of the voxel on the determination line 22 in the upper diagram. The horizontal axis represents coordinates on the Z axis, and the vertical axis represents luminance V. In each coordinate value on the Z axis, the luminance V may be an average value of luminance values of voxels forming a plane perpendicular to the Z axis in the determination region 23.

  In medical images, it is often possible to estimate the brightness of an object, that is, an organ or tissue. This estimation method uses this property. As shown in FIG. 9, a correspondence table 91 between objects and the luminance of voxels constituting the objects is prepared in advance so that the image processing algorithm execution unit 141 can use the correspondence table 91. In the object luminance correspondence table 91, object 1, object 2, and object 3 correspond to skin object 31, brain object 33, and skull object 32, respectively.

The object processing execution unit 145 sequentially refers to the luminance of the voxels on the determination line 22 from the designated position Z ₀ on the determination line 22 in the positive direction of the Z axis. When the object processing execution unit 145 refers to the object luminance correspondence table 91 and detects the first voxel within the luminance range of the object, the object processing execution unit 145 stores the coordinate value of the voxel as the object position. In the example of FIG. 8, the positions (voxels) indicated by Z _{1 to} Z ₇ are positions that satisfy this condition.

The visualization processing execution unit 144 acquires the voxel coordinates Z _{1 to} Z ₇ from the object processing execution unit 145, and the initial position Z coordinates Z _{1 to} Z ₇ specified first together with the visualization image of the right diagram or the left diagram in FIG. Candidate objects are indicated by specifying voxels. For example, when the user inputs an instruction to move the candidate position in the Z direction using the input device 13, the input operation acquisition unit 142 acquires the instruction and sends it to the object processing execution unit 145.

The object processing execution unit 145 having obtained this information sequentially sends the above-described coordinates of Z _{1 to} Z _{7 to} the visualization processing execution unit 144 each time the candidate position advances one step in the positive direction of the Z axis. The visualization process execution unit 144 displays a new candidate position by specially displaying the voxel of the acquired coordinate value.

In this way, the visualization process execution unit 144 moves the candidate position to be displayed from Z _{0 to} Z ₇ . When the display candidate position moves to the target position, the user inputs information indicating that the candidate position displayed by the input device 13 is selected. The input operation acquisition unit 142 sends a determination that the displayed candidate position is the designated position to the designated area calculation unit 143. The designated area calculation unit 143 defines the candidate position as the designated position. A region of interest designation step (S18) determines a region of interest based on this designated position.

In FIG. 8, not all voxels at positions Z _{1 to} Z ₇ are actually included in the corresponding objects shown in the object luminance correspondence table 91. Specifically, the position Z ₂ and the position Z ₆ do not exist in the brain 33 but appear near the boundary between the skin 31 and the skull 32. If the type of object to be specified is not specified and displayed, there is no problem in specifying such a position. This is because, if the vicinity of the boundary position of the object can be clearly indicated, the user can easily specify the region of interest based on the position.

  The visualization process execution unit 144 may specify a position outside the corresponding object on the basis of the distance between adjacent positions, and exclude the position from the candidate positions to be displayed. Specifically, the object processing execution unit 145 excludes the target position from the candidate positions when the distance between the target position and the next position is equal to or less than a predetermined threshold.

In a preferred configuration, the visualization processing execution unit 144 preferably displays images in different directions at the same time as shown in FIGS. 5A to 5F, and further highlights portions other than candidate positions in the object in the display image. . For example, the visualization processing execution unit 144 smoothes and analyzes the luminance of the voxel in the volume data, and highlights and displays the voxel having the same luminance as the candidate position in the voxels near the candidate position. For example, when the designated position is the position Z ₄ , the visualization processing execution unit 144 performs visualization display that emphasizes the voxel having the luminance V ₃ .

  The visualization display that emphasizes the voxel is displayed by changing the visualization method so that VR is applied to the voxel having the luminance value and MIP is applied to the other voxels as described in step 12 in FIG. There is a method of changing the color and transparency of a voxel having a specific luminance value and other voxels in the same visualization method.

  The object processing execution unit 145 may determine the range of the object at the candidate position with reference to the object luminance correspondence table 91. The object processing execution unit 145 detects voxels within the corresponding luminance range within a range continuous from the candidate position. The range formed by these voxels is estimated (determined) as the range of the object.

  The object process execution unit 145 performs this process in the determination area 23 or the entire head. Based on this information, the visualization processing execution unit 144 displays the object separately from other parts in the determination area 23 or in the entire head as shown in FIGS. 5A to 5F.

  The object estimated based on the voxel brightness described with reference to FIGS. 7 and 8 (at least a part thereof) differs from the configuration in which the object is defined in the volume data, and is an actual object, that is, a target object. It may not exactly match the organs and tissues of In the object image display (S16) in this configuration, the object image estimated by the object processing execution unit 145 is displayed as described above.

  The object image display (S16) is a process for specifying a region (position) serving as a reference for the region of interest designation (S18) by the user thereafter. Therefore, the candidate object to be displayed does not necessarily match the actual target object. However, in order to more accurately support the specification of the region of interest (S18), the object processing execution unit 145 does not determine the candidate object by estimating based on the luminance, but according to the definition information in the volume data. It is preferable to determine candidate objects. Note that both the predefined object and the estimated object may be displayed.

<Specify region of interest>
Next, the region of interest designation (S18) according to the user input will be described in more detail. Step 18 specifies the region of interest in more detail within the selected object. If the user determines that none of the displayed candidate objects is to be used (N in S17), the user designates a two-dimensional designated position (S13) in the initial visualized image, and then performs steps 14 to 18 This step 19 is skipped and executed.

  In step 19, the user inputs the shape, size, and position of the region of interest from the input device 13. The input operation acquisition unit 142 acquires an input from the input device 13, and the designated region calculation unit 143 calculates a region of interest from the user input received from the input operation acquisition unit 142. It should be noted that the shape, position, and size of the region of interest are not specified in any order, and the region of interest should be displayed on the display device each time so that it can be specified and corrected multiple times.

  First, shape specification will be described. The user can use a basic figure such as a point, polyhedron, sphere, cylinder, or cone, a closed curve drawn freehand, or the like as the shape of the region of interest. For example, the user designates one figure from basic figures prepared in advance. The designated region calculation unit 143 defines the selected figure as the shape of the region of interest.

  In an example of a method of using a freehand closed curve or polyhedron, the user draws a two-dimensional shape on a two-dimensional plane of one viewpoint, and the designated region calculation unit 143 extends the two-dimensional shape in the depth direction. Is defined as a designated area. In another example, the user draws a two-dimensional shape in each of cross-sectional images from different viewpoints, for example, an MPR image or a CMPR image. The designated area calculation unit 143 defines a three-dimensional shape having these as an outline as the shape of the region of interest.

  As another preferable designation method, a three-dimensional region is extracted by an automatic extraction method (automatic extraction algorithm). A typical automatic extraction method is the Region Growing method. Region Growing refers to one or a plurality of seed points (starting points) and refers to adjacent pixels, and determines whether or not a predetermined Growing condition is met based on information such as luminance. In this case, the adjacent pixels are set in the area, and the area is expanded using all the pixels in the area as a seed.

  In this case, the user can use the two-dimensional position (S13) designated by the input device 13 or designated in the initial visualization image as the seed point. The designated region calculation unit 143 performs the Region Growing process using the designated seed point and a predetermined threshold and condition or a threshold and condition adjusted via the input device 13. As another automatic extraction method, a Watershed method, a Snakes method, or the like can be used.

  In addition, when the target object is designated (Y in S17 and object designation in S18), the user can designate the object as a region of interest. Upon receiving an input for designating the selected object as a three-dimensional shape via the input device 13 and the input operation acquisition unit 142, the designated area calculation unit 143 obtains information on the voxels belonging to the current object from the object processing execution unit 145. And define the region as the shape of the region of interest.

  The position of the region of interest can be specified using the two-dimensional position specified in step 13 as a reference. From there, the movement in the display of a plurality of two-dimensional images projected from a plurality of viewpoints such as MPR and CMPR is repeatedly determined. In addition, the pixel displayed on the screen visualized by MIP, that is, the point on the Z axis at the XY position designated on the display screen or the point of the maximum luminance in the determination area 23 is used as the designated point. May be. The size of the region of interest is specified via the input device 13 and the input operation acquisition unit 142. For example, the direction key of the keyboard, the specification using the mouse / wheel, the specification using the numerical value, and the like are used as input.

  When specifying a region of interest in an image including a selected object and other regions, it is preferable that the user can select which region the specified position is to be moved. In one example, different operation methods of the input device 13 are applied to the movement in the candidate object and the movement in other areas.

  For example, an operation that only rotates the wheel is an input of movement in the selected object, and a wheel rotation while pressing a key (for example, Ctrl) on the keyboard is an input of movement in an area other than the selected object. Alternatively, the user moves the designated position with the input device 13 after inputting from the input device 13 to select one of the areas. These determinations are made by the input operation acquisition unit 142.

<Storage of region of interest data>
In storing the designated region of interest, the designated region calculation unit 143 defines the region of interest in association with the volume data, and stores it in the internal memory 20 (and the external storage device 10 if necessary). An example of the method stores all the three-dimensional positions (three-dimensional coordinates) of voxels in the region of interest as the region-of-interest information in association with the volume data. Another example generates mask data having the same number of voxels as the volume data, gives different mask values to voxels in the region of interest and voxels outside the region of interest, and stores the mask data.

  The generation of the mask data gives, for example, a value 1 to voxels in the region of interest and a value 0 to voxels outside the region of interest. If the region of interest has a specified shape such as a sphere, cube, or cylinder, the reference coordinates (center coordinates and feature points (rectangular corner coordinates)), specified shape information, and the size of the region of interest If stored, the region of interest can be reproduced from the volume data based on the information.

  Next, when a region extracted by a three-dimensional region automatic extraction method such as the Region Growing method is used as a region of interest, a parameter including one or a plurality of reference voxels (seed) in the extraction method is stored. It is possible to define the same region of interest with less data than storing mask data. Note that if the region of interest is enlarged / reduced with a three-dimensional shape, the parameter includes the information (enlargement / reduction rate).

  In addition, when using the Watershed method as a three-dimensional region automatic extraction method, as in Region Growing, when storing parameters including one or more reference voxels (seeds) and thresholds, and using the Snakes method , Storing an energy function and parameters including a plurality of control points.

  In the storage of the region of interest, there is a case where detailed information on the region of interest is not necessary. When extracting an organ or diseased part and measuring its maximum diameter, accuracy in voxel units is required. For example, there is no need for a detailed shape or detailed position, but a rough position and a rough three-dimensional range. In some cases, it is only necessary to memorize only.

  As described above, even when the region of interest is specified in detail, the specified region calculation unit 143 is complicated when storing the region of interest (data defining) by reducing the amount of information during storage. The three-dimensional shape may be deformed into a preset specified shape, and the deformed region of interest (data defining) may be stored. For example, the designated area calculation unit 143 stores the coordinates of the reference position (for example, the coordinates of the center and feature points), information on the prescribed shape, and the size of the area of interest. Since the appropriate storage method varies depending on the situation, such as the data size, the ease of handling the stored data, the number of regions of interest, etc., it is important to use a storage method that suits the purpose of use.

  Although the present invention has been described in detail with reference to the accompanying drawings, the present invention is not limited to such specific configurations, and various modifications and equivalents within the spirit of the appended claims Includes configuration.

  For example, as described above, it is preferable that the image processing apparatus displays, as candidate objects, objects that overlap with the determination area determined according to the user input. As a result, an object suitable as a candidate object can be selectively displayed. However, depending on the design, the image processing apparatus may not use the determination area. For example, the image processing apparatus may sequentially or collectively display all objects defined in the volume data.

DESCRIPTION OF SYMBOLS 1 Medical image processing system, 10 External storage device, 11 Image processing apparatus, 12 Display apparatus 13 Input device, 15 Internal memory, 20 Internal memory 21 Image processing algorithm execution part, 21 Two-dimensional designation | designated position, 22 Determination straight line 23 Determination area | region, 24 candidate positions, 31 skin objects, 31 skin objects 32 skull objects, 33 brain objects, 61 initial visualization images 62 to 64 candidate object images, 91 object luminance correspondence table 141 image processing algorithm execution unit, 142 input operation acquisition unit 143 designated area Calculation unit, 144 Visualization processing execution unit, 145 Object processing execution unit 151 Volume data, 152 Initial visualization image data 153 Determination area specification data, 154 Candidate object specification data 155 Candidate object image data, 156 Heart area image data

Claims (15)

An image processing system for three-dimensional medical image data, comprising an input acquisition unit for acquiring user input, a data storage device, an image processing unit, and a display device,
The data storage device stores the three-dimensional medical image data,
The image processing unit visualizes the three-dimensional medical image data to generate one or a plurality of candidate object image data that clearly indicate one or a plurality of candidate objects, and the one or a plurality of candidates Storing object image data in the data storage device;
The display device displays an image of the one or more candidate object image data stored in the data storage device;
The image processing unit selects an object for designating a three-dimensional region of interest from the one or more candidate objects according to a user selection input for the displayed image,
The input acquisition unit is an image processing system that receives a designation input of the three-dimensional region of interest from the user for a display image of the selected object. The image processing unit visualizes the three-dimensional medical image data to generate initial visualized image data, stores the initial visualized image data in the data storage device,
The display device displays an image based on the initial visualized image data,
The image processing unit determines a determination region in the three-dimensional medical image data according to a user input with respect to the displayed initial visualized image,
The image processing system according to claim 1, wherein a part of each of the one or more candidate objects is included in the determination region.   The image processing system according to claim 2, wherein the determination area has a three-dimensional shape.   The image processing system according to claim 3, wherein the determination area is an area that extends along a straight line including a user-specified two-dimensional position in the initial visualization processing image and has a constant cross section.   The image processing system according to claim 1, wherein the one or more candidate objects are defined in advance in the three-dimensional medical image data. The image processing unit estimates one or more objects in the three-dimensional medical image data based on voxel luminance;
The image processing system according to claim 1, wherein the one or more candidate objects are included in the estimated one or more objects. The image processing system estimates the one or more objects in the determination region based on voxel brightness in the determination region,
The image processing system according to claim 2, wherein the image processing system generates one or a plurality of image data that clearly indicates each of the one or a plurality of candidate objects in the determination region. The image processing unit generates two candidate object image data having different directions for each of the one or more candidate objects, and stores the candidate object image data in the data storage device,
The image processing system according to claim 1, wherein the display device displays the two candidate object images simultaneously. The image processing unit generates two candidate object image data having different directions for each of the one or more candidate objects, and stores the candidate object image data in the data storage device,
The display device simultaneously displays the two candidate object images;
The image processing system according to claim 2, wherein one of the two candidate object images clearly indicates a position of the candidate object in the determination area.   The image processing system according to claim 9, wherein the display device sequentially displays a plurality of sets of the two candidate object images. The image processing unit stores, as image data of the three-dimensional region of interest, parameters of a three-dimensional region extraction process in the data storage device;
The said image processing part produces | generates the image data of the said three-dimensional region of interest in the said three-dimensional medical image data by the said three-dimensional area image extraction process using the said parameter. Image processing system. In the image display of 3D medical image data in a medical image processing system, a program for causing a processor to execute a process for supporting the designation of a 3D region of interest by a user.
Visualizing the three-dimensional medical image data to generate one or a plurality of candidate object image data specifying each of one or a plurality of candidate objects, and storing the one or a plurality of candidate object image data as data Memorize in the device,
Outputting the one or more candidate object image data stored in the data storage device to a display device;
According to a user selection input for the displayed one or more candidate object images, an object for specifying the three-dimensional region of interest from the one or more candidate objects,
Receiving a designation input of the three-dimensional region of interest from the user for the image display of the selected object. The process is
Visualizing the three-dimensional medical image data to generate initial visualized image data, storing the image data in the data storage device,
Outputting the initial visualized image data to the display device;
In accordance with a user input for the displayed initial visualized image, a determination region in the three-dimensional medical image data is determined,
The program according to claim 12, wherein a part of each of the one or more candidate objects is included in the determination area. An image processing method in a three-dimensional medical image data processing system,
Storing the three-dimensional medical image data in a data storage device;
Visualizing the three-dimensional medical image data to generate one or a plurality of candidate object image data specifying each of one or a plurality of candidate objects, and storing the one or a plurality of candidate object image data in the data storage Memorize in the device,
Displaying an image based on the one or more candidate object image data stored in the data storage device;
According to a user selection input for the displayed one or more images, an object for designating a three-dimensional region of interest from the one or more candidate objects is selected,
An image processing method for receiving a designation input of the three-dimensional region of interest from the user for a display image of the selected object. Visualizing the three-dimensional medical image data to generate initial visualized image data, storing the initial visualized image data in the data storage device,
Displaying the initial visualized image data;
In accordance with a user input for the displayed initial visualized image, a determination region in the three-dimensional medical image data is determined,
The image processing method according to claim 14, wherein a part of each of the one or more candidate objects is included in the determination area.

Images