JP2007000267A - Area extracting apparatus, area extracting method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technique by which the original condition of a standard model before fitting action can be more effectively adjusted when applying a model fitting method on a plurality of objects. <P>SOLUTION: An area extracting apparatus 1 extracts the first object area BJ1 of a plurality of object areas from measured data by means of the model fitting method using the first standard model MS1 of a plurality of standard models, and obtains the variable power quantity and/or moving quantity of the first model MS1 at that time as the relative relationship between the first standard model and the first object area. Further, the apparatus 1 corrects the original condition of the second standard model MS2, (MS3) of a plurality of models based on the relative relationship and extracts the second object area BJ2, (BJ3) of a plurality of object areas from the measured data by means of the model fitting method using the corrected second standard model. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a technique for extracting a region related to a desired object from measurement data.

  In recent years, in-vivo three-dimensional images taken by an X-ray CT (Computed Tomography) apparatus or an MRI (Magnetic Resonance Imaging) apparatus are widely used in the medical field. As a result, information on internal organs and the like can be visually grasped, and an improvement in diagnostic accuracy can be expected. However, on the other hand, since the three-dimensional image used for diagnosis is composed of tens to hundreds of slice images, only information necessary for diagnosis is obtained from such an enormous amount of information. Is a big burden for doctors.

  Therefore, there is an increasing demand for quantitative or automatic diagnosis using a computer, and research on a computer aided diagnosis (CAD Computer-aided diagnosis) system has been actively conducted.

  In order to perform diagnosis support by a computer, it is an important issue to accurately extract information necessary for diagnosis, that is, an organ region or a shape.

  As an organ region extraction method, a model fitting method for finding a target contour by deforming a standard model prepared in advance based on the energy minimization principle has been proposed (Non-patent Document 1).

  According to this method, since a corresponding region is extracted by deforming a model having a feature in advance, which is a standard model, high-precision extraction processing is possible.

Tsagaan Bigalma, Akinobu Shimizu, Hidefumi Obata, “Development of a method for extracting kidney regions from abdominal CT images using a three-dimensional deformable model”, IEICE Transactions, D-II, January 2002, Vol. D-II, No.1, pp.140-148

  The quality of the extraction result obtained by the above model fitting method depends on the initial state (initial position, etc.) of the standard model before the fitting operation. That is, in the model fitting method, it is important to perform initial alignment and the like with high accuracy.

  By the way, in such an extraction process, not only a single target object but also a plurality of target objects are extracted, that is, an area (target object area) corresponding to a plurality of target objects is applied by applying a model fitting technique. It can be considered to extract from a three-dimensional image. Specifically, it is conceivable that standard models corresponding to each of a plurality of objects are prepared in advance, and each object is extracted from a three-dimensional image using the plurality of standard models.

  In this case, in order to obtain a good extraction result, it is required to accurately perform initial alignment between each of the plurality of objects and a plurality of standard models corresponding to each of the plurality of objects.

  However, in order to accurately perform the initial alignment as described above, if the initial alignment process is performed individually for each of a plurality of objects, it is necessary to repeat the same initial alignment process for the number of objects. There is a problem that it is very time-consuming.

  Therefore, an object of the present invention is to provide a technique capable of more efficiently adjusting the initial state of the standard model before the fitting operation when the model fitting method is applied to a plurality of objects. It is in.

  In order to solve the above-mentioned problem, the invention of claim 1 is an area extraction apparatus for extracting a plurality of object areas from measurement data, the means for storing a plurality of standard models relating to different objects, respectively, Means for extracting a first object region of the plurality of object regions from the measurement data by a model fitting technique using a first standard model of the standard models; and The amount of magnification and / or movement of the first standard model when fitting to the first object region is obtained as a relative relationship between the first standard model and the first object region. A correction means for correcting an initial state of a second standard model of the plurality of models based on the relative relationship, and the corrected second standard model, A second object region among the serial plurality of object area, characterized in that it comprises means for extracting the model fitting method from the measurement data.

  According to a second aspect of the present invention, in the region extracting apparatus according to the first aspect of the invention, the correction means corrects an initial position of the second standard model.

  According to a third aspect of the present invention, in the region extracting apparatus according to the first or second aspect of the present invention, the correction means corrects an initial size of the second standard model.

  The invention of claim 4 is an area extraction method for extracting a plurality of object areas from measurement data, wherein the plurality of objects are obtained by using a first standard model among a plurality of standard models for different objects. A step of extracting a first object area of the object area from the measurement data by a model fitting technique, and the first standard model when fitting the first standard model to the first object area A scaling amount and / or a moving amount of the first standard model and the first object region as a relative relationship, and based on the relative relationship, Using the step of correcting the initial state of the second standard model and the second standard model after correction, the second object area of the plurality of object areas is calculated as the total object area. Characterized in that it comprises a step of extracting the model fitting method from the data.

  The invention according to claim 5 is a region extraction method for extracting a plurality of object regions from measurement data in a computer, wherein a first standard model among a plurality of standard models for different objects is used, A step of extracting a first object area of a plurality of object areas from the measurement data by a model fitting technique, and the first standard model when fitting the first standard model to the first object area. Obtaining a scaling amount and / or a moving amount of the standard model as a relative relationship between the first standard model and the first object region, and based on the relative relationship, the plurality of models And correcting the initial state of the second standard model, and using the corrected second standard model, the second pair of the plurality of object regions. Wherein the objects area is a program for executing a step of extracting the model fitting method from the measured data, a region extraction method comprising.

  According to the first to fifth aspects of the invention, when the model fitting method is applied to a plurality of objects, the initial state of the second standard model can be adjusted more efficiently.

  In particular, according to the second aspect of the present invention, the initial position of the second standard model can be corrected more efficiently.

  In particular, according to the third aspect of the invention, the initial size of the second standard model can be corrected more efficiently.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<A. Configuration>
FIG. 1 is a diagram showing an outline of a region extraction device 1 according to an embodiment of the present invention. The region extraction device 1 is a device that extracts a desired region (specifically, a three-dimensional region) from measurement data (specifically, three-dimensional measurement data (also referred to as three-dimensional measurement data)).

  As shown in FIG. 1, the region extracting apparatus 1 includes a personal computer (hereinafter simply referred to as “personal computer”) 2, a monitor 3, an operation unit 4, and a mounting unit 5.

  The personal computer 2 includes a control unit 20, an input / output I / F 21, and a storage unit 22.

  The input / output I / F 21 is an interface (I / F) for transmitting / receiving data between the monitor 3, the operation unit 4, the mounting unit 5, and the personal computer 2. Send and receive.

  The storage unit 22 is composed of, for example, a hard disk and stores a software program (hereinafter simply referred to as “program”) PG for executing region extraction.

  The control unit 20 mainly includes a CPU, a ROM 20a, a RAM 20b, and the like, and is a part that performs overall control of each unit of the personal computer 2.

  The monitor 3 is configured by a CRT, for example, and visually outputs a display image generated by the control unit 20.

  The operation unit 4 includes a keyboard and a mouse, and transmits various signals to the user (user) according to various operations to the input / output I / F 21.

  The mounting unit 5 can detachably mount a storage medium such as the memory card 51. Various data or programs stored in the memory card 51 attached to the attachment unit 5 can be taken into the control unit 20 or the storage unit 22 via the input / output I / F 21.

  In this embodiment, from the three-dimensional measurement data (three-dimensional image) acquired by the X-ray CT apparatus, each region (object region) BJ1, BJ2, corresponding to a plurality of (here, three) objects. The case where BJ3 (refer FIG. 2) is extracted with the area | region extraction apparatus 1 is illustrated. Specifically, areas BJ1, BJ2, and BJ3 corresponding to the hip bone and kidney are extracted as object areas. FIG. 2 is a schematic diagram showing a part of the human body (near the waist).

  Specifically, first, among the plurality of object regions BJ1, BJ2, and BJ3, the object region (region corresponding to the hip bone) BJ1 (see FIG. 2) is extracted by the model fitting method, and then the other objects. An example is shown in which regions (specifically, regions corresponding to two left and right kidneys) BJ2 and BJ3 are extracted by the model fitting method. Corresponding standard models MS1, MS2, and MS3 (see FIG. 15) are used for extracting the object areas BJ1, BJ2, and BJ3, respectively.

  In order to distinguish the standard models before and after deformation at the time of model fitting, for each standard model MSi (i = 1,..., N; N is the number of standard models), the standard model before deformation is MPi, It is also expressed as a standard model MQi after deformation. For example, for the standard model MS1, the standard model before deformation is expressed as MP1, and the standard model after deformation is expressed as MQ1 (see FIG. 16).

  Next, various functions of the region extraction device 1 will be described.

  FIG. 3 is a block diagram illustrating various functions of the region extraction device 1. These various functions are realized by executing a predetermined program PG using various hardware such as a CPU in the control unit 20.

  As shown in FIG. 3, the region extraction device 1 includes a model storage unit 31, a measurement data input unit 32, a temporary region extraction unit 33, an initial position determination unit 34, a model fitting unit 35, a relative relationship acquisition unit 36, and a model correction. A section 37 and an extraction area output section 38 are provided.

  The model storage unit 31 stores a plurality of standard models MSi relating to different objects. The plurality of standard models MSi are stored together with the relative positional relationship between them.

  The measurement data input unit 32 inputs three-dimensional measurement data (volume data) acquired by a CT apparatus or an MRI apparatus.

  The provisional region extraction unit 33 has a function of provisionally extracting a region (object) to be extracted using a method different from the model fitting method. In this embodiment, the candidate area (provisional area) of the object area BJ1 is extracted using the area expansion method.

  The initial position determination unit 34 has a function of determining the initial arrangement position of the standard model using the provisionally extracted object region. In this embodiment, the initial position determination unit 34 determines the initial position of the standard model MS1.

  The model fitting unit 35 has a function of extracting an object region by deforming a standard model using a model fitting method. The model fitting unit 35 includes a first model fitting unit 35a and a second model fitting unit 35b. The first model fitting unit 35a performs a model fitting process on the standard model MS1 serving as a reference to extract the object region BJ1, and the second model fitting unit 35b performs a model fitting process on the other standard models MS2 and MS3. This is executed to extract the object areas BJ2 and BJ3.

  The relative relationship acquisition unit 36 calculates the amount of scaling and / or movement of the standard model MS1 when the standard model MS1 is fitted to the object region BJ1 by the model fitting method, and the relative amount between the standard model MS1 and the object region BJ1. As a relation, it has a function to be sought.

  The model correcting unit 37 has a function of correcting the initial state (initial position and initial size) of the standard models MS2 and MS3 of the plurality of models based on the relative relationship obtained by the relative relationship acquiring unit 36. is doing. The second model fitting unit 35b executes the model fitting process using the standard models MS2 and MS3 corrected by the model correcting unit 37, and extracts the object areas BJ2 and BJ3.

  The extraction area output unit 38 has a function of displaying and outputting the extracted object areas BJ1, BJ2, and BJ3 on the monitor 3.

<B. Operation>
<B1. Outline of operation>
Next, the operation of the region extraction apparatus 1 described above will be described in detail.

  FIG. 4 is a flowchart showing the overall operation of the region extraction apparatus 1.

  As shown in FIG. 4, the region extraction apparatus 1 extracts the target specified by the user (user) from the input image (three-dimensional image) by executing the processes from step S1 to step S7. The extraction target is designated in advance by the user (user) before the process of step S1 is started. As the designation method, for example, an operation of the operation unit (for example, mouse) 4 or the like from an item list (for example, an item list in which various internal organs, bones, and the like are enumerated) stored in the storage unit 22 in a database is stored. Thus, it is possible to adopt a mode in which the user (user) selects a desired item. In the following, it is assumed that the hip bone and the left and right kidneys are designated as extraction objects.

  First, region extraction processing for extracting the corresponding target region BJ1 of one standard model MS1 is executed (steps S1, S2, S3). Specifically, a region BJ1 corresponding to the hip bone is extracted by the model fitting method. The initial position of the standard model MS1 in this first model fitting process is determined using the region expansion method (steps S1 and S2).

  Thereafter, a region extraction process is performed to extract the corresponding target regions BJ2 and BJ3 of the other standard models MS2 and MS3 (steps S4, S5, and S6). Specifically, regions BJ2 and BJ3 corresponding to the kidney are extracted by the model fitting method. However, in the model fitting process for extracting the regions BJ2 and BJ3, the initial positions and the like of the standard models MS2 and MS3 are determined in a simpler manner (steps S4 and S5) than in steps S1 and S2. Specifically, the initial states of the other standard models MS2 and MS3 are changed based on the magnification (magnification) and movement (translational displacement and rotational displacement) of the standard model MS1 in the region extraction operation (step S3). Is done.

  In step S7, the extracted object areas BJ1, BJ2, and BJ3 are displayed and output on the monitor 3. In addition, data representing the extraction result is output to and stored in the memory card 51 and the storage unit 22 so that they can be easily used thereafter.

  Hereinafter, the region extraction processing (steps S1, S2, S3) of the target region BJ1 and the region extraction processing (steps S4, S5, S6) of the other target regions BJ2, BJ3 will be described in detail.

<B2. Extraction Process for First Object Area BJ1>
First, the region extraction process regarding the first object region BJ1 will be described.

  As described above, the object region BJ1 is finally extracted by the model fitting method (step S3). However, when determining the initial position of the standard model in the model fitting method, a method different from the model fitting method (specifically, a region expansion method) is used (steps S1 and S2). Specifically, the object area BJ1 is provisionally extracted from the three-dimensional image using the area expansion method (step S1), and the initial position of the standard model is obtained using the object area BJ1 temporarily extracted. (Initial position when the standard model is arranged in the measurement data) is determined, and the standard model is arranged (step S2). Thereafter, the object region BJ1 is extracted from the measurement data by applying the model fitting method using the standard model MS1 moved to the initial position (step S3).

  Hereinafter, each of steps S1 to S3 will be described in detail.

<Provisional area extraction processing>
In step S1, an area (object area) corresponding to the object designated by the user (user) is provisionally extracted from the input three-dimensional image (volume data).

  As described above, the provisional area extraction process (step S1) aims to tentatively extract an object designated by a user (user).

  FIG. 5 is a flowchart showing details of provisional region extraction processing (step S1) using the region expansion method. FIG. 6 is a diagram showing a three-dimensional image TDP1 obtained by imaging a human body with an X-ray CT apparatus and a slice image PM at a predetermined position among a plurality of slice images SD constituting the three-dimensional image TDP1. As shown in FIG. 6, the three-dimensional image acquired by the X-ray CT apparatus is composed of a plurality of (for example, several tens to several hundreds) slice images SD showing a cross-section of a human body. Yes. Each slice image SD is an image visualized by displaying the amount of X-ray absorption (CT value) at each point (each pixel) in grayscale.

  FIG. 7 is an enlarged view of the vicinity of adjacent pixels centered on the pixel M1 on the slice image PM shown in FIG. In FIG. 7, as adjacent pixels of the pixel M1 of the slice image PM, in addition to the adjacent 8 pixels M2 to M9 in the slice image PM of the same hierarchy, the adjacent 9 pixels U1 to U9 in the slice image PU and the adjacent in the slice image PD Nine pixels D1 to D9 are shown. The slice image PU is a slice image adjacent to the slice image PM in the + z direction (in other words, a slice image immediately above the slice image PM), and the slice image PD is −z with respect to the slice image PM. It is a slice image adjacent in the direction (in other words, a slice image immediately below the slice image PM).

  In the following, the provisional region extraction processing using the region expansion method will be described with a specific example of the case where the target region BJ1 displayed in the slice image PM of FIG. 6 is provisionally extracted.

  First, in step S11, the expansion start pixel in the object area is specified by the user (user). Specifically, on an arbitrary slice image (here, slice image PM) on which an object (object region) to be extracted is displayed, an arbitrary operation in the target region is performed by operating the operation unit (eg, mouse) 4 or the like. Identify points. More specifically, when it is desired to extract the object area BJ1, an arbitrary point (for example, M1) in the object area BJ1 on the slice image PM may be specified with a mouse or the like (see FIG. 6). By this operation, the pixel corresponding to the specific point M1 on the slice image PM becomes the expansion start pixel.

  Next, in step S12, a pixel adjacent to the expansion start pixel is detected as an expansion candidate pixel. Specifically, if the expansion start pixel is specified as M1 in step S11, all the pixels adjacent to the pixel M1 are the expansion candidate pixels. That is, a total of 26 pixels, that is, the pixels M2 to M9 on the slice image PM, the pixels U1 to U9 on the slice image PU, and the pixels D1 to D9 on the slice image PD are extension candidate pixels.

  Next, in step S13, it is determined whether or not an extended candidate pixel is detected in step S12. If an extended candidate pixel is detected in step S12, the process proceeds to step S14.

  In step S14, it is determined whether or not the density value of the detected extension candidate pixel is within a predetermined range. Specifically, the density value range of the object (such as the hip bone) is predetermined for each object, and the density value of the expansion candidate pixel is within the density value range of the extraction target object (target area). Whether or not is determined using the upper and lower thresholds. Accordingly, when it is determined that the expansion candidate pixel has a density value within a predetermined range, the process proceeds to step S15.

  In step S15, region expansion is performed using the expansion candidate pixel as a pixel in the extraction target object.

  On the other hand, if it is determined in step S14 that the extension candidate pixel does not have a density value within the predetermined range, the area extension of the extension candidate pixel is not performed.

  Next, in step S16, it is determined whether or not there is an extension candidate pixel that has not passed through the extension candidate pixel density determination step (step S14).

  If there is an extension candidate pixel that has not undergone the density determination step (step S14), the processes of steps S14 and S16 are performed on the unfinished extension candidate pixel.

  On the other hand, if the density determination step (step S14) has been completed for all the expansion candidate pixels, the process returns to step S12, and the pixels expanded in step S15 (in other words, the pixels in the region). A pixel that is further adjacent to the specified pixel) and has not yet been determined whether or not it exists in the region is detected as a new expansion candidate pixel. Thereafter, the above-described steps S13 to S16 are performed on the newly detected extended candidate pixel. Steps S12 to S16 are repeated as much as possible to detect a new extension candidate pixel. When a new extension candidate pixel cannot be detected, the provisional region extraction process ends (step S13). By such an expansion process, the temporary extraction area is sequentially expanded to the upper hierarchical image and the lower hierarchical image, and further expanded outward in the same hierarchical image.

  As described above, the steps S12 to S16 are repeatedly executed, so that the region expansion is gradually performed three-dimensionally from the expansion start pixel M1, and the object designated by the user (user) is provisionally extracted. .

<Initial position determination process>
Next, in step S2, the initial position of the standard model MS1 is determined using the object region temporarily extracted in step S1. The initial position process (step S2) will be described with reference to FIGS.

  FIG. 8 is a flowchart showing details of the initial position determination process (step S2) using the barycentric point. FIG. 9 is a diagram showing the object region BJ1 (object OBa) included in the three-dimensional image (measurement data), and FIG. 10 is a diagram conceptually showing how the standard model MS1 is moved to the initial position. It is. FIG. 10 also shows the positional relationship between the object region and the standard model when the standard model is initially arranged. In FIG. 9 and FIG. 10 and the like, the object region BJ1 (object OBa) and the corresponding standard model BJ1 are simplified (or abstracted) as appropriate for convenience of illustration.

  First, in step S21, the center of gravity GZ of the object OBa temporarily extracted in step S1 is calculated.

  Next, in step S22, the standard model is initially placed so that the center point GH of the standard model used in model fitting described later matches the center point GZ calculated in step S21. For example, as shown in FIG. 10, the standard model MS1 is moved so that the center of gravity GH of the standard model MS1 matches the center of gravity GZ of the extraction target object OBa, and the standard model MS1 is initially arranged in the measurement data.

  When the initial placement of the standard model using the barycentric point is completed, the process proceeds to model fitting processing (step S3).

<Model fitting process>
The model fitting process in step S3 uses a “standard model” that is a model of a general (standard) object area prepared in advance and uses information (shape, etc.) obtained from the object area in the measurement data. It is a process to deform. In the present application, the standard model after the deformation by the model fitting process (in other words, the standard model reflecting the object area information) is also referred to as “individual model”.

  In step S3, the object region is extracted by deforming the standard model MS1 arranged at the initial position by the model fitting method.

  FIG. 11 is a flowchart showing details of the model fitting process (step S3). FIG. 12 is a schematic diagram in which the vicinity of the region RQ (see FIG. 10) centered on the point Q1 on the contour of the extraction target object OBa is enlarged, and FIG. 13 shows some slice images PQ1, PQ2, and PQ4. FIG. FIG. 14 is a schematic diagram in which control points are connected by virtual springs.

  Note that the standard model of the extraction target object (target object region) is composed of, for example, a minute polygon (for example, a triangle) and is stored in the storage unit 22 or the like. The standard model composed of polygons can express the surface shape of the model by the three-dimensional coordinates of the vertices of each polygon. In addition, of the vertices of the polygon in the standard model, all vertices or some representative vertices are also referred to as “control points”.

  Hereinafter, the model fitting process will be described in detail with reference to FIGS.

  First, in step S31, processing for detecting a point corresponding to the control point Cj of the standard model MS1 (hereinafter also referred to as “corresponding point”) from the three-dimensional image TDP1 is performed (see FIG. 11). Here, a method is used in which a pixel existing nearest to the control point Cj among the pixels (boundary points) indicating the outline (boundary) of the extraction target object OBa is a corresponding point of the control point Cj. Note that the outline (boundary) of the extraction target object OBa is acquired by performing an edge extraction process in each slice image SD.

  For example, in FIG. 12 (elevation), the corresponding point of the control point C1 is a boundary point (pixel) Q1 that is present nearest to the control point C1. Here, in FIG. 12, only boundary points in a predetermined plane (two-dimensional space) parallel to the xz plane are shown, but actually, the control point C1 in the three-dimensional space as shown in FIG. The boundary point existing closest to is a corresponding point. In FIG. 13, the projection points in the slice images PQ1, PQ2, and PQ4 of the control point C1 are indicated by crosses.

  Next, in step S32, an arbitrary one of the control points Cj of the standard model (hereinafter also referred to as “movement target point”) (for example, the control point C1) is moved by a minute amount L in one direction (for example, the A1 direction). (See FIG. 12).

  Further, in step S33, the total energy Ue of the model (hereinafter also referred to as “temporary deformation model”) in a state where the movement target point is moved and temporarily deformed in step S32 is calculated.

  The total energy Ue is represented by the sum of the external energy Fe relating to the distance between the control point Cj and the corresponding point and the internal energy Ge for avoiding excessive deformation, as shown in the equation (1). External energy Fe and internal energy Ge will be described later.

  Next, in step S34, it is determined whether or not the movement target point has been moved in all directions. For example, if the movement direction of the movement target point in all directions in the three-dimensional space is 26 directions (direction toward 26 pixels adjacent to the movement target point (pixel)), the control point C1 is moved in all 26 directions. It is determined whether or not.

  If the creation of the temporary deformation model in which the movement target point (control point C1) is moved in all 26 directions has not been completed, the movement direction of the movement target point is changed and moved again by a small amount L, and different movement direction patterns are obtained. A temporary deformation model is created and the total energy Ue of each temporary deformation model is calculated (steps S32 and S33).

  If it is determined in step S34 that the movement in all directions has been completed, the process proceeds to step S35.

  In step S35, a temporary deformation model that minimizes the total energy Ue is selected from the generated temporary deformation models of all patterns. In other words, a temporary deformation model that minimizes the total energy Ue is selected from the temporary deformation models generated by moving a certain control point Cj in 26 directions. For example, according to the action of internal energy Ge and external energy Fe as will be described later, when only external energy Fe is considered without considering internal energy Ge, the control point moves in a direction approaching the corresponding point. The temporary deformation model to be selected is selected. However, when considering the internal energy Ge, a temporary deformation model in which the control point moves in a direction different from the direction approaching the corresponding point may be selected.

  Next, in step S36, it is determined whether or not all the control points Cj have been moved. Specifically, it is determined whether or not the movement of the minute amount L has been completed for all the control points Cj of the standard model, and if there is an unfinished control point (also referred to as an incomplete point), the movement target point Is changed to the control point (incomplete point), and the operations in steps S32 to S35 are repeated to create a temporary deformation model in which the movement of all the control points Cj is completed. On the other hand, if the movement of all the control points has been completed, the process proceeds to step S37.

  In step S37, it is determined whether or not to end the model fitting process. Specifically, when the condition is satisfied on condition that the average value of the distances between the plurality of control points and the corresponding points in the temporary deformation model in which the movement of all the control points Cj is completed is equal to or less than a predetermined value In addition, the model fitting process may be terminated. According to this, when each control point approaches the corresponding point to a predetermined extent, the same process can be ended. Alternatively, or in addition to this, the amount of change in the total energy Ue between the previous temporary deformation model (which has completed the movement of all control points Cj) and the current temporary deformation model is equal to or less than a predetermined amount. Whether or not it may be used as a criterion for end determination. According to this, a process can be complete | finished when the total energy does not change so much even if a control point is moved.

  If it is determined in step S37 that the model fitting process is not finished, the process proceeds to step S38.

  In step S38, it is determined whether the unit processing loop has been executed a predetermined number of times W (for example, 10 times) using the processing executed in steps S32 to S37 described above as a unit processing loop. If the predetermined number of times W is not executed, the processing loop of steps S32 to S37 is repeated again until the predetermined number of times W is executed. If the predetermined number of times W is executed, the process proceeds to step S39. That is, the unit processing loop is repeated until a temporary deformation model that completes the movement of all control points is created W a predetermined number of times.

  In step S39, the corresponding points determined in step S31 are updated. Specifically, the corresponding point is updated with the pixel (boundary point) closest to each control point having moved W a predetermined number of times by the above processing loop as a new corresponding point for each control point, and the step is again performed. The processes of S32 to S39 are repeated. According to such “update of corresponding point”, it is possible to optimize the corresponding point even when the nearest boundary point of the control point changes with the movement of the control point.

  On the other hand, if it is determined in step S37 that the model fitting process is to be terminated, the finally obtained temporary deformation model is determined as an individual model corresponding to the extraction target object, and the process in step S3 is terminated.

  In such a model fitting process (step S3), an individual model corresponding to the extraction target region is created by gradually deforming the standard model by a small amount L for each control point.

  Here, the external energy Fe and the internal energy Ge constituting the total energy Ue will be described.

  The external energy Fe is expressed as in Expression (2) using the square of the distance between each control point Cj and the corresponding point Qj corresponding to each control point Cj.

  Here, α is a constant, Nt is the number of control points, and | Cj−Qj | is the distance between the control point Cj and the corresponding point Qj.

  Since the temporary deformation model in which the external energy Fe becomes very large, that is, the temporary deformation model in which the distance between the control point Cj and the corresponding point Qj is larger than that before the movement, the total energy Ue becomes large. It becomes difficult to be employed in the above-described step S35 (step of employing movement of the control point that minimizes the total energy Ue).

  Assuming that the internal energy Ge is connected to the control points Cj by virtual springs SPR (SPR1, SPR2, SPR3,...) As shown in FIG. It is expressed as follows.

  Here, β is a constant, K is a spring coefficient of a virtual spring, Nh is the number of virtual springs, and w is a displacement amount of each virtual spring.

  According to Equation (3), excessive movement of each control point Cj is expressed as an increase in energy stored in the virtual spring SPR. For example, if one control point Cz moves to a certain point Vz and the relative displacement with other control points increases, energy from the extension of each virtual spring is stored in the virtual springs SPR1, SPR2, and SPR3. The internal energy Ge and thus the total energy Ue increases.

  In such a temporary deformation model with excessive deformation, the internal energy Ge increases and the total energy Ue also increases. Therefore, in the above-described step S35 (step of adopting the movement of the control point that minimizes the total energy Ue). It becomes difficult to be adopted.

  In other words, the temporary deformation model that reduces the internal energy Ge and decreases the total energy Ue is selected in step S35, thereby obtaining an effect of preventing excessive deformation due to movement of each control point Cj. it can.

  That is, by introducing such internal energy Ge, the control point Cj can be moved without impairing the shape of the standard model, that is, the shape of each polygon constituting the standard model.

  According to the above-described processing, the temporary deformation model that minimizes the total energy Ue expressed by the sum of the external energy Fe and the internal energy Ge is a temporary deformation model that optimally deforms the standard model (“optimum deformation”). Also referred to as “model”), that is, determined as an individual model. And the area | region which exists in the corresponding position of the said individual model is extracted as an extraction object area | region. According to this, an appropriate region extraction process is realized.

<B3. Extraction Processing for Other Object Areas BJ2 and BJ3>
Next, an area extraction process regarding other object areas BJ2 and BJ3 will be described.

  As described above, the region extraction operation for one standard model MS1 among the plurality of standard models MS1, MS2, and MS3 is executed to extract the object region BJ1 (steps S1, S2, and S3). The area extraction operation is performed on the other standard models MS2 and MS3 by applying the model fitting method, and the object areas BJ2 and BJ3 are extracted.

  By the way, when extracting the object areas BJ2 and BJ3, it is also assumed that the same initial alignment process (steps S1 and S2) as described above using the area expansion method as described above is performed.

  However, since the initial position determination process using the above-described region expansion method has a relatively large amount of calculation, if the same initial position determination process is repeatedly performed for a plurality of standard models, the processing load increases. There is a problem that it ends up. Even when a method other than the region expansion method is used, the same problem occurs when the individual initial position determination processing for each standard model has a relatively large processing load.

  Therefore, in this embodiment, when the model fitting method is applied to the other standard models MS2 and MS3, the amount of magnification (magnification) and the amount of movement (translational displacement and translation) of the standard model MS1 in the region extraction operation (step S3). A method of changing the initial state (initial position and initial size) of the other standard models MS2 and MS3 based on (rotational displacement) will be described. According to such a method, the initial states of the standard models MS2 and MS3 can be set relatively easily (efficiently), and the object regions BJ2 and BJ3 can be extracted by model fitting.

Therefore, first, in step S4, the relative relationship (movement amount and scaling amount) between the standard model MP1 before deformation in step S3 and the standard model MQ1 after deformation is obtained. In detail,
(1) Change in displacement between standard model MP1 before deformation and standard model MQ1 after deformation (more specifically, the amount of movement of standard model MS1 when the model fitting technique is applied (step S3)), and (2) deformation A change in size between the previous standard model MP1 and the deformed standard model MQ1 (more specifically, the amount of magnification of the standard model MS1 when the model fitting technique is applied (step S3)) is obtained. Since the standard model MQ1 after the deformation can be regarded as equivalent to the object area BJ1, the relative relationship between the standard model MP1 and the standard model MQ1 after the deformation is obtained by comparing the standard model MP1 before the deformation and the object. This is equivalent to obtaining a relative relationship with the region BJ1.

  In step S5, the initial positions and initial sizes of the other standard models MS2 and MS3 when extracting the object areas BJ2 and BJ3 based on this relative relationship (specifically, the moving amount and the zooming amount). Correct the size.

  Here, with the initial position determination operation (step S2) of the first standard model MS1, the other standard models MS2 and MS3 are assumed to have been subjected to the same position change as the standard model MS1. The changed standard models MS2 and MS3 are also expressed as MP2 and MP3, respectively. In other words, in step S2 prior to step S5, the three standard models MS1 to MS3 are moved while maintaining the initial relative positional relationship and become standard models MP1 to MP3 (see FIG. 15). . As will be described later, the standard model MP2 is corrected to the standard model MF2 and the standard model MP3 is corrected to the standard model MF3 by the correction operation in step S5 (see FIG. 16).

  FIG. 15 is a conceptual diagram showing the standard model MS1 (MP1) and the like before the deformation process (step S3) by model fitting. FIG. 16 is a conceptual diagram showing the standard model MP1 before the deformation process, the standard model MQ1 after the deformation process, and the like. In FIG. 16, the standard model MP1 before the deformation process (step S3) is indicated by a broken line, and the standard model MQ1 after the deformation process (step S3) is indicated by a solid line.

  The deformation (MP1 → MQ1) of the standard model MS1 in step S3 is roughly divided into a change in displacement and a change in size. More specifically, such a standard model deformation includes three elements: a translation element, a rotation element, and a scaling element. Hereinafter, these elements and operations for correcting the elements will be described with reference to FIGS. 17 to 19.

FIG. 17 is a diagram showing a state in which there is a translational displacement represented by a translational displacement vector AX = (x, y, z) ^T (T means transposition) between the standard models MP1 and MQ1 before and after deformation. It is. The vector AX is a vector representing the translational displacement between the center of gravity position GP1 before deformation of the standard model MS1 and the center of gravity position GQ1 after deformation. When such a translational displacement exists, the other standard models MS2 (MP2) and MS3 (MP3) are moved by applying the same displacement amount as the translational displacement vector AX. As a result, the standard model MP2 is corrected to the standard model MF2, and the standard model MP3 is corrected to the standard model MF3. In other words, the standard models MP2 and MP3 are translated.

FIG. 18 is a diagram showing a state in which there is a rotational displacement represented by a rotational displacement vector Θ = (θ, φ, ψ) ^T between the standard models MP1 and MQ1 before and after the deformation. The vector Θ is a vector representing a rotational displacement around a predetermined center of rotation (here, a case where it coincides with the center of gravity is illustrated). When such rotational displacement exists, the other standard models MS2 (MP2) and MS3 (MP3) are moved around the rotational center by applying the rotational displacement. As a result, the standard model MP2 is corrected to the standard model MF2, and the standard model MP3 is moved to the standard model MF3. In other words, the standard models MP2 and MP3 are rotationally moved.

  FIG. 19 is a diagram illustrating a state in which there is a zoom amount indicated by the zoom factor α between the standard models MP1 and MQ1 before and after the deformation. The scaling factor α is a value obtained by dividing the size of the standard model MQ1 after deformation by the size of the standard model MP1 before deformation. When the scaling factor α is larger than 1, it means that the standard model MS1 is enlarged and deformed, and when the scaling factor α is smaller than 1, it means that the standard model MS1 is reduced and deformed.

  If there is a relationship expressed by the scaling factor α, the other standard models MS2 (MP2) and MS3 (MP3) are scaled (enlarged or reduced) by the scaling factor α, and the standard model MP1 The standard model MP2 is modified to the standard model MF2 by moving the distance between the standard model MP2 and the standard model MP1 and the standard model MP3 so that the distance between the standard model MP1 and the standard model MP3 is α, and the standard model MP3 is the standard model MF3. To be corrected. In other words, the standard models MP2 and MP3 are enlarged or reduced, and moved according to the magnification.

  Reference is again made to FIG. Generally speaking, the deformation (MP1 → MQ1) of the standard model MS1 in step S3 is a deformation obtained by combining various deformations as shown in FIGS. By appropriately performing various modifications as described above on the standard models MP2 and MP3, the standard models MP2 and MP3 become standard models MF2 and MF3 as shown in FIG. In FIG. 16, the standard models MP2 and MP3 before the correction process (step S5) are represented by broken lines, and the standard models MF2 and MF3 after the correction process are represented by solid lines.

  More specifically, first, an affine transformation as shown in FIG. 16 for transforming the standard model MP1 before the transformation into the standard model MQ1 after the transformation is obtained. For example, some (plural) of the control points of the standard model MS1 can be selected as “representative control points”, and the affine transformation matrix can be obtained based on the coordinate values before and after the deformation of the representative control points. Then, the standard models MP2 and MP3 are corrected using the affine transformation. According to this, it is possible to obtain the standard models MF2 and MF3 after being subjected to the combined transformation of scaling (enlargement or reduction), rotational movement, and parallel movement as described above.

  Thereafter, in step S6, model fitting processing is executed using the standard models MF2 and MF3 after the correction processing. The process in step S6 is the same as the process in step S3. In the process of step S6 (model fitting process), the standard model MF2 is transformed into the standard model MQ2 (see FIG. 16), and the standard model MF3 is transformed into the standard model MQ3. The standard models MQ2 and MQ3 (shown by two-dot chain lines in FIG. 16) after model fitting (after deformation) are based on the fact that they can be regarded as equivalent to the object regions BJ2 and BJ3, respectively. The existence area of the model MQ2 is extracted as the object area BJ2, and the existence area of the standard model MQ3 after the deformation is extracted as the object area BJ3.

  In this way, the object areas BJ2 and BJ3 corresponding to the standard models MS2 and MS3, respectively, are extracted from the three-dimensional measurement data.

  With the above operation, the three object regions BJ1, BJ2, and BJ3 can be extracted using the model fitting method.

  In the above embodiment, prior to the operation of extracting the object areas BJ2 and BJ3 corresponding to the standard models MS2 and MS3 using the model fitting technique, based on the relative relationship between the standard model MS1 and the object area BJ1. Other standard models MS2 and MS3 are modified. Accordingly, initial alignment and the like for the standard models MS2 and MS3 can be easily realized. For example, with respect to the standard models MS2 and MS3, initial alignment using the region expansion method is unnecessary and efficient.

  In particular, since initial alignment and the like for a plurality of standard models (for example, MS2 and MS3) other than the first standard model can be performed collectively using relative relationships, initial expansion using the extended region method or the like can be performed individually. It is more efficient than the case where alignment is performed.

  Further, in particular, by correcting the initial sizes of the standard models MS2 and MS3 in the model fitting process, the standard model is changed from a state having a size relatively close to the original object regions BJ2 and BJ3 in the three-dimensional measurement data. Since the process of fitting MS2 and MS3 to the object regions BJ2 and BJ3 can be started, the extraction accuracy in the model fitting process can be improved.

<C. Modification>
Although the embodiments of the present invention have been described above, the present invention is not limited to the contents described above.

  For example, in the above, the case where the initial alignment for the standard model MS1 is performed using the region expansion method (steps S1, S2, etc. in FIG. 4) is illustrated, but the present invention is not limited to this. Specifically, a specific feature point (for example, the tip of a sharp portion) in the object is extracted in advance from the three-dimensional measurement data (three-dimensional image), and the standard feature MS1 and the three-dimensional measurement data are extracted. The initial position of the standard model MS1 may be corrected so as to correct the “deviation” regarding the feature point.

  In the above embodiment, both the amount of scaling and the amount of movement of the standard model MP1 when fitting the standard model MP1 to the object region BJ1 are relative to each other (standard model MP1 and object region BJ1). However, the present invention is not limited to this. For example, only the movement amount of the standard model MP1 when fitting the standard model MP1 to the object region BJ1 may be obtained as a relative relationship between them, or the standard when fitting the standard model MP1 to the object region BJ1. Only the scaling amount of the model MP1 may be obtained as a relative relationship between them.

  Furthermore, in the above embodiment, the case where both the initial position and the initial size of the standard models MP2 and MP3 are corrected when the object regions BJ2 and BJ3 are extracted by applying the model fitting method is illustrated. It is not limited to this. For example, only the initial positions of the standard models MP2 and MP3 may be corrected, or only the initial sizes of the standard models MP2 and MP3 may be corrected.

  Moreover, in the said embodiment, although the case where a target object area | region was extracted from the three-dimensional image acquired by the X-ray CT apparatus was illustrated, it is not limited to this, From the image acquired by other apparatuses, such as an MRI apparatus, You may make it extract a target object area | region.

  Moreover, in the said embodiment, although the case where the measurement data stored in the memory card 51 is read is assumed, it is not limited to this, Measurement data stored in another storage device (for example, server computer) is networked. It is also possible to receive and acquire the information via, for example.

It is a figure which shows the outline | summary of the area | region extraction apparatus which concerns on embodiment. It is a schematic diagram which shows a part of human body (near the waist | hip | lumbar part). It is a block diagram which shows the various functions of an area | region extraction apparatus. It is a flowchart which shows the whole operation | movement of an area | region extraction apparatus. It is a flowchart which shows the temporary area | region extraction process using an area | region expansion method. It is a figure which shows the three-dimensional image of the human body by X-ray CT apparatus. It is an enlarged view near pixel M1 on a predetermined slice image. It is a flowchart which shows the initial position determination process using a gravity center point. It is a figure which shows the object contained in a three-dimensional image. It is a flowchart which shows a model fitting process. It is a figure which shows a mode that a standard model is moved to an initial position. It is a figure which shows the relationship between the standard model initially arranged, and an extraction object. It is a figure which shows the slice images PQ1, PQ2, and PQ4. It is the schematic diagram which connected between control points with the virtual spring. It is a figure which shows standard model MS1 (MP1) etc. before the deformation | transformation process (step S3) by model fitting. It is a figure which shows each standard model before and behind a deformation | transformation process. It is a figure which shows the translation element in a correction process. It is a figure which shows the rotation element in a correction process. It is a figure which shows the variable magnification element in a correction process.

Explanation of symbols

1 region extraction device 35 model fitting unit Θ rotational displacement vector α scaling factor AX translational displacement vector BJ1, BJ2, BJ3 object region Cj, Ck, Cz control point MF2, MF3, MPi, MQi, MSi standard model PD, PM, PQ1 , PQ2, PQ4, PU, SD Slice image Qj Corresponding point TDP1 3D image

Claims (5)

An area extraction apparatus for extracting a plurality of object areas from measurement data,
Means for storing a plurality of standard models for different objects;
Means for extracting a first object region of the plurality of object regions from the measurement data by a model fitting method using a first standard model of the plurality of standard models;
When the first standard model is fitted to the first object region, the amount of scaling and / or movement of the first standard model is expressed as the first standard model and the first object region. A means for obtaining a relative relationship with
Correction means for correcting an initial state of a second standard model of the plurality of models based on the relative relationship;
Means for extracting a second object region of the plurality of object regions from the measurement data by a model fitting method using the second standard model after correction;
An area extracting apparatus comprising: The region extraction device according to claim 1, wherein
The area extracting apparatus, wherein the correcting means corrects an initial position of the second standard model. In the area extraction device according to claim 1 or 2,
The region extracting apparatus characterized in that the correcting means corrects an initial size of the second standard model. An area extraction method for extracting a plurality of object areas from measurement data,
Extracting a first object region of the plurality of object regions from the measurement data by a model fitting method using a first standard model of a plurality of standard models for different objects;
When the first standard model is fitted to the first object region, the amount of scaling and / or movement of the first standard model is expressed as the first standard model and the first object region. A step for obtaining a relative relationship with
Modifying an initial state of a second standard model of the plurality of models based on the relative relationship;
Extracting the second object region of the plurality of object regions from the measurement data by a model fitting technique using the second standard model after correction;
A region extraction method comprising: On the computer,
An area extraction method for extracting a plurality of object areas from measurement data,
Extracting a first object region of the plurality of object regions from the measurement data by a model fitting method using a first standard model of a plurality of standard models for different objects;
When the first standard model is fitted to the first object region, the amount of scaling and / or movement of the first standard model is expressed as the first standard model and the first object region. A step for obtaining a relative relationship with
Modifying an initial state of a second standard model of the plurality of models based on the relative relationship;
Extracting the second object region of the plurality of object regions from the measurement data by a model fitting technique using the second standard model after correction;
A program for executing a region extraction method provided.

Images