JP4909792B2 - Image interpretation support apparatus, method, and program - Google Patents

Description

  The present invention relates to an apparatus and method for supporting interpretation of a bent luminal structure in a three-dimensional image, and a program for causing a computer to execute the method.

  In the field of three-dimensional medical image processing, a technique for generating an endoscopic image by performing central projection from a viewpoint on a tube core line of a bent luminal organ is known (for example, Patent Documents 1, 2, and 3). 4). In addition, a method of generating a stretched and unfolded image in which a luminal organ is linearly expanded and further the tube is opened to expand the inner wall side (for example, Patent Document 4), the luminal organ is linearly expanded, A technique (for example, Patent Document 1) that generates an elongated cross-sectional image that is obtained by cutting a pipe vertically is known.

  As a specific display method for these images, an endoscopic image in which a longitudinal section image (expanded section image) of a blood vessel and a tube core line direction formed by the central projection method are centered on a CRT monitor is projected. The image is displayed at the same time, and the viewpoint position of the endoscopic image is superimposed on the longitudinal cross-sectional image with an arrow, and the arrow is moved to another viewpoint with the mouse, so that the endoscopic image is displayed. A method of switching and displaying has been proposed (for example, Patent Document 1).

  In addition, a straightened colon image (elongated cross-sectional image) and a 3D image (internal view) of the colon centered on the tube core direction at a position selected along the straightened colon image. A method for displaying (mirror view) is proposed (for example, Patent Document 2).

  In addition, a straight view of the blood vessel (elongated cross-sectional image), a Perpendicular view with a cross section perpendicular to the tube core line of the blood vessel, and the tangent direction of the same cross section or tube core line as the Perpendicular view at the position specified in the straight view or Perpendicular view There has been proposed a method of creating and displaying an arbitrary cross-sectional image with a cross-section parallel to (for example, Patent Document 3).

Furthermore, a developed image (elongated developed image) obtained by projecting a tubular tissue by a bent cylindrical projection method and a method of superimposing and displaying surface information representing a correspondence relationship between positions and directions of both images on a central projected image of the tubular tissue. Has been proposed (for example, Patent Document 4).
Japanese Patent Laid-Open No. 11-318884 JP-T-2001-511031 JP 2004-283373 A JP 2006-65397 A

  The above endoscopic image is suitable for grasping the details of the local area inside the luminal organ, but cannot represent a wide area inside the luminal organ. It is not suitable for grasping the whole picture. On the other hand, the above-described expanded development image and expanded cross-sectional image are suitable for grasping the entire inside of the luminal organ, but since the bent luminal organ is linearly expanded, The shape of the organ is not preserved, and it is not suitable for observing in detail the part of interest inside the hollow organ.

  Therefore, in view of the structural characteristics of the luminal organ and the characteristics of the endoscopic image, the expanded expanded image, and the expanded cross-sectional image, when observing the image representing the inside of the luminal organ, first, the expanded expanded image It is considered that it is preferable that the workflow is such that when an overall observation is performed with an expanded cross-sectional image and a portion to be noticed is found, the details of the portion are observed with an endoscopic image.

  In contrast, in the methods described in Patent Documents 1 and 2 described above, the viewpoint position of the central projection image is determined from the specified position in the expanded cross-sectional image, but the generated central projection image projects the direction of the tube core line. The center projection image does not necessarily include a portion corresponding to the designated position in the expanded cross-sectional image.

  Similarly, in the method described in Patent Document 3, the cross-sectional position of the arbitrary cross-sectional image is determined from the designated position in the expanded cross-sectional image, but the same cross-section as the Perpendicular view at that position or the cross-section in the tangential direction of the tube core line. A section corresponding to the designated position is not necessarily included in the cross-sectional image, and even if it is included, it is not always displayed at a position suitable for observing the central portion of the arbitrary cross-sectional image.

  The method described in Patent Document 4 is a method for superimposing and displaying the mutual positional relationship between the development view and the cross-sectional view on each image, and the center projection image is displayed according to the designated position in the expanded development image. It does not suggest a workflow to generate.

  The present invention has been made in view of the above circumstances, and an overall observation is performed with an expanded development image or an expanded cross-sectional image representing a state in which a bent luminal structure is expanded, and attentions found in the image are found. An object of the present invention is to provide an image interpretation support apparatus, method, and program for more efficiently realizing a workflow of performing detailed observation with a projected image of a position.

In the first image interpretation support apparatus of the present invention, a bent luminal structure (TO) is represented as schematically shown in FIG. 1 and FIG. In each of a plurality of viewpoints (VP _m ; m = 1, 2,...) On the core line (CL) passing through the center of the structure (TO) in the three-dimensional image (V), the viewpoint (VP _m ), the line-of-sight vector for each of a plurality of line-of-sight vectors (R1 _mn ; m = 1, 2,..., n = 1, 2,...) starting from the center line (CL). (R1 _mn) using image information of the vicinity of the upper and / or visual axis vector (R1 _mn), the pixel value of the visual axis vector (R1 _mn) the structure is projected projected pixels on (PP _mn) Each of the projection pixels (PP _mn ) is determined so that the arrangement order of the viewpoints (VP _m ) and the arrangement order of the plurality of line-of-sight vectors (R1 _mn ) at the viewpoints (VP _m ) are maintained. Expansion expanded image generating means for generating an expanded expanded image (DV) arranged on the projection plane, the position (PP _mn ) of each projection pixel in the expanded expanded image (DV), and the three-dimensional image (V) Corresponding information storage means for storing correspondence information that can specify the correspondence relationship with each of the line-of-sight vectors (VP _m ), and position designation that accepts designation of the position of interest (for example, PP _uv ) in the decompressed expanded image (DV) Based on the correspondence information stored in the correspondence means and the correspondence information storage means, the line-of-sight vector (R1 _uv ) corresponding to the target position (PP _uv ) accepted by the position designation acceptance means is identified, and the identification is set to the viewpoint the starting point (VP _u) of the line-of-sight vector (R1 _uv), approximately in the center to center projection, or the start point of the specified viewing vector direction of projecting the direction of the visual axis vector (R1 _uv) ( The position of VP _u ) Then, by parallel projection with the direction of the line-of-sight vector (R1 _uv ) as the projection direction, the position in the three-dimensional image corresponding to the target position (PP _uv ) (arbitrary position on R1 _uv ) is approximately centered Projection image generation means for generating a projection image projected on the screen is provided.

  According to the first image interpretation support method of the present invention, the computer is configured such that at each of a plurality of viewpoints on a core line passing through the center of the structure in a three-dimensional image in which a bent tubular structure is represented, The structure on the line-of-sight vector is projected using image information on the line-of-sight vector and / or in the vicinity of the line-of-sight vector for each of a plurality of line-of-sight vectors starting from the viewpoint toward the peripheral direction of the core line A decompressed expanded image in which each of the projection pixels is arranged on a projection plane is determined so that a pixel value of the projection pixel is determined and the arrangement order of the viewpoints and the arrangement order of the plurality of line-of-sight vectors at the viewpoints are maintained. A process of generating, a process of storing correspondence information that can identify a correspondence relationship between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image, Attention And specifying the line-of-sight vector corresponding to the received position of interest based on the stored correspondence information, using the start point of the specified line-of-sight vector as a viewpoint, and the direction of the line-of-sight vector To the target position by central projection with the center of the projection direction or parallel projection with the position of the start point of the identified line-of-sight vector as the center of the viewpoint and the direction of the line-of-sight vector as the projection direction. And a process of generating a projection image in which the corresponding position in the three-dimensional image is projected substantially at the center.

  A first image interpretation support program according to the present invention causes a computer to execute the first image interpretation support method.

  Details of the first image interpretation support apparatus, method, and program of the present invention will be described below.

  Specific examples of the “bent luminal structure” include a luminal organ of a living body such as a large intestine.

  “A plurality of line-of-sight vectors from the viewpoint to the peripheral direction of the core line” exist radially from the viewpoint, and the range (center angle) of the radial direction is 120 degrees, 180 degrees, 360 degrees, etc. Is optional.

  Specific examples of “image information” include signal intensity (CT value, etc.), density value, luminance value, color signal value (R, G, B, etc.) obtained at the time of photographing / acquisition of the three-dimensional image, volume Examples include opacity used in rendering processing.

  A process of “determining a pixel value of a projection pixel” and a process of “generating a projection image in which a position in the three-dimensional image corresponding to the target position is projected to substantially the center (by central projection / parallel projection)” As a specific example of this, there is a process of setting a plurality of search points along the line of sight and obtaining a pixel value of a projection pixel based on the set plurality of search points and / or image information in the vicinity thereof. For example, the average value of the pixel values of the search points, the MIP (Maximum Intensity Projection) process that extracts and projects the maximum value of the pixel values of the search points, and the minimum value are extracted. Ray casting to obtain the pixel value of the projection pixel by integrating Minpac (Minimum Intensity Projection) processing and projecting opacity (Opacity) and luminance values at each search point along the line of sight Volume rendering processing and the like by grayed method.

  As a specific example of the processing of “generating an expanded and developed image in which each projection pixel is arranged on the projection plane so that the arrangement order of each viewpoint and the arrangement order of a plurality of line-of-sight vectors at each viewpoint are maintained” A projection pixel for each line-of-sight vector in FIG. 3 is projected onto the projection plane as a single line by coordinate transformation, and a method of arranging the lines at each viewpoint on the projection plane in the order of viewpoints, or a short with the tangent at each viewpoint as an axis. Projection pixels for each line-of-sight vector at the viewpoint are arranged (projected) in the coordinate system of the cylindrical projection plane, the cylindrical projection planes are connected in the order of viewpoints, and the connected cylindrical projection plane is further converted into a projection plane by coordinate transformation. An expanding method (for example, see JP-A-8-249492) and the like can be mentioned.

  “Correspondence information that can specify the correspondence between each position in the expanded expanded image and each line-of-sight vector in the three-dimensional image” is the position of one or more projection pixels in the expanded expanded image and their projections. What is necessary is just to combine suitably the thing which linked | related the visual line vector corresponding to a pixel, and the information which linked | related the positional relationship of the up-down direction and / or left-right direction of the said expansion | deployment expansion image, and the positional relationship of the said viewpoint and a visual line vector. . Here, as a specific example of the latter, the distance from the first viewpoint to the last viewpoint on the core line, the distance between the viewpoints, the range of the radial direction of the line-of-sight vector (center angle), and the angle formed between adjacent line-of-sight vectors The angle etc. are mentioned.

  The process of “identifying the line-of-sight vector corresponding to the position of interest based on the correspondence information” is proportional to the above correspondence information, using the number of pixels in the vertical and horizontal directions, the image size, the pixel interval, etc. of the expanded expanded image as appropriate. This can be realized by performing calculation or the like.

  Note that the correspondence information is obtained by associating the positions of all the projection pixels in the decompressed expanded image with the line-of-sight vector corresponding to each projection pixel, and by referring to this, the line-of-sight vector corresponding to the target position is obtained. It may be specified.

In the second image interpretation support apparatus of the present invention, a bent luminal structure (TO) is represented as schematically shown in FIG. 1 and FIG. 2 as an example of the processing target and processing result images. In each of a plurality of viewpoints (VP _m ; m = 1, 2,...) On the core line (CL) passing through the center of the structure (TO) in the three-dimensional image (V), the viewpoint (VP _m ), the line-of-sight vector for each of a plurality of line-of-sight vectors (R1 _mn ; m = 1, 2,..., n = 1, 2,...) starting from the center line (CL). (R1 _mn) using image information of the vicinity of the upper and / or visual axis vector (R1 _mn), the pixel value of the visual axis vector (R1 _mn) the structure is projected projected pixels on (PP _mn) Each of the projection pixels (PP _mn ) is determined so that the arrangement order of the viewpoints (VP _m ) and the arrangement order of the plurality of line-of-sight vectors (R1 _mn ) at the viewpoints (VP _m ) are maintained. Expansion expanded image generating means for generating an expanded expanded image (DV) arranged on the projection plane, the position (PP _mn ) of each projection pixel in the expanded expanded image (DV), and the three-dimensional image (V) Corresponding information storage means for storing correspondence information that can specify the correspondence relationship with each of the line-of-sight vectors (VP _m ), and position designation that accepts designation of the position of interest (for example, PP _uv ) in the decompressed expanded image (DV) Based on the correspondence information stored in the correspondence means and the correspondence information storage means, the line-of-sight vector (R1 _uv ) corresponding to the target position (PP _uv ) received by the position designation reception means is specified, and the line-of-sight On the vector (R1 _uv ), the main projection position (for example, P _uv ) is high enough that the degree of contribution to the determination of the pixel value of the projection pixel (PP _uv ) corresponding to the line-of-sight vector (R1 _uv ) satisfies a predetermined criterion Determine the central projection or The line projection, characterized in that the main projection position (P _uv) is provided a projection image generating means for generating a projection image to be projected approximately in the center.

  In the second image interpretation support method of the present invention, the computer is configured such that at each of a plurality of viewpoints on a core line passing through the center of the structure in a three-dimensional image in which a bent tubular structure is represented, The structure on the line-of-sight vector is projected using image information on the line-of-sight vector and / or in the vicinity of the line-of-sight vector for each of a plurality of line-of-sight vectors starting from the viewpoint toward the peripheral direction of the core line A decompressed expanded image in which each of the projection pixels is arranged on a projection plane is determined so that a pixel value of the projection pixel is determined and the arrangement order of the viewpoints and the arrangement order of the plurality of line-of-sight vectors at the viewpoints are maintained. A process of generating, a process of storing correspondence information that can identify a correspondence relationship between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image, Attention Based on the received correspondence information and the stored correspondence information, the line-of-sight vector corresponding to the received position of interest is specified, and the pixel value of the projection pixel corresponding to the line-of-sight vector on the line-of-sight vector Determining a main projection position that is high enough to satisfy a predetermined criterion, and generating a projection image so that the main projection position is projected almost at the center by central projection or parallel projection. It is characterized by performing.

  A second image interpretation support program according to the present invention causes a computer to execute the second image interpretation support method.

  The details of the second image interpretation support apparatus / method / program of the present invention will be described below, focusing on the differences from the first image interpretation support apparatus / method / program.

  The method of “determining a main projection position on the line-of-sight vector that is high enough to satisfy a predetermined criterion for determining the pixel value of the projection pixel corresponding to the line-of-sight vector” determines the pixel value of the projection pixel It depends on the processing. For example, in the case of processing for obtaining the average value of pixel values of search points on the line-of-sight vector or MIP processing, the position of the search point at which the pixel value is maximum, and in MinIP processing, the search at which the pixel value is minimum. In the case of point position and volume rendering processing, the position of the search point with the maximum opacity can be set as the main projection position.

  In the method of “generating a projection image so that the main projection position is projected at approximately the center by central projection or parallel projection”, if the main projection position is projected at approximately the center, the viewpoint and line of sight at the time of projection The direction is arbitrary.

In the third image interpretation support device of the present invention, a bent luminal structure (TO) is represented as schematically shown in FIG. 3 and FIG. Each of a plurality of viewpoints (VP _mn ; m = 1, 2,..., N = 1, 2,...) Inside the structure (TO) in the three-dimensional image (V) is a starting point. , each viewpoint (VP _mn) each curved surface a plurality of line-of-sight vector directed in the circumferential direction of the one side of the (CS) (R2 _mn) through a visual axis vector (R2 _mn) above and / or visual axis vector (R2 _mn ) is used to determine the pixel value of the projection pixel (PP _mn ) on which the structure (TO) is projected on the line-of-sight vector (R2 _mn ), and the respective viewpoints (VP _mn ) as sorted is held in the extended cross-sectional image generation means for generating an extended cross-sectional image obtained by placing each of the projection pixels (PP _mn) on the projection plane (SV), wherein said extension section Correspondence information storage for storing a correspondence information which can identify the correspondence between the image the position of each projection pixel in (SV) (PP _mn) and the sight line vector of the 3-dimensional image (V) (R2 _mn) Based on the correspondence information stored in the correspondence information storage means, the position designation acceptance means for accepting designation of the position of interest (for example, PP _uv ) in the expanded cross-sectional image (V) accepted target position (PP _uv) to identify the line-of-sight vector corresponding (R2 _uv), on the visual axis vector (R2 _uv), projected pixel corresponding to the visual axis vector (R2 _uv) of (PP _uv) the degree of contribution to the determination of the pixel value determines the main projection position high enough to meet the predetermined criteria (e.g., P _uv), the central projection or parallel projection, the main projection position (P _uv) is projected approximately in the center Projection to generate projection images Characterized by providing the image generating means.

  According to a third image interpretation support method of the present invention, each of the viewpoints has a computer starting from each of a plurality of viewpoints inside the structure in the three-dimensional image in which the bent tubular structure is represented. The structure on the line-of-sight vector is projected using image information on the line-of-sight vector and / or in the vicinity of the line-of-sight vector for each of a plurality of line-of-sight vectors toward the peripheral direction on one side of the curved surface passing through Determining a pixel value of the projection pixel and generating an expanded cross-sectional image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the viewpoints is maintained; A process for storing correspondence information that can specify a correspondence relationship between a position of a projection pixel and each line-of-sight vector in the three-dimensional image; a process for receiving designation of a position of interest in the expanded cross-sectional image; Based on correspondence information A main projection that identifies the line-of-sight vector corresponding to the received position of interest and that has a high degree of contribution to the determination of the pixel value of the projection pixel corresponding to the line-of-sight vector on the line-of-sight vector satisfies a predetermined criterion A position is determined, and a projection image is generated by center projection or parallel projection so that the main projection position is projected almost at the center.

  A third image interpretation support program of the present invention causes a computer to execute the third image interpretation support method.

  The details of the third image interpretation support apparatus / method / program of the present invention will be described below, focusing on the differences from the first and second image interpretation support apparatuses / methods / programs.

  As a specific example of the “curved surface passing through each viewpoint”, a curved surface passing through a core line passing through the central portion of the tubular structure can be considered.

  As a specific example of the process of “generating an expanded cross-sectional image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the viewpoints is maintained”, the curved surface passes through the core line, When a plurality of viewpoints are set on a plurality of lines in a direction perpendicular to the core line on the curved surface, projection pixels corresponding to line-of-sight vectors at each viewpoint on one line on the curved surface are set for each line. , A method of projecting as a single line on the projection plane by coordinate transformation, and projecting onto the projection plane so that the order of the lines on the curved surface is maintained. Examples of the method are described.

  “Correspondence information that can specify the correspondence between each position in the expanded cross-sectional image and each line-of-sight vector in the three-dimensional image” is the position of one or more projection pixels in the expanded cross-sectional image and their projections. A combination of a line-of-sight vector corresponding to a pixel, and information that associates a positional relationship in the vertical and / or horizontal direction of the elongated cross-sectional image with a positional relationship of the viewpoint (start point) of the line-of-sight vector, as appropriate do it. Here, specific examples of the latter include the distance between the viewpoints located at both ends of the on-luminal structure, the interval between the viewpoints, and the like.

  The process of “identifying the line-of-sight vector corresponding to the position of interest based on the correspondence information” is proportional using the above correspondence information and the number of pixels in the vertical and horizontal directions of the expanded cross-sectional image, the image size, the pixel interval, and the like as appropriate. This can be realized by performing calculation or the like.

  The correspondence information is obtained by associating the positions of all the projection pixels in the expanded cross-sectional image with the line-of-sight vector corresponding to each projection pixel, and by referring to this, the line-of-sight vector corresponding to the target position is obtained. It may be specified.

In the method of “generating a projection image so that the main projection position is projected at approximately the center by central projection or parallel projection”, if the main projection position is projected at approximately the center, the viewpoint and line of sight at the time of projection The direction is arbitrary. For example, as schematically shown in FIG. 4, a point (Q) on the core line (CL) passing through the center of the structure (TO) closest to the main projection position (P _uv ) is used as a viewpoint, and the viewpoint A method of generating a projection image by central projection with the direction from (Q) toward the main projection position (P _uv ) as the approximate center of the projection direction, or with this point (Q) as the approximate center of the viewpoint, A method is conceivable in which a projection image is generated by parallel projection with the direction from Q) toward the main projection position (P _uv ) as the projection direction.

  According to the present invention, the position of each projection pixel in an expanded development image or an expanded cross-sectional image obtained by linearly extending a bent tubular structure, and each of the three-dimensional images in generating those images. By storing correspondence information that can identify the correspondence relationship with the line-of-sight vector, based on this correspondence information, the line-of-sight vector corresponding to the specified attention position is specified, and on the basis of the identified line-of-sight vector, Generate a projection image by central projection or parallel projection where the target position is the center of projection, or based on the identified line-of-sight vector, determine the main projection position that greatly contributes to the determination of the projection pixel in the line-of-sight vector, It is possible to generate a projection image by central projection or parallel projection in which the determined main projection position is the center of projection. Therefore, if the position of interest found in the image is specified after performing the overall observation with the expanded development image or the expanded cross-sectional image, the position of the position of interest in the three-dimensional image can be further specified. Instead, a projected image whose center of projection is the focus position or main projection position is automatically generated and can be observed. The workflow of details becomes more efficient.

  Hereinafter, a three-dimensional medical image processing system in which an image interpretation support apparatus according to an embodiment of the present invention is mounted will be described with reference to the drawings.

  FIG. 5 is a hardware configuration diagram showing an overview of a three-dimensional medical image processing system. As shown in the figure, in this system, a modality 1, an image storage server 2, and an image processing workstation 3 are connected via a network 9 in a communicable state.

  The modality 1 acquires a three-dimensional medical image (voxel data) V representing the subject, and specifically includes a CT apparatus, an MRI apparatus, an ultrasonic diagnostic apparatus, and the like.

  The image storage server 2 is a computer that stores and manages a three-dimensional medical image V acquired by the modality 1 and a medical image generated by image processing at the image processing workstation 3 in an image database. And database management software (for example, ORDB (Object Relational Database) management software).

  The image processing workstation 3 is a computer that performs image processing on the three-dimensional medical image V acquired from the modality 1 or the image storage server 2 in response to a request from the interpreter and displays the generated image. In particular, it includes an input device such as a keyboard and a mouse for inputting a request from a radiogram interpreter, a main memory having a capacity capable of storing the acquired three-dimensional medical image V, and a display for displaying the generated image. . The image interpretation support device of the present invention is mounted on the image processing workstation 3, and the processing in this device is realized by executing a program installed from a recording medium such as a CD-ROM. The program may be installed after being downloaded from a server connected via a network such as the Internet.

  The storage format of image data and communication between devices via the network 9 are based on a protocol such as DICOM (Digital Imaging and Communications in Medicine).

FIG. 6 is a block diagram showing the functions of the image interpretation support apparatus according to the first embodiment of the present invention of the image processing workstation 3. As shown in the figure, this apparatus includes a large intestine extraction unit 11, a core line extraction unit 12, an expanded and developed image generation unit 13, a correspondence information storage unit 14, a viewpoint setting reception unit 15, a position designation reception unit 16, and a projection image generation unit 17. The large intestine extraction unit 11 extracts the image V _LI representing the large intestine region from the inputted three-dimensional image V, and the core line extraction unit 12 extracts the core line CL passing through the central part of the V _LI tube of the large intestine. The expanded expanded image generation unit 13 extracts and generates an expanded expanded image DV based on the image V _LI representing the large intestine region, the viewpoint information VP received by the initial viewpoint position or the viewpoint setting reception unit 15, and the core line CL. In addition, the correspondence information PR (described later) is stored in the correspondence information storage unit 14, the position designation receiving unit 15 receives the designation of the position of interest POI in the decompressed expanded image DV, and the projection image generating unit 17 performs the expanded expanded image. Based on DV, correspondence information PR, attention position POI , To generate a projection image PV to the target position POI and the center of the projection. Details of each processing unit and the like will be described below.

  In the present embodiment, it is assumed that the three-dimensional image V is obtained by imaging by CT of the human abdomen.

First, the large intestine extraction unit 11 uses a known method for each of the axially disconnected images obtained from the three-dimensional image V and including the abdomen of the human body and including a section perpendicular to the body axis (axial). A process for separating the body from the body surface based on the body surface is performed. As a specific example, there is a method of performing binarization processing on the input axial sectional image, extracting a contour by contour extraction processing, and extracting the inside of the extracted contour as a body (human body) region, etc. . Next, a binarization process using a threshold is performed on the in-vivo region, and a candidate for a large intestine region in each axial dislocation image is extracted. Specifically, since air is contained in the large intestine, binarization is performed by setting a threshold value for extracting the air region, and the air region in the body of each axial position image is extracted. Finally, as schematically shown in FIG. 7A, only a portion where the air region is connected between the axial dislocation images is extracted as an image V _LI representing the large intestine region. FIG. 7B is an example of an image generated by performing volume rendering processing on the extracted large intestine region.

The core line extraction unit 12 extracts the core line CL passing through the central part of the colon tube from the extracted large intestine region _VLI by a known method. As a specific example, a three-dimensional thinning process is performed on a binarized image obtained by extracting only a portion to which an air region is connected, that is, a binarized image representing a large intestine region, which is obtained in the last processing of the large intestine extraction unit 11. The method (refer the said patent document 3) etc. which are performed etc. are mentioned.

Note that a user interface for manually correcting the processing results of the large intestine extraction unit 11 and the core line extraction unit 12 may be provided. As a specific example, for each axial cut image obtained from the three-dimensional image V, the extracted large intestine region V _LI and core line CL are highlighted and displayed on the axial cut image, and the user can An interface that corrects a region or core line (a point on the image) in which the extraction result is incorrect for each cut image by operating the mouse of the image processing workstation 3 or the like can be given.

  FIG. 8 schematically shows details of processing performed by the decompressed and expanded image generation unit 13.

First, the viewpoint information VP that can specify the positions of a plurality of viewpoints VP _m (m = 1, 2,..., M) and a curve in which the viewpoint VP _m should be set, that is, information on the colon core CL, are input. The positions of a plurality of viewpoints VP _m for generating the decompressed expanded image DV are determined (# 21). Here, the input viewpoint information VP is, for example, information on the initial position (for example, the end point of the extracted large intestine core line) of the decompressed expanded image DV generated first. Extension expansion image generation unit 13, from the initial position along the input core CL, at predetermined intervals, the viewpoint VP _1, VP ₂ of a predetermined number M, · · ·, set the VP _M. Note that the viewpoint interval can be given, for example, as an activation parameter of the program of the decompressed expanded image generation unit 13. Alternatively, information indicating the amount of movement in the direction of the curve (core line CL) with respect to the currently displayed expanded expanded image DV can also be an example of the viewpoint information VP. In this case, the point at which each VP _m viewpoint corresponding to the currently displayed decompressed expanded image DV is moved by the input movement amount along the core line CL is the viewpoint of the expanded expanded image DV to be generated next. VP _m position.

Next, the angle between the line-of-sight vector R1 _mn (n = 1, 2,..., N) starting from each viewpoint VP _m and the tangent vector TV _m of the curve (core line) CL at each viewpoint VP _m . θ _mn is determined (# 22). In the present embodiment, a common angle of 90 degrees is used for all line-of-sight vectors R1 _mn . Note that the tangent vector TV _m is a vector in the tangential direction of the curve (core line) CL at each viewpoint VP _m as shown in FIG. 1, and a plurality of viewpoints VP _m along the curve (core line) CL. of which is from the end of the point of view VP _M of direction towards the top of the viewpoint VP ₁ of.

As described above, after determining the angle θ _m in each line-of-sight vector, the decompressed expanded image generation unit 13 determines N line-of-sight vectors R1 _mn at each viewpoint VP _m (# 23). This processing can be performed as parallel processing for each viewpoint VP _m . Figure 9 is a line-of-sight vector R1 _mn of extension expansion image generation unit 13 sets, which is viewed from the tangential direction of the viewpoint VP _m, as shown in FIG., Line vectors R1 _mn are perspective viewpoint VP _m Is set as radial. The central angle at this time can be a value given in advance by an activation parameter or the like, or a value input by the user. FIG. 9A shows an example in which the central angle is 120 degrees, (b) is 180 degrees, and (c) is 360 degrees. When the central angle is 360 degrees, as shown in FIG. The vectors R1 _m1 and R _mN may be matched.

Next, extension expansion image generation unit 13, for each line vectors R1 _mn, using the image information in the vicinity of the line-of-sight vector R1 _mn on and / or line-of-sight vector R1 _mn, the three-dimensional image V on the line vectors R1 _mn The pixel value of the projection pixel PP _mn onto which the represented structure is projected is determined (# 24). This processing can also be performed as parallel processing for each line-of-sight vector R1 _mn . In this embodiment, volume rendering processing using a ray casting method is used as an example of a method for determining the pixel value of the projection pixel PP _mn . Specifically, it is obtained as follows.

The opacity α and the color information c are defined in advance for each tissue of the subject based on the CT value and the like, and based on this, the opacity α and the color information c in each voxel of the three-dimensional medical image V are voxeled. Keep it as data. First, a plurality of search points P _i (i = 1, 2,..., I) obtained by sampling the three-dimensional medical image V at a predetermined interval are set along the line-of-sight vector R1 _mn . Next, the luminance value b (P _i ) at each search point P _i is calculated by the following equation (1).

Here, h is a shading coefficient by diffuse reflection. In this embodiment, environmental light and specular reflection light are not considered. Further, N (P _i) is the normal vector of the plane defined by the gradient of the CT value at the search point P _i, L is a unit direction vector of the light source S given in advance from the search point P _i, "- "Is an inner product of vectors, and c (P _i ) is color information at the search point P _i , and is obtained by interpolation calculation using color information of neighboring voxels as necessary.

Further, the opacity α (P _i ) at each exploration point P _i is obtained by interpolation calculation using the opacity of neighboring voxels as necessary.

Then, as shown in the following equation (2), the product of the luminance value b (P _i ) and the opacity α (P _i ) at each exploration point P _i is added, and the accumulated value of the opacity α is a predetermined value. When the threshold value is reached or the line-of-sight vector R1 _mn exits from the target three-dimensional medical image V, the processing for the line-of-sight vector R1 _mn is terminated, and the addition result is the pixel of the projection pixel PP _mn in the line-of-sight vector R1 _mn . Determine as value C.

After calculating the pixel value of the projection pixel PP _mn for each line-of-sight vector R1 _{mn at} each viewpoint VP _m by the above processing, the decompressed expanded image generation unit 13 projects the projection pixel at each line-of-sight vector R1 _mn for each viewpoint VP _m. Projecting PP _mn onto the projection plane as one line in the order in which the line-of-sight vectors R1 _mn are arranged, that is, in the order of n = 1, 2,..., N, generates an expanded and developed image DV (# 25). Here, the arrangement order of the lines on the projection plane is the arrangement order of the viewpoints VP _m , that is, the order of m = 1, 2,. FIG. 2 schematically shows the projection position of each projection pixel PP _{mn in} the projection plane (expanded developed image DV). Incidentally, FIG. 10 illustrates a positional relationship between the projection pixel PP _mn by line-of-sight vectors R1 _mn in orientation and coordinate plane of extension developed view DV viewpoint VP _m gaze vector R1 _mn in the coordinate space with the origin. As shown in FIG. 10A, in the coordinate space where the viewpoint VP _m is the origin, the direction of the tangent vector TL _m is the z axis, and the direction of the line of sight vector R1 _m1 is the x axis, the direction of the line of sight vector R1 _m1 is used as a reference. The angle between the line-of-sight vectors R1 _m1 and R1 _mn is β, and the radiation angle (center angle) from the line-of-sight vectors R1 _m1 to R1 _mN at the viewpoint VP _m is β ₀ . On the other hand, as shown in FIG. 10B, the left-right projection of the projection pixel PP _m1 in the line-of-sight vector R1 _m1 is w and the width in the left-right direction (the above-described line direction perpendicular to the curve (core line CL)) of the expanded unfolded image DV. The position is 0, and the projected position in the left-right direction of the projection pixel PP _{mn at} the line-of-sight vector R1 _mn is x. Here, when a rotation matrix that rotates β degrees around the z axis in FIG. 10A is RotZ (β), the following equation (3) is established.

Therefore, based on this equation (3), that is, the radiation angle from the line-of-sight vector R1 _m1 to R1 _mN of the angle β formed by the line-of-sight vector R1 _{mn at} each viewpoint VP _m and the reference line-of-sight vector (for example, R1 _m1 ). The ratio to β ₀ matches the ratio of the position in the left-right direction (line direction) of the projection pixel PP _mn corresponding to the line-of-sight vector R1 _mn in the developed image DV to the width in the left-right direction (line direction) of the developed image DV. In addition, the position of each projection pixel PP _mn in the left-right direction (line direction) on the developed image DV can be determined.

On the other hand, for the projection position in the direction corresponding to the up-down direction (core line CL direction) of the expanded expanded image DV, that is, the direction in which the viewpoint VP _m is arranged, for example, the above line at the viewpoint VP ₁ is the top of the expanded expanded image DV, the line and the bottom of the extension developed view DV in the viewpoint VP _N, during which the viewpoint VP _2, VP _3, the line in ··· VP _M-1, the viewpoint VP _m in the coordinate space of the three-dimensional medical image V What is necessary is just to arrange | position according to a space | interval.

  The image in FIG. 13A is an example of the decompressed expanded image DV generated by the above-described processing by the expanded expanded image generation unit 13.

The correspondence information storage unit 14 is a predetermined area of the main storage device of the image processing workstation 3. In this embodiment, the correspondence information storage unit 14 corresponds to the position in the coordinate system of the three-dimensional image V of the _first viewpoint VP ₁ of the decompressed expanded image DV. Assume that it is stored as information PR.

  The viewpoint setting accepting unit 15 provides a user interface that accepts an input of a movement amount in a curve (core line CL) direction with respect to the currently displayed expanded development image DV. As a specific example, after selecting the expanded / unfolded image DV by clicking the expanded / unfolded image DV displayed on the display of the image processing workstation 3, the mouse wheel is rotated, the keyboard is A user interface that accepts the amount of rotation and the number of times of pressing (pressing time) as an input of the amount of movement in the direction of the curve (core line CL) with respect to the expanded expanded image DV by pressing an arrow key can be mentioned. Note that the conversion from the rotation amount and the number of times of pressing (pressing time) to the moving amount may be prepared in advance. In addition, a user interface that displays an image of the large intestine viewed from the outside of the body (see FIG. 7B) and an image in which the core line CL is superimposed on the display, and allows setting of a new viewpoint position for the image. It may be.

The position designation receiving unit 16 provides a user interface that accepts designation of a position in the currently displayed decompressed expanded image DV. For example, by clicking on the position of interest POI in the expanded expanded image DV illustrated in FIG. 13A with the mouse, the position designation receiving unit 16 causes the clicked position POI in the coordinate system of the expanded expanded image DV (for example, The coordinates of PP _uv ) in FIG. 2 are acquired.

The projection image generation unit 17 specifies the line-of-sight vector R1 _uv corresponding to the target position POI (PP _uv ) received by the position designation reception unit 16 based on the correspondence information PR stored in the correspondence information storage unit 14. A position (R1 _uv) in the three-dimensional image V corresponding to the target position PP _uv is obtained by central projection with the start point VP _u of the identified line-of-sight vector R1 _uv as the viewpoint and the direction of the line-of-sight vector R1 _uv as the center of the projection direction. A projection image PV is generated in which an arbitrary position above is projected at the center. Details will be described below.

[1] Specification of the position of the viewpoint corresponding to the target position in the coordinate system of the three-dimensional image The projection image generation unit 17 first expands and expands the image DV of the _target position POI (PP _uv ) received by the position designation reception unit 16. from the coordinates in the coordinate system, the target position POI (PP _uv) to identify what number of pixels in the vertical direction of extension developed view DV (direction of the core wire CL). Here, it is the u-th pixel. Next, the correspondence information PR stored in the correspondence information storage unit 14, that is, the position in the coordinate system of the three-dimensional image V of the _first viewpoint VP1 of the expanded expanded image DV, and the expanded expanded image generation unit 13 in the present embodiment. The position of the u-th viewpoint, that is, the viewpoint VP _u corresponding to the projection pixel PP _uv on the core line CL, in the coordinate system is specified based on the viewpoint interval given as the program start parameter.

[2] Identification of the direction of the line-of-sight vector corresponding to the target position in the coordinate system of the three-dimensional image On the other hand, the projection image generation unit 17 expands and expands the _target position POI (PP _uv ) received by the position designation receiving unit 16 From the coordinates in the DV coordinate system, the pixel number in the left-right direction (line direction) of the expanded expanded image DV is specified from the position of interest POI (PP _uv ). Here, it is the vth pixel. Next, a width x from the first pixel to the vth in the left-right direction (line direction) of the expanded expanded image DV is obtained, and this width x and the image width w of the expanded expanded image DV are determined in advance. Based on the direction of the reference line-of-sight vector R1 _u1 and the radiation angle (center angle) β ₀ from the line-of-sight vectors R1 _u1 to R1 _uN , the angle β between the line-of-sight vectors R1 _u1 and R1 _uv is β ₀ Since x / w, the direction in the coordinate system of FIG. 10A (hereinafter referred to as the coordinate system of the projection calculation plane) of the line-of-sight vector R1 _uv corresponding to the target position PP _uv is obtained by the above equation (3). It is done.

Further, the projection image generation unit 17 obtains the orientation of the line-of-sight vector R1 _uv in the coordinate system of the three-dimensional image V. FIG. 11 schematically shows the relationship between the coordinate system C ₁ of the projection calculation plane and the coordinate system C ₀ of the three-dimensional image V. As shown in the figure, T is a matrix representing the inclination of the coordinate system C ₁ of the projection calculation plane viewed from the coordinate system C ₀ of the three-dimensional image V, and the direction of the line-of-sight vector R1 _uv in the coordinate system of the projection calculation plane is _{Assuming that} C ₁ (R1 _uv ), the direction C ₀ (R1 _uv ) of the line-of-sight vector R1 _uv in the coordinate system of the three-dimensional image V is obtained by the following equation (4).

C ₀ (R1 _uv ) = T * C ₁ (R1 _uv ) (4)
Here, “*” represents a matrix product.

[3] Position C ₀ (VP _u) of the viewpoint VP _u in the coordinate system of the three-dimensional image V obtained by the generated projection image generation unit 17, the above processing of the projected image, line-of-sight vector R1 _uv orientation C ₀ ( with R1 _uv), as schematically illustrated in FIG. 12 (a), the view point the starting point VP _u of the line-of-sight vector R1 _uv, the central projection of the center of the direction of projecting the direction of the sight line vector R1 _uv The projection image PV is generated. The specific content of the central projection method is a volume rendering method in this embodiment. The calculation method of the projection pixel for each line-of-sight vector in this central projection is the same as the above formulas (1) and (2). The image in FIG. 13B is an example of the center projection image PV created by the above processing, and the position in the three-dimensional image V corresponding to the target position (PP _uv ) (arbitrary position on R1 _uv ). Is projected at the center of the image.

The projection image generation unit 17, the center position of the viewpoint shown in FIG. 12 (b) to schematically represent the line of sight vector R1 _uv starting point as (VP _u), the projection of the direction of the sight line vector R1 _uv orientation The parallel projection image PV may be generated by parallel projection.

  Next, the workflow realized by the image interpretation support apparatus of the present invention will be described using the flowchart of FIG. 14, the block diagram of FIG.

  First, when a user such as an image diagnostician performs an operation of selecting a desired three-dimensional image V from an image list to be diagnosed displayed on the image processing workstation 3 (# 1), the axis of the three-dimensional image V is selected. A cross-sectional image by the dislocation is displayed on the display.

Next, when the user selects “colon analysis” from the menu screen displayed on the display of the image processing workstation 3 in order to make a diagnosis based on the three-dimensional image V (# 2), the colon extracting unit 11 However, the image V _LI representing the large intestine region is extracted from the inputted three-dimensional image V (# 3), and the core line extraction unit 12 extracts the core line CL passing through the central part of the V _LI tube of the large intestine (# Four). Here, the user confirms the extraction result on the screen, and manually corrects the extraction result as necessary.

After the core line is confirmed, when the user selects “Generate expanded expanded image” from the menu screen, the expanded expanded image generation unit 13 determines the lower end of the core line CL (the end portion of the large intestine) based on the activation parameter of the program. Is determined as the initial development position (viewpoint) of the decompressed expanded image DV, and M viewpoints VP _m (m = 1, 2,..., M) are set on the core line CL based on the positions. line vectors R1 _mn at each of the viewpoints VP _m which is (n = 1,2, ···, n ) and to determine the angle theta _m of the angle between the tangent vector TV _m at the viewpoint VP _m. Next, the viewpoint VP by the parallel processing for each _m, and determines the direction of the sight line vector R1 _mn at each viewpoint VP _m, further, by the parallel processing for each line vectors R1 _mn, line vectors R1 _mn on and / or line-of-sight vector R1 using the image information of the vicinity of the _mn, structure represented in the three-dimensional image V on the line vectors R1 _mn determines the pixel values of projection pixels PP _mn projected. Then, for each viewpoint VP _m, the projection pixel PP _mn in each line vectors R1 _mn, the arrangement order of the line-of-sight vector R1 _mn, i.e. n = 1, 2, · · ·, in the order of N, as one line, each of the viewpoints VP order of _m, i.e., m = 1, 2, · · ·, by projecting the projection plane as order of M is retained, to generate an extended deployment image DV (# 6). At this time, the coordinates in the coordinate system of the three-dimensional image V of the _first viewpoint VP ₁ are stored in the correspondence information storage unit 14 as correspondence information PR.

  The generated expanded development image DV is displayed on the display of the image processing workstation 3. For example, when the user wants to observe an expanded development image DV as shown in FIG. 13 (a) and wants to observe a previous image along the core line CL, the user advances the mouse wheel after clicking the expanded development image DV. When rotated, the viewpoint setting accepting unit 15 acquires the amount of rotation, accepts it as an input of the amount of movement in the direction of the curve (core line CL) with respect to the expanded expanded image DV, and the expanded expanded image generation unit 13 determines the amount of movement as a viewpoint. Deliver it as information VP (# 7; YES).

The decompressed expanded image generation unit 13 sets the viewpoint VP _m of the expanded expanded image DV to be generated next on the basis of the viewpoint movement amount information VP acquired by the viewpoint setting reception unit 15, and thereafter, a new process is performed through the same process as described above. A simple decompressed expanded image DV is generated (# 6). The image processing workstation 3 updates the display of the decompressed expanded image DV to the newly generated image, and the user observes the updated expanded expanded image. At this time, the coordinates in the coordinate system of the three-dimensional image V of the _first viewpoint VP ₁ stored in the correspondence information storage unit 14 are also updated.

Next, when the user clicks and designates a position POI to be noted of the decompressed and expanded image DV as shown in FIG. 13A, for example, the position designation receiving unit 16 clicks on the clicked position POI (for example, FIG. 2). To obtain the coordinates in the coordinate system of the decompressed expanded image DV of (PP _uv ) of (# 8; YES).

Based on the correspondence information PR and the coordinates of the clicked position POI (PP _uv ) in the coordinate system of the expanded and expanded image DV, the projection image generation unit 17 generates the 3D image V of the viewpoint VP _u corresponding to the position PP _uv . The position in the coordinate system is specified, the direction of the line-of-sight vector R _uv corresponding to the position PP _uv in the coordinate system of the projection calculation plane is obtained, and the direction of the line-of-sight vector R _uv in the coordinate system of the projection calculation plane is further determined by 3 The direction of the dimension image V in the coordinate system is converted. Then, based on the viewpoint VP _u and the direction of the line-of-sight vector R _uv in the coordinate system of the three-dimensional image V, a projection image PV by central projection or parallel projection is generated (# 9). The generated projection image PV is displayed on the display of the image processing workstation 3.

As described above, in the three-dimensional medical image processing system in which the image interpretation apparatus according to the first embodiment of the present invention is mounted, the decompressed and expanded image generation unit 13 includes the image V _LI representing the large intestine region, the initial viewpoint The decompressed expanded image DV is generated based on the viewpoint information VP and the core line CL received by the position or viewpoint setting receiving unit 15, and the correspondence information PR is stored in the correspondence information storage unit 14, and the position designation receiving unit 15 decompresses. Upon receiving the designation of the target position POI in the developed image DV, the projection image generating unit 17 generates a projection image PV having the target position POI as the center of projection based on the decompressed developed image DV, the correspondence information PR, and the target position POI. To do. Therefore, if the attention position POI found in the image DV is specified after performing the overall observation with the decompressed and expanded image DV, the position of the attention position POI in the three-dimensional image V is further specified. Therefore, a projection image DV centered on the target position POI is automatically generated and can be observed, so part and detail of the entire image representing the bent tubular structure can be observed. The workflow is more efficient.

  Next, as a second embodiment of the present invention, a case where an expanded cross-sectional image SV is generated instead of the expanded expanded image DV will be described. FIG. 15 is a block diagram showing functions of an image interpretation support apparatus according to the second embodiment of the present invention. As shown in the figure, the expanded expanded image generation unit 13 is replaced with an expanded cross-section generation unit 18. Further, the contents of the correspondence information PR and the contents of the processing of the projection image generation unit 17 are also different from the above embodiment. Hereinafter, these differences will be mainly described.

The expanded cross-sectional image generation unit 18 has a plurality of viewpoints VP _mn (m = 1, 2,..., M, n = 1, 2,..., N) on the curved surface CS passing through the core line CL of the large intestine V _LI. the per line vectors R2 _mn to start, using the image information in the vicinity of the line-of-sight vector R2 _mn on and / or line-of-sight vector R2 _mn, the projection pixel PP _mn of colon V _LI is projected on the eye vector R2 _mn A pixel value is determined, and an expanded cross-sectional image SV in which each of the projection pixels PP _mn is arranged on the projection plane is generated so that the arrangement order of the viewpoints VP _mn is maintained.

  FIG. 16 schematically shows details of processing performed by the elongated cross-sectional image generation unit 18. FIG. 17 schematically shows how the viewpoint and the line-of-sight vector are set by this processing.

First, as in the first embodiment, (a VP _mC) a plurality of viewpoints on the core line CL of the large intestine V _LI sets a (# 31).

Next, a line-of-sight vector R2 _mC (n = 1, 2,..., N) starting from a plurality of viewpoints VP _mC on the core line CL and a tangent vector TV _m of the curve (core line) CL at each viewpoint VP _{mC Is} determined (# 32). In the present embodiment, it is assumed that the angle θ _mC is 90 degrees for all the line-of-sight vectors R2 _mC , and is given as a program start parameter of the elongated cross-sectional image generation unit 18. Note that the tangent vector TV _m is a tangential vector of the core line CL in each viewpoint VP _mC, along the core line CL, the top of the viewpoint VP _1C from the end of the viewpoint VP _MC of the plurality of viewpoints VP _mC The direction is toward. It should be noted that which side of the large intestine V _LI is the viewpoint vector R2 _mC is given as the above parameter. Details will be described later (see FIG. 19).

Next, for each viewpoint VP _mC (m = 1, 2,..., M), on the line perpendicular to the tangent vector TV _m and the line of sight vector R2 _mC , the viewpoint VP _mn (n = 1, 2,... ·, set N), the determined parallel line vectors R2 _mn gaze vector R2 _mC that starting from the viewpoints VP _mn (# 33). This processing can be performed as parallel processing for each viewpoint VP _mC . Note that the curved surface CS in FIG. 17 represents a curved surface passing through all the viewpoints VP _mn .

Extension sectional image generation unit 18, in the same manner as in the first embodiment, the parallel processing of each line vectors R2 _mn, structures represented in the three-dimensional image V on the line vectors R2 _mn is projected projection The pixel value of the pixel PP _mn is determined (# 34).

Finally, the line for each, i.e., for each value the same perspective m viewpoint VP _mn, the projection pixel PP _mn in each line vectors R2 _mn, the arrangement order of the line-of-sight vector R2 _mn, i.e. n = 1, 2, · .., N are projected onto the projection plane as one line in the order of N to generate an expanded cross-sectional image SV (# 35). Here, the arrangement order of the lines on the projection plane is the arrangement order of the viewpoints VP _m , that is, the order of m = 1, 2,. FIG. 2 schematically shows the projection position of each projection pixel PP _{mn in} the projection plane (elongated sectional image SV).

Here, in this embodiment, the line-of-sight vectors R2 _mn (n = 1, 2,..., N) are parallel, and _{therefore the} line-of-sight lines VP _mn (n = 1, 2,... , N) is the pixel interval of the projection pixel PP _mn (n = 1, 2,..., N) in the expanded cross-sectional image SV as it is (see FIG. 20).

On the other hand, for the projection position in the direction corresponding to the vertical direction (the core line CL direction) of the expanded cross-sectional image SV, that is, the viewpoint VP _mC (the direction in which m = 1, 2,..., M is arranged), for example, the viewpoint VP _1C The above line at the top of the expanded cross-sectional image SV and the above line at the viewpoint VP _NC as the bottom of the expanded cross-sectional image SV, and the above lines at the viewpoints VP _2C , VP _3C ,... VP _{(M-1) C May} be arranged according to the interval of the viewpoints VP _mC in the coordinate space of the three-dimensional medical image V.

FIG. 18 is an example of an expanded slice image SV generated by the above processing by the expanded slice image generation unit 18. As shown in the figure, since the expanded cross-sectional image SV is a cross section of the large intestine V _LI by the curved surface CS, the display in a direction perpendicular to the core line CL rather than the expanded expanded image DV showing the state where the tube of the large intestine V _LI is cut open The range becomes narrower.

Therefore, a plurality of elongated sectional images SV may be combined so that the entire circumference direction of the large intestine _VLI can be confirmed. FIG. 19 schematically shows an example thereof. The figure shows a cross section obtained by slicing the large intestine _VLI in a direction perpendicular to the core line CL. As shown, the central angle a of the three 180-degree extension sectional images SV _1, SV _2, SV _3, generates with overlapping ends by 60 degrees. Curved CS1, CS2, CS3, respectively, is a curved surface that contains the point of view of time of generation of the extended cross-sectional image SV _1, SV _2, SV ₃ is. Thus, with overlapping right and left end portions of the elongated cross-sectional image by generating a plurality of elongated cross-sectional image _{SV 1, SV 2, SV 3} , can be avoided that the region of interest is hard to see over a plurality of images.

In the present embodiment, the correspondence information storage unit 14 stores, as correspondence information PR, the position in the coordinate system of the three-dimensional image V of the first viewpoint VP _1C of the expanded _slice image SV among the viewpoints VP _mC passing through the core line CL. Shall.

The projection image generation unit 17 specifies the line-of-sight vector R2 _uv corresponding to the attention position POI (for example, PP _uv ) received by the position designation reception unit 16 based on the correspondence information PR stored in the correspondence information storage unit 14. On the line-of-sight vector R2 _uv , a main projection position (for example, P _uv ) is determined so that the degree of contribution to the determination of the pixel value of the projection pixel PP _uv corresponding to the line-of-sight vector R2 _uv satisfies a predetermined criterion. A projection image is generated so that the main projection position P _uv is projected at the center by projection or parallel projection. Details will be described below.

[1] Specifying the position in the coordinate system of the three-dimensional image of the viewpoint on the line corresponding to the target position and on the core line First, the projected image generation unit 17 receives the _target position POI (PP _uv received by the position designation receiving unit 16 ) In the coordinate system of the expanded cross-sectional image SV, the number of pixels in the vertical direction (direction of the core line CL) of the target position POI (PP _uv ) is specified. Here, it is the u-th pixel. Next, the correspondence information PR stored in the correspondence information storage unit 14, that is, the position in the coordinate system of the three-dimensional image V of the first viewpoint VP _1C of the expanded slice image SV, and the expanded slice image generation unit 18 in the present embodiment. The position of the u-th viewpoint, that is, the viewpoint VP _uC corresponding to the projection pixel PP _uC on the core line CL in the coordinate system of the u-th viewpoint is specified based on the viewpoint interval given as the program start parameter.

[2] Specifying the position and orientation of the viewpoint / line-of-sight vector corresponding to the target position in the coordinate system of the projection calculation plane The projection image generation unit 17 then moves the coordinate system (projection) in each line as shown in FIG. The position of the viewpoint VP _uv and the direction of the line-of-sight vector R2 _uv in the coordinate system on the calculation plane) are obtained.

20A shows the position of the viewpoint VP _uv and the direction of the viewpoint line-of-sight vector R2 _uv in the coordinate system of the projection calculation plane with the viewpoint VP _uC as the origin, and FIG. 20B shows an expanded cross section. This represents the position of the projection pixel PP _uv by the line-of-sight vector R2 _uv in the coordinate plane of the image SV.

As shown in FIG. 20A, the viewpoint VP _uC on the core line CL, that is, the midpoint of the line where the viewpoints VP _u1 , VP _u2 ,..., VP _uN are arranged is the origin, the direction of the tangent vector TL _u is the z axis, and the viewpoint If the direction of the line where VP _un (n = 1, 2,..., N) is arranged is the x axis, the direction of the line-of-sight vector R2 _uv is the y axis direction.

Next, the position of the viewpoint VP _uv in the coordinate system of the projection calculation plane is specified. From the coordinates in the coordinate system of the expanded expanded image DV of the _target position POI (PP _uv ) received by the position designation receiving unit 16, the projection image generation unit 17 determines that the _target position POI (PP _uv ) is in the horizontal direction of the expanded cross-sectional image SV. The pixel number in the (line direction) is specified. Here, it is the vth pixel. Next, as shown in FIG. 20B, from the middle point O in the left-right direction (line direction) of the expanded sectional image SV, that is, from the _Cth pixel (projection pixel PP _uC passing through the core line) to the _vth . Find the width x between them. Here, since the line-of-sight vectors _R2un (n = 1, 2,..., N) are parallel to each other, the width w / 2 from the midpoint O to the projection pixel _PPuN in FIG. , The width x from the pixel (projection pixel PP _uC passing through the core line) to the v-th is the width from the viewpoint VP _uC to the viewpoint VP _uN and the viewpoint VP _uC to the viewpoint VP _{uv in FIG.} It becomes the width to. In FIG. 20A, a vector representing the direction of the line where the viewpoints VP _un (n = 1, 2,..., N) are arranged is VL, and a vector representing the position of the viewpoint VP _uv in the line direction is P _x . Then, the following expression (5) is established.

The P _x is the vector representing the position of the viewpoint VP _uv in the projection calculation plane.

[3] Determination of main projection position on line-of-sight vector corresponding to target position Next, as shown in FIG. 20A, the projection image generation unit 17 _changes the line-of-sight vector R2 _uv to the line-of-sight vector R2 _uv . The main projection position P _uv is determined so high that the degree of contribution to the determination of the pixel value of the corresponding projection pixel PP _uv satisfies a predetermined criterion. In the present embodiment, since volume rendering processing is performed when calculating the pixel value of the projection pixel PP _mn , the point having the maximum opacity among the search points on the line-of-sight vector R2 _uv is determined as the main projection position P _uv. And decide. Here, since the main projection position P _uv is obtained as a position in the coordinate system of the projection calculation plane, it is converted into a position in the coordinate system of the three-dimensional image V as follows.

FIG. 21 schematically shows the relationship between the coordinate system C ₁ of the projection calculation plane and the coordinate system C ₀ of the three-dimensional image V. As in FIG. 11 in the first embodiment, T represents a matrix representing the inclination of the coordinate system C ₁ of the projection calculation plane viewed from the coordinate system C ₀ of the three-dimensional image V, and the core CL is obtained in advance. If the position of the viewpoint VP _uC in the coordinate system C ₀ of the three-dimensional image V is C ₀ (VP _uC ), the position C ₀ (P _uv ) of the main projection position P _uv in the coordinate system of the three-dimensional image V is It is obtained by equation (6).

C ₀ (P _uv ) = T * C ₁ (P _uv ) + C ₀ (VP _uC ) (6)
Here, “*” represents a matrix product.

[4] Projection Image Generation The projection image generation unit 17 obtains the position C ₀ (VP _uC ) and the main projection position C ₀ (P _uv ) of the viewpoint VP _uC in the coordinate system of the three-dimensional image V obtained by the above processing. using viewpoint VP _uC position C ₀ to (VP _uC) and viewpoint, and the center position C ₀ (VP _uC) from the projection such that a direction in the main projection position C ₀ (P _uv) orientation of the viewpoint VP _uC A projection image PV is generated by central projection.

The projection image generation unit 17, the position C ₀ of the viewpoint VP _uC the (VP _uC) centered perspective, directed from the viewpoint VP _uC position C ₀ (VP _uC) in the main projection position C ₀ (P _uv) orientation The projection image PV may be generated by parallel projection with the projection direction as.

Further, since the position C ₀ (P _uv ) in the coordinate system of the three-dimensional image V of the main projection position P _uv is obtained in [3] above, an arbitrary position in the coordinate system of the three-dimensional image V is taken as a viewpoint, and the main position is derived therefrom. A central projection may be performed with the direction toward the projection position C ₀ (P _uv ) as the center of the projection direction, or a projection pixel by a vector from the arbitrary position toward the main projection position C ₀ (P _uv ). May be the center of the projected image, and parallel projection may be performed with the direction of the vector as the projection direction.

  As described above, in the second embodiment of the present invention, the position of interest is specified in the expanded slice image SV generated by the expanded slice image generation unit 18. As in the case of the decompressed and expanded image DV in the embodiment, the projection image PV is obtained without further specifying the position or the like of the target position in the three-dimensional image V by the user.

  In addition, what changed variously in the range which does not deviate from the meaning of this invention with respect to the system configuration | structure in the said embodiment, a processing flow, a user interface, etc. is also contained in the technical scope of this invention. Moreover, said embodiment is an illustration to the last and all the above description should not be utilized in order to interpret the technical scope of this invention restrictively.

  For example, an image processing server (not shown) is separately connected to the network 9 without performing part or all of the processes of the large intestine extraction unit 11, the core line extraction unit 12, and the decompressed expanded image generation unit 13 on the image processing workstation 3. The image processing server may perform the processing in response to the processing request from the image processing workstation 3.

In addition, on the line-of-sight vector R1 _uv corresponding to the target position PP _uv of the first embodiment of the present invention, the degree of contribution to the determination of the pixel value of the projection pixel PP _uv corresponding to the line-of-sight vector R1 _uv is a predetermined reference. A main projection position (e.g., P _uv ) that is high enough to be satisfied is determined, and a projection image is generated so that the main projection position P _uv is projected almost at the center by central projection or parallel projection from an arbitrary point. Also good.

  Furthermore, in the above embodiment, only the minimum necessary information is stored as the correspondence information PR. However, each pixel of the decompressed expanded image DV or the decompressed cross-sectional image SV and the viewpoint corresponding to the pixel are displayed. The position and the direction of the line-of-sight vector may be stored in association with each other. At this time, these pieces of information in each coordinate system of the projection calculation plane and the three-dimensional image may be stored. Further, the main projection position on each line-of-sight vector may be stored.

The figure which represented typically the setting method of the gaze vector in the 1st aspect of this invention. A diagram that schematically shows the positional relationship of the projected pixels in the expanded developed image and the expanded cross-sectional image. The figure which represented typically the setting method of the gaze vector in the 2nd aspect of this invention. The figure which represented typically the production | generation method of the projection image in the 2nd aspect of this invention 1 is a schematic configuration diagram of a three-dimensional medical image processing system in which an expansion / development image generation apparatus according to an embodiment of the present invention is mounted. The block diagram which showed typically the structure of the image interpretation assistance apparatus in the 1st Embodiment of this invention. Diagram for explaining large intestine extraction processing A diagram that schematically shows the flow of the decompressed expanded image generation process Illustration for explaining the radiation angle of the line-of-sight vector A diagram that schematically shows the positional correspondence between the direction of the line-of-sight vector and the projected pixel in the expanded expanded image The figure which represented typically the relationship between the coordinate system of the projection calculation plane in the 1st Embodiment of this invention, and the coordinate system of a three-dimensional image. (A) A diagram schematically showing central projection and (b) parallel projection (A) A diagram showing an example of an expanded image and (b) a central projection image The flowchart which showed an example of the image interpretation workflow in embodiment of this invention The block diagram which showed typically the structure of the image interpretation assistance apparatus in the 2nd Embodiment of this invention. Diagram that schematically shows the flow of the expanded cross-sectional image generation process The figure which represented typically the setting method of the gaze vector in the 2nd Embodiment of this invention. A diagram showing an example of an expanded cross-sectional image The figure showing an example of the display range of each image at the time of generating a plurality of expansion section images A diagram schematically showing the positional correspondence between the direction of the line-of-sight vector and the projected pixel in the expanded cross-sectional image The figure which represented typically the relationship between the coordinate system of the projection calculation plane in the 2nd Embodiment of this invention, and the coordinate system of a three-dimensional image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Modality 2 Image storage server 3 Image processing workstation 11 Large intestine extraction part 12 Core line extraction part 13 Expansion expansion image generation part 14 Corresponding information storage part 15 Viewpoint setting reception part 16 Position designation reception part 17 Projection image generation part 18 Expansion cross-section image generation Part

Claims (9)

In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And an expanded expanded image generating means for generating an expanded expanded image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
Correspondence information storage means for storing correspondence information capable of specifying a correspondence relationship between the position of each projection pixel in the expanded development image and each line-of-sight vector in the three-dimensional image;
Position designation accepting means for accepting designation of a position of interest in the decompressed expanded image;
Based on the correspondence information stored in the correspondence information storage means, the line-of-sight vector corresponding to the position of interest received by the position designation reception means is specified, the start point of the specified line-of-sight vector is used as a viewpoint, and the line-of-sight The central projection with the vector direction as the approximate center of the projection direction, or the parallel projection with the position of the start point of the identified line-of-sight vector as the approximate center of the viewpoint and the direction of the line-of-sight vector as the projection direction, An image interpretation support apparatus, comprising: a projection image generation unit configured to generate a projection image in which a position in the three-dimensional image corresponding to a position of interest is projected substantially at the center. In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And an expanded expanded image generating means for generating an expanded expanded image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
Correspondence information storage means for storing correspondence information capable of specifying a correspondence relationship between the position of each projection pixel in the expanded development image and each line-of-sight vector in the three-dimensional image;
Position designation accepting means for accepting designation of a position of interest in the decompressed expanded image;
Based on the correspondence information stored in the correspondence information storage unit, the line-of-sight vector corresponding to the target position received by the position designation reception unit is specified, and the projection pixel corresponding to the line-of-sight vector on the line-of-sight vector A main projection position whose degree of contribution to the determination of the pixel value is high enough to satisfy a predetermined criterion is determined, and a projection image is generated by central projection or parallel projection so that the main projection position is projected almost at the center. An image interpretation support apparatus comprising a projection image generation unit. A plurality of lines of sight toward a peripheral direction on one side of a curved surface passing through each viewpoint, starting from each of a plurality of viewpoints inside the structure in the three-dimensional image representing the bent tubular structure For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints An expanded cross-sectional image generating means for generating an extended cross-sectional image in which each of the projection pixels is arranged on the projection plane,
Correspondence information storage means for storing correspondence information that can specify the correspondence between the position of each projection pixel in the elongated cross-sectional image and each line-of-sight vector in the three-dimensional image;
Position designation accepting means for accepting designation of a position of interest in the elongated cross-sectional image;
Based on the correspondence information stored in the correspondence information storage unit, the line-of-sight vector corresponding to the target position received by the position designation reception unit is specified, and the projection pixel corresponding to the line-of-sight vector on the line-of-sight vector A main projection position whose degree of contribution to the determination of the pixel value is high enough to satisfy a predetermined criterion is determined, and a projection image is generated by central projection or parallel projection so that the main projection position is projected almost at the center. An image interpretation support apparatus comprising a projection image generation unit. Computer
In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And a process of generating an expanded and developed image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
A process of storing correspondence information that can specify a correspondence relationship between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the decompressed expanded image;
Based on the stored correspondence information, the line-of-sight vector corresponding to the received position of interest is identified, the starting point of the identified line-of-sight vector is used as a viewpoint, and the direction of the line-of-sight vector is approximately the center of the projection direction. In the three-dimensional image corresponding to the target position by the center projection or parallel projection with the position of the start point of the identified line-of-sight vector as the center of the viewpoint and the direction of the line-of-sight vector as the projection direction. A computer-aided image interpretation support method, comprising: generating a projection image in which the position of the image is projected substantially at the center. Computer
In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And a process of generating an expanded and developed image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
A process of storing correspondence information that can specify a correspondence relationship between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the decompressed expanded image;
Based on the stored correspondence information, the line-of-sight vector corresponding to the received position of interest is identified, and the degree of contribution to the determination of the pixel value of the projection pixel corresponding to the line-of-sight vector is determined on the line-of-sight vector. A computer that determines a main projection position that is high enough to satisfy a predetermined standard, and generates a projection image so that the main projection position is projected substantially at the center by central projection or parallel projection. Image interpretation support method using Computer
A plurality of lines of sight toward a peripheral direction on one side of a curved surface passing through each viewpoint, starting from each of a plurality of viewpoints inside the structure in the three-dimensional image representing the bent tubular structure For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints To generate an expanded cross-sectional image in which each of the projection pixels is arranged in the projection plane so that
A process of storing correspondence information that can identify a correspondence relationship between the position of each projection pixel in the elongated cross-sectional image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the elongated cross-sectional image;
Based on the stored correspondence information, the line-of-sight vector corresponding to the received position of interest is identified, and the degree of contribution to the determination of the pixel value of the projection pixel corresponding to the line-of-sight vector is determined on the line-of-sight vector. A computer that determines a main projection position that is high enough to satisfy a predetermined standard, and generates a projection image so that the main projection position is projected substantially at the center by central projection or parallel projection. Image interpretation support method using On the computer,
In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And a process of generating an expanded and developed image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
Processing for storing in the storage means correspondence information that can specify the correspondence between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the decompressed expanded image;
Based on the stored correspondence information, the line-of-sight vector corresponding to the received position of interest is identified, the starting point of the identified line-of-sight vector is used as a viewpoint, and the direction of the line-of-sight vector is approximately the center of the projection direction. In the three-dimensional image corresponding to the target position by the center projection or parallel projection with the position of the start point of the identified line-of-sight vector as the center of the viewpoint and the direction of the line-of-sight vector as the projection direction. And a process of generating a projection image in which the position of the image is projected at substantially the center. On the computer,
In each of a plurality of viewpoints on the core line passing through the central portion of the structure in the three-dimensional image representing the bent tubular structure, a plurality of lines of sight toward the peripheral direction of the core line starting from the viewpoint For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints And a process of generating an expanded and developed image in which each of the projection pixels is arranged on a projection plane so that the arrangement order of the plurality of line-of-sight vectors at each viewpoint is maintained.
Processing for storing in the storage means correspondence information that can specify the correspondence between the position of each projection pixel in the decompressed expanded image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the decompressed expanded image;
Based on the correspondence information stored in the storage unit, the line-of-sight vector corresponding to the received position of interest is specified, and on the line-of-sight vector, the pixel value of the projection pixel corresponding to the line-of-sight vector is determined. Determining a main projection position whose degree of contribution is high enough to satisfy a predetermined criterion, and executing a process of generating a projection image so that the main projection position is projected substantially at the center by central projection or parallel projection An image interpretation support program characterized by On the computer,
A plurality of lines of sight toward a peripheral direction on one side of a curved surface passing through each viewpoint, starting from each of a plurality of viewpoints inside the structure in the three-dimensional image representing the bent tubular structure For each vector, using the image information on and / or in the vicinity of the line-of-sight vector, a pixel value of a projection pixel on which the structure on the line-of-sight vector is projected is determined, and the order of arrangement of the viewpoints To generate an expanded cross-sectional image in which each of the projection pixels is arranged in the projection plane so that
Processing for storing in the storage means correspondence information that can specify the correspondence between the position of each projection pixel in the elongated cross-sectional image and each line-of-sight vector in the three-dimensional image;
Processing for accepting designation of a position of interest in the elongated cross-sectional image;
Based on the correspondence information stored in the storage unit, the line-of-sight vector corresponding to the received position of interest is specified, and on the line-of-sight vector, the pixel value of the projection pixel corresponding to the line-of-sight vector is determined. Determining a main projection position whose degree of contribution is high enough to satisfy a predetermined criterion, and executing a process of generating a projection image so that the main projection position is projected substantially at the center by central projection or parallel projection An image interpretation support program characterized by