JP6292863B2 - Image processing apparatus and image processing method - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing method, and more particularly, to adjustment of an image processing condition (template).

  Conventionally, medical imaging apparatuses such as X-ray CT (Computed Tomography) apparatus, X-ray diagnostic apparatus, ultrasonic diagnostic apparatus, MRI (Magnetic Resonance Imaging) apparatus, etc. have a disease or examination purpose or a state at the time of patient imaging. Shooting is performed under various shooting conditions according to the above. In addition, the medical image generated by the medical imaging apparatus is an image suitable for diagnosis (analysis result image) in which the positional relationship, orientation, angle, image density, or display range of the organ is adjusted using the image processing apparatus. Image processing is performed so that In an image processing apparatus, image processing conditions applied to an image of a patient are stored as a template, and the same template is applied to an image of another patient. A function of obtaining an analysis result image is provided.

  For example, Patent Document 1 discloses a medical image diagnostic apparatus that supports image processing work by creating and displaying a work flowchart representing a series of work procedures from preparation to end of image collection and display using a graphic processor. Is described. By using the medical image diagnostic apparatus and the like described in Patent Literature 1, even an unskilled person can proceed with work necessary for image processing to some extent.

JP 2001-000397 A

  However, medical images that are the basis of image processing differ depending on the imaging range in the body axis direction at the time of imaging, FOV (Field Of View) differences, individual patient differences, etc. Even if the template is directly applied to another patient's image, it is difficult to obtain the same analysis result image. In particular, when the difference between the original images is large, such as greatly different physiques, the template adjustment work takes a lot of time and effort.

  The present invention has been made in view of the above problems, and an image processing apparatus and an image processing method capable of easily and quickly adjusting an image processing parameter group included in a predefined template. The purpose is to provide.

  In order to achieve the above-described object, the first invention stores a template table that stores an image processing template, information on an original image to which the template is applied, and feature amount information used in the template in association with each other. A template selection unit that selects the template to be applied to the processing target image, a feature amount extraction unit that extracts a feature amount of the processing target image, and features of the processing target image extracted by the feature amount extraction unit A template conversion unit that converts a value of each parameter included in the template based on a difference between the amount and a feature amount of the original image to which the template selected by the template selection unit is applied, and the template conversion unit. Image processing is executed by applying the converted parameter value to the processing target image An image processing apparatus comprising: the image processing unit.

  According to a second aspect of the present invention, an image processing template, information on an original image to which the template is applied, a template table that stores information relating to feature amount used in the template in association with each other, A template selection unit that selects the template to be applied; an original image acquisition unit that acquires a plurality of original images to which the template selected by the template selection unit is applied; and a plurality of images acquired by the original image acquisition unit A synthesized image generating unit that synthesizes an original image and generates a synthesized image that approximates the processing target image, and each of the templates included in the template based on a synthesis ratio of the original image in the synthesized image generated by the synthesized image generating unit. A template conversion unit that adjusts parameter values, and the template conversion unit An image processing apparatus characterized by the value of the parameter and an image processing unit that executes image processing by applying to the processing target image.

  According to a third aspect of the invention, the computer applies the image processing template, the information of the original image to which the template is applied, and the feature amount information used in the template in association with each other, and the image processing target image Based on the difference between the step of selecting a template, the step of extracting the feature amount of the processing target image, the feature amount of the extracted processing target image, and the feature amount of the original image to which the selected template is applied And converting each parameter value included in the template, and executing the image processing by applying the converted parameter value to the processing target image. This is an image processing method.

  According to a fourth aspect of the present invention, a computer stores an image processing template, information on an original image to which the template is applied, and feature amount information used in the template in association with each other in a storage unit; The step of selecting the template to be applied, the step of acquiring a plurality of original images to which the selected template is applied from the storage unit, and synthesizing the acquired plurality of original images to generate a composite image that approximates the processing target image A step of converting each parameter value included in the template based on a composition ratio of the original image in the composite image, and applying the converted parameter value to the processing target image. And a step of executing image processing.

  With the image processing apparatus and the image processing method of the present invention, it is possible to easily and quickly adjust the parameter values of the predefined image processing template in consideration of the difference between the image to which the template is applied and the image to be processed. It is possible to obtain a desired analysis image.

The figure which shows the whole structure of the image processing apparatus 100 The figure which shows the function structure of the image processing apparatus 100, and the data which the memory | storage part 103 hold | maintains The figure which shows an example of the parameter table 31 The figure which shows an example of the template table 32 The figure which shows an example of the landmark table 33 The figure which shows an example of the template adjustment data 34 The figure which shows an example of the template conversion data 35 between models Diagram explaining adjustment of parameter value via model The flowchart explaining the whole flow of the analysis image generation process which the medical image processing apparatus 100 performs The flowchart explaining the flow of the template conversion process in 1st Embodiment. The figure which shows an example of the template table 32a in 2nd Embodiment. The flowchart explaining the flow of the template conversion process in 2nd Embodiment The figure explaining adjustment of the parameter value based on the feature-value difference of several images The flowchart explaining the flow of the template conversion process in 3rd Embodiment The figure explaining adjustment of the parameter value based on the composition ratio of a composite image

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

[First Embodiment]
First, the configuration of an image processing system 1 to which the image processing apparatus 100 of the present invention is applied will be described with reference to FIG.

  As shown in FIG. 1, the image processing system 1 includes an image processing device 100 having a display device 107 and an input device 109, an image database 111 connected to the image processing device 100 via a network 110, and an image photographing device 112. With.

  An image processing apparatus 100 according to the present invention is a computer that performs processing such as image generation and image analysis. As shown in FIG. 1, the image processing apparatus 100 includes an external device such as a CPU 101 (Central Processing Unit) 101, a main memory 102, a storage device 103, a communication interface (communication I / F) 104, a display memory 105, and a mouse 108. Interface (I / F) 106, and each unit is connected via a bus 113.

  The CPU 101 calls a program stored in the main memory 102 or the storage device 103 to the work memory area on the RAM of the main memory 102 and executes the program, drives and controls each unit connected via the bus 113, and the image processing apparatus Various processes performed by 100 are realized.

  Further, the CPU 101 applies an image processing condition defined in a template created for an individual at an arbitrary timing to another processing target image, and generates an analysis image (see FIGS. 9 and 10). Execute. Details of the analysis image generation processing will be described later.

  The main memory 102 includes a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. The ROM permanently holds a computer boot program, a program such as BIOS, data, and the like. The RAM temporarily stores programs, data, and the like loaded from the ROM, the storage device 103, and the like, and includes a work area used by the CPU 101 for performing various processes.

  The storage device 103 is a storage device that reads and writes data to and from an HDD (hard disk drive) and other recording media, and stores a program executed by the CPU 101, data necessary for program execution, an OS (operating system), and the like. . As for the program, a control program corresponding to the OS and an application program are stored. Each of these program codes is read by the CPU 101 as necessary, transferred to the RAM of the main memory 102, and executed as various processing execution units.

The communication I / F 104 includes a communication control device, a communication port, and the like, and mediates communication between the image processing apparatus 100 and the network 110. The communication I / F 104 performs communication control with the image database 111, another computer, or an image capturing apparatus 112 such as an X-ray CT apparatus or an MRI apparatus via the network 110.
The I / F 106 is a port for connecting a peripheral device, and transmits / receives data to / from the peripheral device. For example, a pointing device such as a mouse 108 or a stylus pen may be connected via the I / F 106.

  The display memory 105 is a buffer that temporarily accumulates display data input from the CPU 101. The accumulated display data is output to the display device 107 at a predetermined timing.

  The display device 107 includes a display device such as a liquid crystal panel and a CRT monitor, and a logic circuit for executing display processing in cooperation with the display device, and is connected to the CPU 101 via the display memory 105. The display device 107 displays display data stored in the display memory 105 under the control of the CPU 101.

  The input device 109 is an input device such as a keyboard, for example, and outputs various instructions and information input by the operator to the CPU 101. The operator interactively operates the image processing apparatus 100 using external devices such as the display device 107, the input device 109, and the mouse 108.

  The network 110 includes various communication networks such as a LAN (Local Area Network), a WAN (Wide Area Network), an intranet, and the Internet, and connects the image database 111, a server, other information devices, and the like to the image processing apparatus 100. Mediate.

  The image database 111 accumulates and stores image data photographed by the medical image photographing device 112. In the image processing system 1 shown in FIG. 1, the image database 111 is connected to the image processing apparatus 100 via the network 110, but the image database 111 is provided in, for example, the storage device 103 in the image processing apparatus 100. May be.

Next, the functional configuration and data configuration of the image processing apparatus 100 according to the present invention will be described with reference to FIG.
As illustrated in FIG. 2, the CPU 101 of the image processing apparatus 100 includes an image acquisition unit 21, a template selection unit 22, a feature amount extraction unit 23, a template conversion unit 24, and an image processing unit 25. The storage device 103 stores a parameter table 31, a template table 32, a landmark table 33, template adjustment data 34, inter-model template conversion data 35, and the like.

  The image acquisition unit 21 reads an image to be processed (hereinafter referred to as a processing target image) from the storage device 103 and stores it in the RAM (main memory 102).

  The template selection unit 22 selects a template to be applied to the processing target image.

A template defines a group of parameters used in image processing and their parameter values. FIG. 3 shows an example of a parameter table 31 that defines a parameter group included in each template. As shown in FIG. 3, the parameter table 31 stores image processing parameters P1 _A , P2 _A , P1 _B , P2 _B and their values (parameter values) for each of the templates T _A , T _B ,. . The image processing parameters are various image processing items such as image enlargement, rotation, density value, and the like.

  The template selection unit 22 selects a template to be applied to the processing target image from a plurality of templates previously created by the operator and stored in the storage device 103, such as a template table 32 shown in FIG. . The selection of the image processing template by the template selection unit 22 is performed by a user operation via the input device 109, for example.

FIG. 4 shows an example of the template table 32. The template table 32 stores image processing templates and related information about each template. The related information includes an image to which the template is applied (hereinafter referred to as an original image) and landmark information (feature information) that is referred to when the template is applied. For example, in the example of FIG. 4, the template T _A and the original image IMG _A and landmark information L _A is stored in association. This indicates that the image processing is performed on the basis of the landmark information L _A when applying the template T _A with respect to the original image IMG _A. Similarly, the template T _B stores the original image IMG _B and landmark information L _B in association with each other, the template T _C stores the original image IMG _C and landmark information L _C in association with each other, and the template T _X Stores the original image IMG _X and the landmark information L _X in association with each other.

The landmark information L _A , L _B , L _C ,... Is a data group that represents the feature amount of the image. Each landmark information L _A , L _B , L _C ,... Defines what amounts are used as feature amounts m1, m2,... For typical parts (bones, organs, etc.) included in the image. And stored in the storage device 103 like the landmark table 33 shown in FIG. Examples of the feature quantities m1, m2,... Include quantities related to the orientation and size of the organ (cross-sectional area, circumference, volume, shape, short axis, long axis, center of gravity, etc.), concentration values, and the like. In the example shown in FIG. 5, the landmark information Lx includes feature amounts m1 and m2. Further, for example, the landmark information _{L Y,} include feature quantity m1, m3.

  The feature amount extraction unit 23 extracts a feature amount used in the template from the processing target image. Which feature value is used in the image processing template is determined in the template table 32 (see FIG. 4) as landmark information. Further, the amount of feature included in each landmark information is defined in the landmark table 33 (see FIG. 5).

  The template conversion unit 24 is included in the template based on the difference between the feature amount of the processing target image extracted by the feature amount extraction unit 23 and the feature amount of the original image to which the template selected by the template selection unit 22 is applied. Adjust the value of each parameter.

  In the first embodiment, the template conversion unit 24 performs template conversion via a pre-defined model image. That is, the template conversion unit 24 identifies the original model and the first model image that approximates the original image. Further, the template conversion unit 24 specifies the processing target image and the second model image that approximates the processing target image. First, the parameter value included in the template selected by the template selection unit 23 based on the difference in landmark feature between the original image and the first model image is converted for the first model image. A conversion function used for template conversion between the original image and the first model image is stored in the storage device 103 as template adjustment data 33 shown in FIG.

  The model image is image data that is statistically created and has a skeleton and an organ that are ideally anatomically arranged. For example, they are classified by age, sex, body type, etc., such as infants, children, adult men, and adult women.

The conversion functions F _A (a), F _B (a), F _C (a), and F _X (a) shown in FIG. 6 are the model images ModelA, ModelB, ModelC, and ModelX that are most similar to the original image and the original image. This is a function for matching landmark information (features) between. Specifically, the conversion function is data for adjusting each parameter value (enlargement, reduction, rotation, density value, etc.) included in the template based on the difference in the landmark feature quantity between the original image and the model image. Is a group.

  When the parameter value of the template is adjusted for the first model image, the template conversion unit 24 then converts the parameter value adjusted for the first model image into a second model image that approximates the processing target image. . A conversion function used for parameter conversion between the first model image and the second model image is stored in the storage device 103 as the inter-model template conversion data 35 shown in FIG.

  The inter-model template conversion data 35 is a conversion function for matching landmark information (feature amount) between a certain model image and another model image. Each conversion function of the inter-model template conversion data 35 is calculated and stored in advance based on the feature amount difference between the models.

  Note that the variable a of the conversion function F (a) shown in FIGS. 6 and 7 represents the type of parameter. When there are a plurality of parameters to be adjusted, a plurality of conversion functions are stored as template adjustment data 34 (FIG. 6) or inter-model template conversion data 35 (FIG. 7) for each parameter. In the following description, it is assumed that one parameter included in the template is adjusted using one conversion function F (a), but each conversion function for adjusting a plurality of parameters is defined respectively. Also good. In addition, when a plurality of parameters can be comprehensively adjusted with one conversion function, they may be expressed as a conversion function with a plurality of parameters as variables.

  The template conversion unit 24 converts the parameter value converted for the second model image for the processing target image based on the feature amount difference between the second model image and the processing target image.

  In order to speed up the specification of the model in the template conversion unit 24, it is desirable that the model image is stored in a hierarchical manner based on patient attributes and feature amounts of the model image.

  For example, the model image is preferably hierarchized based on medical information (examination information: gender, age, body type, etc.). By hierarchizing based on the medical information, it is possible to narrow down the medical information registered before photographing the processing target image to a model image that approximates the processing target image to some extent. By obtaining the correlation rate and the similarity from the model image narrowed down to some extent and specifying the most approximate model image, it is possible to speed up the process of specifying the most approximate model.

  Further, the model image may be hierarchized based on the feature amount of the part (landmark) of the model image. When creating a model image, the model image is grouped for each part based on the values of feature quantities (cross-sectional area, perimeter, volume, shape, short diameter / long diameter, center of gravity, etc.). For example, the cross-sectional area S is grouped such that small: 1 ≦ S <9, medium: 10 ≦ S <19, large: S> 20. The template conversion unit 24 may specify a model image that approximates a model image that has the largest number of various feature amounts extracted from the original image or the processing target image and belongs to the common group.

  As described above, the image processing apparatus 100 according to the present embodiment uses the conversion function (template adjustment data 34) used in the past or the template conversion data 35 between models obtained in advance as a processing target image. Convert to be applicable. Thereby, also when using the template applied to the image of the patient from which a physique, age, sex, etc. differ completely, the time which conversion of a template requires can be shortened. Although it is extremely difficult to apply a template created and stored by a user to images with completely different classifications, template conversion can be performed at high speed via an ideal model image created in advance. For this reason, an analysis image can be obtained at high speed.

FIG. 8 is a diagram for explaining the relationship between the feature amount of the image and the parameter value.
When the parameter P _X included in the known template T _X applied to the original image IMG _X is adjusted for the processing target image IMG _Z , the template conversion unit 24 first selects a model image Model _X that approximates the original image IMG _X. Identify. Then, the parameter value P _X in the template T _X is converted to P _α based on the difference between the feature amount of the model image ModelX and the feature amount of the original image IMG _X. This is performed for all parameters, it converts the template T _X to the template T _alpha for the first model image. For this conversion, the conversion function F _X (a) of the template adjustment data 34 shown in FIG. 6 is used.

Next, the template conversion unit 24 identifies the model image ModelZ that approximates the processing target image IMG _Z , and sets the parameter value P in the template T _α based on the difference between the feature amount of the model image ModelZ and the feature amount of the model image ModelX. _α is converted to P _β . This is performed for all parameters, it converts the template T _alpha to the template T _beta for the second model image. For this conversion, the conversion function F _{X → Z} (a) of the inter-model template conversion data 35 shown in FIG. 7 is used.

Further, the template conversion unit 24 converts the parameter value P _M2 in the template T _β into P _Z based on the difference between the feature amount of the model image ModelZ and the feature amount of the processing target image IMG _Z. This is performed for all parameters, it converts the template T _beta to the template Tz for the processing target image. For this conversion, an inverse function F _Z (a) ⁻¹ of the conversion function F _Z (a) related to the second model image ModelZ stored in the template adjustment data 34 shown in FIG. 6 is used.

  The image processing unit 25 applies the template converted by the template conversion unit 24 to the processing target image, and executes image processing. As a result, an analysis image can be generated by performing image processing on the processing target image using a template in which each parameter included in the template applied to another image in the past is adjusted for the processing target image.

Next, the operation of the image processing apparatus 100 will be described.
The CPU 101 of the image processing apparatus 100 reads out the program and data related to the analysis image generation processing shown in FIGS. 9 and 10 from the main memory 102 and executes processing based on the program and data.

  At the start of execution of the analysis image generation process shown in FIG. 9, the image data to be processed is fetched from the image database 111 or the like via the network 110 and the communication I / F 104 and stored in the storage device 103 of the image processing apparatus 100. It is assumed that

  First, the CPU 101 of the image processing apparatus 100 acquires a processing target image (step S101). Next, the CPU 101 receives selection of a template to be applied to the processing target image (step S102). In step S <b> 102, the CPU 101 may read out selectable templates from the storage device 103, display a list on the display device 107, and accept a selection by a user operation from the input device 109. When a template is selected in step S102, the CPU 101 extracts a feature amount of the processing target image (step S103).

  Which feature amount is extracted from the processing target image in step S103 differs depending on the template selected in step S102. The CPU 101 refers to the template table 32 stored in the storage device 103 and acquires landmark information used in the selected template. In addition, the CPU 101 refers to the landmark table 33 stored in the storage device 103 and acquires which feature amount is extracted from the acquired landmark information. The CPU 101 extracts each feature amount defined by landmark information from the processing target image.

Next, the CPU 101 executes template conversion processing (step S104). In the template transformation processing in step S104, CPU 101 is a template _{T X} selected in step S102, converts the value of the parameter included in the template _{T X} to be applicable to the processing target image IMG _Z.

  Details of the template conversion processing in step S104 will be described with reference to the flowchart of FIG.

In the template conversion process, first, the CPU 101 is based on a feature amount difference from a model image ModelX (first model image) that approximates the original image IMG _X to which the template T _X selected in step S102 is applied to the original image IMG _X. Te adjusts the parameter values included in the template _{T X} (step S201).

In the process of step S201, CPU 101 obtains the associated information about the template _{T X} selected from the template table 32 in step S102. Then, a conversion function F _X (a) for matching the landmark information (feature amount) of the original image IMG _{X to} which the template T _X is applied from the template adjustment data 33 (see FIG. 6) and the first model image IMG _X is obtained. get. Then, using the obtained conversion function F _X (a), the template T _X applied to the original image IMG _X is converted into the template T _α for the first model image IMG _X.

  Next, the CPU 101 specifies a second model image ModelZ that approximates the processing target image IMGZ (step S202).

CPU101 converts the template T _alpha for transformed first model image at step S201 for the second model image ModelZ (step S203). Specifically, the CPU 101 acquires an inter-model template conversion function F _{X → Z} (a) between the first model image ModelX and the second model image Model Z from the inter-model template conversion data 34 (see FIG. 7). Then, using the inter-model template conversion function F _{X → Z} (a), the template T _α for the first model image is converted into the template T _β for the second model image ModelZ.

CPU101 based on the difference between the feature quantity of the second model image ModelZ approximating to the processing target image IMG _Z processed image IMG _Z, processing the parameter values included in the template T _beta for the second model image object image obtaining the template _{T Z} for converting them suitable for IMG _Z (step S204). In the process of step S204, the CPU 101 acquires the conversion function F _Z (a) associated with the second model image ModelZ from the template adjustment data 34, and applies the inverse function of F _Z (a) to An adjustment value of a parameter from the image ModelZ to the processing target image IMGZ is estimated. CPU101 as a template a new template _{T Z} after adjustment is stored in the template table 32 in association with the original image IMG _Z or landmark information (features).

Template conversion processing is performed as described above, the template _{T Z} for the target image IMG _Z is required, performs the image processing of the processing target image IMG _Z using the template _{T Z} CPU 101 is determined (step S105 ). Then, the image obtained by the image processing is stored in the storage device 103 and displayed on the display device 107 (step S106), and the series of image processing ends.

  The template conversion processing according to the first embodiment does not adjust a template created in advance based on an ideal reference image, but a template created and stored by a user at an arbitrary timing for an image of a patient. Is suitable for applying to another image. When applying the same template to an image with a large difference from the original image, it is extremely difficult to adjust the parameter value, but there are a model that approximates the original image and a model that approximates the image to be processed. If the parameter value is adjusted through the method, a known conversion function can be used, so that the template can be converted with a smaller adjustment amount. Since data necessary for template conversion between models can be created in advance and stored in the storage device 103, processing time required for template conversion can be shortened.

[Second Embodiment]
Next, a second embodiment of the image processing apparatus 100 according to the present invention will be described with reference to FIGS. In the following description, the description of the same configuration and operation as in the first embodiment will be omitted, and the same reference numerals as those in the first embodiment will be given.

  In the second embodiment, the image processing apparatus 100 adjusts parameter values in the template without using model information in the template conversion process.

FIG. 11 is an example of the template table 32a held by the image processing apparatus 100 according to the second embodiment.
The template table 32a according to the second embodiment includes, as related information about each image processing template, an original image to which the template is applied, landmark information referred to when the template is applied, The adjusted post-adjustment parameter value is stored in association with the original image.

For example, in the example of FIG. 11, the template T _X, together with the original image IMG _A and landmark information L _A as related information, adjusted parameter value Pa is stored in association. Further, the same template T _X, together with the original image IMG _B and landmark information L _B as another related information, adjusted parameter value Pb is stored in association.

  The template conversion process in the second embodiment will be described with reference to FIG.

CPU101 first searches the original image to which the template _{T X} selected in Step S102 of FIG. 9 (step S301). For example, as the original image to which the template _{T X, IMG A, IMG B} , it is assumed that IMG _C is searched. The parameter values Px each original image _IMG A included in the template _{T X,} IMG B, are adjusted respectively according to the feature amount of IMG _C, and is actually applied converted parameter values Pa, Pb, Pc is Shall.

CPU101 multiple original images _IMG A _retrieved, IMG B, acquires the characteristic amount of the landmark information IMG _C from the template table 32a (see FIG. 11) (step S302). Further, the CPU 101 obtains a difference between the feature amount of the landmark information extracted from the processing target image IMG _Z and the feature amount of the landmark information of the plurality of original images IMG _A , IMG _B , and IMG _C searched in step S302 ( Step S303).

Based on the size of the difference between the processing target image IMG _Z and the landmark information of the searched plurality of original images IMG _A , IMG _B , and IMG _C , the CPU 101 determines the searched original images IMG _A , IMG _B , and IMG _C A weighting coefficient to be applied to each applied parameter value Pa, Pb, Pc is obtained (step S304).

In step S304, the differences S _A , S _B , S _C between the feature values m _A , m _B , m _C of the original images IMG _A , IMG _B , IMG _C and the feature values m _z of the processing target image IMG _Z are S _a <If the relationship of _S B _{<S C} is the weight coefficient W1, W2, W3 multiplying converted parameter values Pa applied to each image, Pb, the Pc is W1>W2> the relationship W3.

The CPU 101 acquires the adjusted parameter values Pa, Pb, and Pc applied to the original images IMG _A , IMG _B , and IMG _C from the template table 32a (step S305), and the weighting factors W1, W2, and the weight coefficients calculated in step S304. The parameter value Pz to be applied to the processing target image IMG _Z is calculated by weighting and adding W3 to the adjusted parameter values Pa, Pb, and Pc (W1 · Pa + W2 · Pb + W3 · Pc) (step S306).

CPU101 executes each for each parameter in the template the processing of step S305~ step S306, converts the template _{T X} to the template _{T Z} for the target image IMG _Z.

FIG. 13 shows the relationship between the weighting coefficient and the difference between the image feature amounts.
The difference in feature amount between the processing target image IMG _Z and the original image IMG _A is S _A , the difference in feature amount between the processing target image IMG _Z and the original image IMG _B is S _B , the processing target image IMG _Z and the original image IMG _c the difference in feature amount between the case of _{S C,} parameter values Pa applied to the original image, Pb, W1>W2> weighted addition respectively the weighting factors W1, W2, W3 having the relationship W3 against Pc. Thereby, the parameter value Pz applied to the processing target image IMG _Z is calculated.

As described above, in the template conversion processing according to the second embodiment, the CPU 101 determines the parameters included in the template based on the difference between the feature amounts of the plurality of original images and the processing target image to which the same template is applied. A weighting factor to be multiplied by is calculated. Then, the parameter value Pz to be applied to the processing target image IMG _Z is obtained by performing weighted addition of the adjusted parameter value applied to each original image with the calculated weight coefficient.

  Thereby, template conversion in consideration of individual differences can be performed with a simple configuration having no model. Since the parameter value applied to the processing target image is adjusted based on the difference between the feature amounts of the plurality of images, the accuracy of the analysis image can be improved.

[Third Embodiment]
Next, a third embodiment of the image processing apparatus 100 according to the present invention will be described with reference to FIGS.

The image processing apparatus 100 according to the third embodiment converts a template without using model information in the template conversion process. The image processing apparatus 100 searches for a plurality of original images to which the same template is applied, and generates a composite image that approximates the processing target image IMG _Z by combining the plurality of original images. Based on the composition ratio of the original image in the composite image, the parameter value included in the template is adjusted and applied to the processing target image.

  In the third embodiment, the image processing apparatus 100 holds a template table 32a similar to the template table 32a in the second embodiment (see FIG. 11).

  In other words, in the third embodiment, the image processing apparatus 100 includes, as related information about each image processing template, the original image to which the template is applied, the landmark information referred to when the template is applied, and the template conversion unit 24. And a template table 32a in which the adjusted parameter values converted or adjusted by the above are stored in association with the original image.

A template conversion process according to the third embodiment will be described with reference to FIG.
CPU101 first searches the original image to which the template _{T X} selected in Step S102 of FIG. 9, to acquire the image data of the searched original image (step S401). For example, the original image _IMG A was applied the template _{T X,} IMG B, if the IMG _C is searched to obtain each of these original image _{IMG A,} IMG B, the image data IMG _C from the storage device 103.

The CPU 101 generates a composite image of the plurality of original images IMG _A , IMG _B , and IMG _C acquired in step S401 (step S402). In step S <b> 402, the CPU 101 generates a plurality of composite images with different composition ratios by changing the composition ratios in various ways.

Next, the CPU 101 specifies a composite image that most closely approximates the processing target image IMG _Z from among a plurality of composite images having different composite ratios (step S403). The most approximate image in step S403 may be specified by arithmetic processing of the CPU 101, or the operator may select from the plurality of composite images created in step S402 via the input device 109. Also good. When performing the calculation process, for example, a difference between a plurality of synthesized images and a processing target image may be taken and an image having the smallest difference may be selected.

Next, the CPU 101 acquires adjusted parameter values Pa, Pb, and Pc applied to the original images IMG _A , IMG _B , and IMG _C from the template table 32a (step S404), and the original image in the composite image specified in step S403. Each adjusted parameter value Pa, Pb, Pc is weighted and added at the composite ratio of IMG _A , IMG _B , and IMG _C. For example, when the composite ratio of the original images IMG _A , IMG _B , and IMG _C in the composite image that most closely approximates the processing target image IMG _Z is Wa: Wb: Wc, this composite ratio is used as the weighting factor. Then, the adjusted parameter values Pa, Pb, and Pc applied to the original images IMG _A , IMG _B , and IMG _C are respectively multiplied by weighting factors Wa, Wb, and Wc and added. Thereby, the parameter value Pz to be applied to the processing target image IMG _Z is calculated (step S405).

CPU101 executes each for each parameter in the template the processing of step S404~ step S405, converts the template _{T X} to the template _{T Z} for the target image IMG _Z.

FIG. 15 shows the calculation of the weighting factor by the composition ratio of the original image.
When the original images IMG _A , IMG _B , and IMG _C to which the selected template T _X is applied are searched, the CPU 101 uses these original images IMG _A , IMG _B , and IMG _C as various composition ratios (composition ratio A, composition ratio B). ,...) And composite images 5A, 5B,. Then, a composite image that most closely approximates the processing target image IMG _Z is specified from these composite images 5A, 5B,. For example, when the synthesized image 5B synthesized at the synthesis ratio B is closest to the processing target image IMG _Z , the synthesis ratio B (Wa: Wb: Wc) of the synthesized image 5B is multiplied by each parameter value Pa, Pb, Pc. The weight coefficient. The CPU 101 multiplies the parameter values Pa, Pb, and Pc applied to each original image by weighting factors Wa, Wb, and Wc and adds them. Thereby, the parameter value Pz applied to the processing target image IMG _Z is calculated.

  As described above, in the template conversion process according to the third embodiment, the CPU 101 generates a composite image that approximates the processing target image from a plurality of original images to which the same template is applied, and the template is based on the composite ratio. A weighting factor to be multiplied by the parameter included in is calculated. Then, the parameter value Pz applied to the processing target image IMGZ is obtained by performing weighted addition of the adjusted parameter value applied to each original image using the weighting coefficient calculated according to the composition ratio.

  As a result, even when there is no model image or inter-template template conversion data, or when the relationship between the template parameter and the landmark feature quantity cannot be expressed by a linear conversion function, a composite image that approximates the image to be processed can be obtained. By making it, the change of the template can be estimated from the composition ratio.

[Fourth Embodiment]
Next, a fourth embodiment of the image processing apparatus 100 according to the present invention will be described.

When the processing target image IMG _Z includes a plurality of parts, it is difficult to create a template that can handle all of the parts. For this reason, the image processing apparatus 100 according to the fourth embodiment divides an image into a plurality of parts when the image to be processed includes a plurality of parts, and applies an optimal template to each part. The optimum template for each part is determined by the feature amount of the landmark included in each part.

  Moreover, the template applied to each divided part is subjected to template conversion by any one of the methods in the first to third embodiments. That is, appropriate template conversion is performed according to the landmark feature amount of each part of the processing target image. When a template suitable for each part is created, the CPU 101 applies the template to each part, performs image processing, and creates an analysis image for each part. Thereafter, a final analysis image is obtained by synthesizing the analysis images created for each part.

  Thereby, even if the image includes a plurality of parts, an analysis image can be obtained by applying an appropriate template to each part. In addition, since the template applied to each part is adjusted according to the feature amount of the landmark of each part, it is possible to create an analysis image with high accuracy in consideration of individual differences among patients.

  The preferred embodiments of the image processing apparatus according to the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to such examples. It will be apparent to those skilled in the art that various changes or modifications can be conceived within the scope of the technical idea disclosed in the present application, and these naturally belong to the technical scope of the present invention. Understood.

1... Image processing system 100... Image processing apparatus 101.
102... Main memory 103... Storage device 104... Communication I / F
105... Display memory 106... I / F
107... Display device 108... Mouse 109... Input device 110... Network 111. Image capturing device 21 Image acquisition unit 22 Template selection unit 23 Feature amount extraction unit 24 Template conversion unit 25 Image processing unit 31 Parameter table 32 Template table 33・ Landmark table 34 ・ ・ ・ Template adjustment data 35 ・ ・ ・ Template conversion data between models

Claims (3)

A storage unit that stores a template table that holds an image processing template, information on an original image to which the template is applied, and feature amount information used in the template;
A template selection unit for selecting the template to be applied to the processing target image;
A feature amount extraction unit for extracting a feature amount of the processing target image;
Based on the difference between the feature amount of the processing target image extracted by the feature amount extraction unit and the feature amount of the original image to which the template selected by the template selection unit is applied, each parameter included in the template A template conversion unit for converting values;
An image processing unit that executes image processing by applying the value of the parameter converted by the template conversion unit to the processing target image,
The storage unit
Inter-template template conversion data for converting the template between model images based on the correlation between the feature amounts of a plurality of model images, and parameters included in the template based on the correlation between the feature amounts of the original image and the model image Template adjustment data for adjusting the value of
The template conversion unit
The parameter value included in the template selected by the template selection unit is adjusted for the first model image using the template adjustment data, and the parameter value adjusted for the first model image is converted into the inter-model template conversion data. Is converted into a second model image that approximates the processing target image, and the parameter value converted for the second model image is converted into a processing target image based on a difference in feature quantity between the second model image and the processing target image. images processor you and adjusting to use. A storage unit that stores a template table that holds an image processing template, information on an original image to which the template is applied, and feature amount information used in the template;
A template selection unit for selecting the template to be applied to the processing target image;
An original image acquisition unit that acquires a plurality of original images to which the template selected by the template selection unit is applied from the template table;
Combining a plurality of original images acquired by the original image acquisition unit, and generating a composite image that approximates the processing target image; and
A template conversion unit that adjusts each parameter value included in the template based on a composition ratio of the original image in the composite image generated by the composite image generation unit;
An image processing unit that executes image processing by applying the value of the parameter converted by the template conversion unit to the processing target image;
An image processing apparatus comprising: Computer
Storing an image processing template, information of an original image to which the template is applied, and feature amount information used in the template in a storage unit;
Selecting the template to be applied to the processing target image;
Acquiring a plurality of original images to which the selected template is applied from the storage unit;
Synthesizing a plurality of acquired original images and generating a synthesized image approximating the processing target image;
Converting the value of each parameter included in the template based on a composition ratio of the original image in the composite image;
Performing image processing by applying the converted parameter value to the processing target image;
An image processing method comprising:

Images