JP2013223620A - Apparatus, method, and program for registration processing of medical image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus, method, and program for registration processing of a medical image which are capable of excellently executing registration processing of a morphological image and a functional image.SOLUTION: A registration processing apparatus includes: image data obtaining means 11A and 11B each obtaining morphological image data and functional image data; image data storing means 12A and 12B each storing the morphological image data and the functional image data; a morphological data extracting means 13 for extracting image data of a desired organ from the morphological image data; a morphological image data binarizing means 14 for converting the morphological image data into predetermined high signal values in the desired organ region, and into zeros in a region other than the organ region; and a registration processing means 15 for executing registration processing for the binarized morphological image data and the functional image data from the image data storing means 12B.

Description

  The present invention relates to a medical image registration processing apparatus, method, and program capable of determining what functional state a target portion is in by associating an image mainly having morphological information with an image mainly having functional information It is about.

  In general, a medical image obtained by CT or MRI can represent morphological information (morphological image) of each part of an organ. On the other hand, the CT site is imaged using a contrast agent, or a desired substance is imaged using SPECT or PET, which is a technique for examining the dynamics and organ accumulation by administering a radioactive substance into the blood. Thus, desired function information (function image) of each part can be obtained.

  Therefore, conventionally, a morphological image and a functional image of the same part are acquired, and registration processing (alignment processing: the same applies hereinafter) is performed so as to correct the deviation. There is known a test method that recognizes in what functional state or in which part drug accumulation due to cancer is present (see Patent Document 1 and Non-Patent Document 1 below).

  The above correction processing is performed by calculating the difference between the superimposed morphological image and the functional image for each coordinate position, and moving the two images in parallel or rotating to each other so that the total error amount is minimized. I was trying to set the position.

  Such a conventional technique is particularly effective when the contrast of two images to be superimposed is close to each other.

Japanese Patent Laid-Open No. 10-282268

Medical image processing technology that supports image-guided radiation therapy, Hidetaka Arimura (Kyushu University Graduate School of Medicine, Department of Health Sciences, Department of Medical Quantum Radiation Science), Japanese Society of Medical Physics, Training Materials , Pp25-44, 2010)

  However, when registration is performed by superimposing morphological images such as CT and MRI and functional images such as SPECT and PET, the functional image generally has a feature that the signal value of only the target part is displayed high. In contrast, morphological images do not have such features.

  Therefore, in the above-described prior art, when registration is performed by superimposing the morphological image and the functional image, the size of the error is not necessarily an index indicating the degree of coincidence of the images, and the registration is good. It was difficult to do.

  The present invention has been made in view of such circumstances, and an object thereof is to provide a medical image registration processing apparatus, method, and program capable of satisfactorily performing registration processing of a morphological image and a functional image. To do.

In order to achieve the above object, a medical image registration processing apparatus according to the present invention includes:
A medical image registration processing apparatus that performs registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific subject,
Image data storage means for storing the morphological image data and the functional image data;
Morphological image data extracting means for extracting morphological image data of a desired target region from the morphological image data;
The extracted morphological image data is divided inside and outside the desired target area, and each pixel within the area has a relatively high predetermined signal value, while each pixel outside the area is relatively low. Morphological image data binarization means for performing binarization processing of signal values so as to obtain predetermined signal values;
Registration processing means for performing registration processing between the binarized morphological image data and the functional image data;
It is characterized by comprising.

  Here, a morphological image refers to an image that reflects the shape of an object, such as a CT image or an MRI image, while a functional image refers to an RI image (nuclear medicine image) or a part of an MRI image. An image that reflects the functional strength of the target part. However, in this case, the morphological image and the functional image are terms that are relatively used, and broadly, of the two images related to the registration process, an image that more reflects the shape of the target region is used as the morphological image and the target region. An image that more reflects the function strength is referred to as a function image.

  Further, it is preferable that the morphological image data extracting means performs a masking process on any one of the inner and outer regions of the desired target region.

  Further, the morphological image data binarizing means sets each pixel of the image data in the desired target area to a signal value that is maximum in the image data of the functional image, and image data outside the desired target area. It is preferable that each of the pixels is replaced with 0.

  Further, the registration processing means, when aligning the position of the morphological image data and the functional image data, a rigid registration processing for performing relative rotational movement and parallel movement of the two images, It is preferable to perform a non-rigid registration process in which correction is performed by curving the shape of the image from the feature points of both images.

Further, the morphological image data is image data obtained by a CT apparatus or an MRI apparatus,
The functional image data is image data obtained by contrast medium administration in a CT device, image data obtained by contrast agent administration or diffusion weight enhancement in an MRI device, or SPECT (single radiation weighted tomographic image acquisition device) or PET ( Image data obtained by a positron radiation-weighted tomographic image acquisition device) is preferable.

Further, the medical image registration processing method according to the present invention is a medical image processing for performing registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific subject. An image registration processing method comprising:
A morphological image data extracting step of extracting image data of a desired target region from the morphological image data;
A morphological image data binarization step for binarizing the extracted morphological image data depending on whether the target area is inside or outside;
A registration processing step for performing registration processing between the binarized morphological image data and the functional image data is performed in this order.

Furthermore, the medical image registration processing program according to the present invention is a computer program for performing registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific subject. A medical image registration processing program to be executed in
A morphological image data extracting step of extracting image data of a desired target region from the morphological image data;
A morphological image data binarization step for binarizing the extracted morphological image data depending on whether the target area is inside or outside;
A registration processing step for performing registration processing between the binarized morphological image data and the functional image data;
Is executed in the computer.

  The medical image registration processing apparatus, method, and program according to the present invention perform registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific anatomy. When performing, first, a desired target region is extracted from the morphological image data, the extracted morphological image data is binarized, and the binarized morphological image data and the functional image data are Registration processing is performed.

  That is, in general, when performing registration processing on images reflecting the form, the two images are subtracted and translated in a direction that minimizes the error, or rotated to move the relative positions of the two images. adjust. However, when performing registration processing of the morphological image and the functional image, the functional image has a feature that the signal value of the target portion is displayed high. When the registration process is adjusted using the difference error as an index, reducing the error does not necessarily improve the image alignment.

  On the other hand, according to the registration processing apparatus, method, and program of the present invention, a target region is extracted from a morphological image, and the target region is replaced with a predetermined relatively high signal value. The signal value in the region other than the target region is replaced with a predetermined relatively low signal value. That is, since the functional image is acquired by specifically accumulating the radiopharmaceutical in the target region, the signal value of the region of the target region is the highest. Processing to approximate the value distribution can be performed.

  In this way, since the morphological image and the functional image can be images having similar contrasts with each other, the processing for minimizing the difference value between the two images can be performed, so that the correspondence between the positions of the two images can be obtained. This makes it easy to register both images easily and satisfactorily.

It is a figure which shows concretely the registration processing apparatus which concerns on one Embodiment of this invention. It is a block diagram which shows the hardware constitutions of the registration processing apparatus shown in FIG. It is a block diagram which shows concretely the functional structure of the registration processing apparatus shown in FIG. It is the schematic for demonstrating each image displayed on the image display apparatus shown in FIG. It is a flowchart which shows the registration processing method which concerns on one Embodiment of this invention. It is a figure which shows CT image (morphological image) used in the registration processing method shown in FIG. The binarized morphological image obtained by extracting the morphological image data of the liver region from the morphological image data shown in FIG. 6 and binarizing the extracted morphological image data of the liver region inside or outside the liver region. FIG. It is a figure which shows the SPECT image (functional image) used in the registration processing method shown in FIG. It is a figure which shows the superimposition image obtained by performing a registration process between the functional image shown in FIG. 7, and the binarization form image shown in FIG.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the above-mentioned drawings. First, a medical image registration processing apparatus according to an embodiment of the present invention will be described with reference to FIGS. 1 and 2.

  The medical image registration processing apparatus 1 shown in FIG. 1 includes first medical image data (morphological image) obtained by the first type image diagnostic apparatus and carrying morphological information of a target tissue of a specific subject (specific patient). Data) and second medical image data (functional image data) carrying the functional information of the target tissue of the specific subject obtained by the second type image diagnostic apparatus, and aligning from a computer or the like An image processing device 10, an input device 20 including a mouse and a keyboard, and an output device 30 including an image display device.

  In this embodiment, the first type image diagnostic apparatus is an X-ray CT apparatus, the second type image diagnostic apparatus is a SPECT apparatus, and the target tissue is the liver of a patient suspected of having a liver disease. I will explain to you. In the following description, the first medical image data may be abbreviated as “patient CT” and the second medical image data may be abbreviated as “patient SPECT”.

  A medical image registration processing apparatus 1 shown in FIG. 1 includes an organ or bone to be diagnosed in a medical image space constructed based on three-dimensional medical image data obtained by an image diagnostic apparatus such as an X-ray CT apparatus. It has a function of extracting a region of interest corresponding to a region of interest such as a tumor or creating a color-coded identification image, and includes an arithmetic processing unit 10 made up of a computer, etc., and an input made up of a keyboard 21, mouse 22, etc. The apparatus 20 and the image display apparatus 30 are provided.

  As shown in FIG. 2, the arithmetic processing unit 10 is a DRAM (Dynamic Random Access) in which a CPU (Central Processing Unit) 101 that executes various arithmetic operations such as image processing, a processing program, image data to be processed, and the like are placed. And a chip set 103 that controls input / output of data to / from the CPU 101, the main memory 102, and the like. The image processing apparatus also includes a GPU (Graphic Processing Unit) 105 that configures a display image based on image data and processing instructions sent from the CPU 101, and a VRAM (Video Random Access Memory) 106 that stores a display image configured by the GPU 105. Between the graphics board 104, a storage device 107 including a hard disk, a data bus 108 that mediates the exchange of data between these components, the input device 20, the image display device 30, and the arithmetic processing device 10 And an interface 109 that mediates exchange of data and the like.

  Further, in terms of a functional viewpoint, the arithmetic processing device 10 includes a morphological image data acquisition unit 11A and a functional image data acquisition unit 11B, image data storage units 12A and 12B, and a configuration as illustrated in FIG. The image data extraction unit 13, the morphological image data binarization unit 14, and the registration processing unit 15 are configured.

  These blocks include hardware such as the CPU 101, the main memory 102, the graphics board 104, and the storage device 107 shown in FIG. 2, and various programs executed by the CPU 101 and the GPU 105 (the medical image registration processing program according to the present invention). Each function realized by (including) is made concrete and shown as a component.

  The morphological image data acquisition unit 11A is configured to acquire three-dimensional medical image data obtained by an image diagnostic apparatus such as an X-ray CT apparatus or an MRI apparatus. Hereinafter, a case where the morphological image data acquisition unit 11A is a multi-slice X-ray CT apparatus will be described as an example. In addition, the functional image data acquisition unit 11B is, for example, image data obtained by contrast medium administration in a CT apparatus, image data obtained by contrast medium administration or diffusion weight enhancement in an MRI apparatus, or SPECT (single radiation weighted tomography). Image data obtained by an image acquisition device) or PET (positron radiation-weighted tomographic image acquisition device) is acquired. In the following, a case where the functional image data acquisition unit 11A acquires SPECT image data obtained by SPECT will be described as an example.

  The image data storage units 12A and 12B are configured by a main memory 102 made up of a DRAM (Dynamic Random Access Memory) or the like.

  The morphological image data extraction unit 13 extracts a liver region to be diagnosed in an image space (hereinafter referred to as “voxel space” as appropriate) constructed by CT image data acquired by the image data acquisition unit 11. And a masking processing unit 132 that performs masking processing on a region other than the liver region. In the following, in an image space constructed by SPECT image data, An example of extracting a liver region will be described.

  Further, the morphological image data binarization unit 14 replaces the liver region image data with a signal value that is maximized in the functional image image data, and a region other than the liver region. A specific outside-subject region image conversion unit 142 that replaces the image data with 0 is provided.

  The registration processing unit 15 includes a CT image based on the binarized CT image data of the liver region from the morphological image data binarizing unit 14 and a functional image based on the functional image data from the image data storage unit 12B. It is configured to perform registration processing with a SPECT image of a certain liver region, and includes a rigid registration processing unit 151 and a non-rigid registration processing unit 152.

  The integrated image registered by the registration processing unit 15 is binarized from the morphological image from the image data storage unit 12A, the functional image from the image data storage unit 12B, and the morphological image data binarization unit 14. Together with the morphological image, it is sent to the image display device 30.

  In this embodiment, as shown in FIG. 4, a plurality of display areas (four display areas W1 to W4 are illustrated in FIG. 4) are simultaneously displayed in the image display unit 31. It is configured. That is, the display area W1 is an area in which a CT image that is a morphological image from the image data storage unit 12A is displayed, and the display area W2 is a binarized morphological image from the morphological image data binarization unit 14. Is displayed, the display area W3 is an area where a functional image from the image data storage unit 12B is displayed, and the display area W4 is registered by the registration processing unit 15. It is an area where the integrated image is displayed.

  Next, a medical image registration processing method according to an embodiment of the present invention will be described with reference to the flowchart of FIG. The basic procedure in the medical image registration processing method according to the present embodiment is executed by the above-described medical image registration processing apparatus 1 that operates according to the medical image registration processing program according to an embodiment of the present invention. Is.

  That is, this embodiment method executes a step (S1) of acquiring a morphological image and a functional image, executes a step (S2) of extracting a region of the liver region in the morphological image, and extracts the region extracted from the morphological image. A step of binarizing data (S3) is executed, and a step of registering the binarized morphological image and the functional image (S4) is executed.

  In this embodiment, in order to improve the accuracy of registration processing, after extracting the liver region from the morphological image, a single signal value is applied to the liver region and the other regions in step 3 (S3). The registration processing is executed in a state (pseudo shape) in which the signal value distribution of the morphological image resembles the signal value distribution of the functional image. This is also clear from the fact that the binarized form image shown in FIG. 7 has a shape similar to the functional image shown in FIG.

  Normally, when the registration process is performed on an image reflecting the form, the position adjustment is performed by moving the two images in parallel in the direction that minimizes the error or rotating the images. However, such a method is particularly good when the target is the alignment of two images with similar contrast. For example, when registration processing is executed for the same two images, the parallel movement amount and the rotational movement amount of the images are adjusted so that the difference error becomes zero. However, functional images use radiopharmaceuticals that accumulate specifically in the target organ, so that the signal value of the target organ is the highest, whereas the morphological image has such a signal value distribution. Not. Therefore, when performing registration processing for morphological images and functional images, even if the method of adjusting image integration using the above-described difference error as an index is executed, the fact that the error is low does not necessarily mean that the images match well. It cannot be said that it represents, but the alignment accuracy is poor.

  Therefore, in the present embodiment, the target organ (here, the liver) is extracted from the morphological image, the internal region is replaced with a predetermined signal value, and the signal value of the external region is replaced with 0, thereby the morphological image. Is performed (simulated) with the signal value distribution of the function image. By performing such binarization processing, good results can be obtained even by a method of performing registration processing between two images and performing image integration adjustment using the above-described difference error as an index.

  Hereinafter, the method of this embodiment will be described for each of the above steps.

<1> A morphological image and a functional image are acquired (S1).
As described above, CT image data including liver region data obtained by a multi-slice X-ray CT apparatus is acquired (see morphological image data, see FIG. 6, the crescent-shaped white line region at the bottom of the figure is the subject mounting table ) And SPECT image data including liver region data obtained by SPECT (functional image data in which the function of the liver region is emphasized, see FIG. 8) is acquired. The acquired image data is stored in the corresponding image data storage units 12A and 12B, respectively.

<2> Extracting the region of the liver region from the morphological image (S2)
Here, region extraction is a technique for automatically or semi-automatically recognizing a target region.

  In the present embodiment, the CT image region extraction method includes various methods such as a threshold processing method, region expansion method, watershed method, active contour model method (snakes method, level set method, etc.), and model base method. A well-known method can be adopted.

More specifically, for example, Japanese Patent Application No. 2001-152523 (Japanese Patent Laid-Open No. 2002-345802, Japanese Patent No. 3486615) already disclosed by the present applicant describes a region extraction method for CT images.
In this region extraction method, first, from the data of a plurality of images obtained by capturing substantially the same region of a living body part under different measurement states, for each pixel at a corresponding position between the images, for each image, A measurement value for a predetermined characteristic is determined. Next, in the multi-dimensional feature space with the feature by image as each coordinate axis, the sample distribution of each pixel related to the region of the living body part is obtained, and the degree of correlation between the feature by image with respect to this sample distribution By applying a distribution function that takes into account, the range of the maternal distribution is estimated, and the region of the tissue is extracted. As a result, it is possible to perform the region extraction of the living body part uniquely and stably.

<3> The image data of the region extraction image is binarized in the morphological image (S3).
As described above, the extracted CT image data in the liver region is replaced with the maximum signal value in the image data, and the CT image data outside the extracted liver region is replaced with 0. (See FIG. 7). Thus, in step 2 (S2), the target liver region is extracted from the morphological image, and in step 3 (S3), the signal value other than the liver region is set so that the liver region has a high signal value. By replacing the signal value for each pixel so that it becomes 0, the liver region is set so that the signal value is high while the signal value is low outside the liver region, and the contrast is set to a high state. A process of making an image resemble the signal value distribution of a functional image (hereinafter referred to as a fake process) is performed.

  <4> Registration processing between the binarized form image and the function image is performed (S4).

Here, registration refers to a process of superimposing these images so that the positions of the same object are aligned between different images.
In this process, the transformation function is calculated so that the similar parts (feature points) between the two images are best matched. FIG. 9 shows a result of the registration of the morphological image and the functional image through the registration process.

  General registration procedures are: (1) pre-processing (eg noise attenuation), (2) determination of corresponding feature points between two images, (3) calculation of conversion function based on corresponding feature points, (4) Interpolation after conversion and (5) calculation of similarity between images are performed in this order. In this embodiment, first, rough alignment is performed by rigid registration, and then fine alignment is performed using non-rigid registration.

  That is, here, the rotational movement amount and the parallel movement amount of one of the images are calculated so that the distribution error of the signal value generated between the morphological image and the functional image is corrected by the rigid registration process. Then, based on this calculated value, image processing for reducing the inter-image error between the two is performed.

  That is, alignment is performed assuming that the liver region to be aligned is a rigid body (assuming that the direction and position of the object change, but the shape and size of the object do not change). One motion parameter is estimated and alignment is performed).

  In this way, the coordinate alignment of the CT image (morphological image) and the SPECT image (functional image) is performed using the rigid registration process, but particularly in soft tissue such as the liver, between the two images. Since deformation has occurred, alignment between the two images is performed by absorbing the deformation using a non-rigid registration process.

  The non-rigid registration process in the present embodiment uses a three-dimensional spline interpolation method to correct a signal value distribution error that occurs between a CT image (morphological image) and a SPECT image (functional image). Image processing for reducing the image error between the two images. That is, after the rigid registration process is performed, the corresponding point search is performed again, and local deformation (image blurring or signal error) is expressed by a motion model based on three-dimensional B-spline interpolation. It is calculated by Gauss-Newton method.

  The binarized morphological image (see FIG. 7) is used only for alignment, and after the alignment is performed, it is superimposed on the functional image (see FIG. 8) (for example, addition calculation processing for each pixel) The morphological image is a morphological image (see FIG. 6) stored in the image data storage unit 12A.

  As described above, in this embodiment, since the non-rigid registration process is performed after the rigid registration process is performed, the alignment between the two images is reliably performed.

  As mentioned above, although one Embodiment of this invention was described, this invention is not limited to the said embodiment, A mode can be changed variously.

  For example, in the above embodiment, the case where the morphological image is an X-ray CT image, the functional image is a SPECT image, and the target tissue is a liver is described, but the morphological image may be an MRI image or the like. Even when an image obtained by contrast medium administration on a CT device, an image obtained by contrast agent administration or diffusion weighting on an MRI device, or an image obtained by PET (positron radiation weighted tomographic image acquisition device) is used. The present embodiment can be usefully applied.

  In addition to the liver, target tissues can be organs such as the heart, pancreas, kidney, and spleen, and biological tissues such as the brain, lungs, and tumors. There is no particular difficulty in applying the present invention when another tissue is the target tissue, and the above-described procedure of the present invention can be applied in substantially the same manner regardless of the type of the target tissue.

  In addition to the case where the masking process is performed only on the outside of the target area, the masking process can be performed on the target area (liver area) so that the inside and outside of the target area can be well identified.

DESCRIPTION OF SYMBOLS 1 Medical image registration processing apparatus 10 Image processing apparatus 20 Input apparatus 21 Keyboard 22 Mouse 30 Output apparatus 11A Morphological image data acquisition part 11B Functional image data acquisition part 12A, 12B Image data storage part 13 Morphological image data extraction part 14 Morphological image Data binarization unit 15 Registration processing unit 101 CPU
102 Main memory 103 Chipset 104 Graphics board 105 GPU
106 VRAM
DESCRIPTION OF SYMBOLS 107 Memory | storage device 108 Data bus 109 Interface 131 Specific subject extraction part 132 Masking process part 141 Specific subject internal region image conversion part 142 Specific subject external region image conversion part 151 Rigid body registration processing part 152 Non-rigid body registration processing part W1 ~ W4 display area

In order to achieve the above object, a medical image registration processing apparatus according to the present invention includes:
A medical image registration processing apparatus that performs registration processing between morphological image data carrying morphological information of a target tissue of a specific subject and functional image data carrying functional information of the target tissue of the specific subject. ,
Image data storage means for storing the morphological image data and the functional image data;
Morphological image data extracting means for extracting the region of the target tissue in an image space constructed by the morphological image data;
The image space from which the target tissue region is extracted is divided into the inside and outside of the target tissue region, and each pixel outside the region has a relatively high predetermined signal value. Then, morphological image data binarization means for performing binarization processing of the signal value so as to obtain a relatively low predetermined signal value,
A region of the target tissue in the binarized form image data, and the registration processing means for performing location registration processing to match the said target tissue region in the functional image data,
It is characterized by comprising.

Moreover, the anatomical image data binarizing means and each pixel of the image data in the area of the target tissue, the maximum and becomes the signal value in the image data of the functional image, the image data of the area outside of the target tissue It is preferable that each of the pixels is replaced with 0.

The target tissue is preferably a liver tissue.

Further, the registration processing means, when aligning the position of the target tissue region in the morphological image data and the target tissue region in the functional image data, relative rotation and parallel movement of both the images. It is preferable to perform a rigid registration process for performing the movement and a non-rigid registration process for performing correction by curving the shape of the image from the detected feature points of both images.

Furthermore, the registration method for medical images according to the present invention includes:
A medical image registration processing method for performing registration processing between morphological image data carrying morphological information of a target tissue of a specific subject and functional image data carrying functional information of the target tissue of the specific subject. ,
From image space constructed by the form image data, and forms image data extraction step of extracting the realm of the target tissue,
A morphological image data binarization step that binarizes in the image space in which the region of the target tissue is extracted, depending on whether the region of the target tissue is inside or outside;
A region of the target tissue in the binarized form image data, a registration processing step of performing a registration process to align the area of the target tissue in the functional image data, and carrying out in this order Is.

Furthermore, a medical image registration processing program according to the present invention includes:
Registration process of medical image in which registration processing of morphological image data carrying the morphology information of the target tissue of the specific subject and functional image data carrying the functional information of the target tissue of the specific subject is executed by a computer A program,
In the image space constructed by the form image data, and forms image data extraction step of extracting the realm of the target tissue,
A morphological image data binarization step for binarizing the image space from which the region of the target tissue has been extracted depending on whether the region of the target tissue is inside or outside;
A registration processing step of performing a position registration process to align the a region of the target tissue in the binarized form the said target tissue region functional in the image data the image data,
Is executed in the computer.

Furthermore, the registration method for medical images according to the present invention includes:
A medical image registration processing method for performing registration processing between morphological image data carrying morphological information of a target tissue of a specific subject and functional image data carrying functional information of the target tissue of the specific subject. ,
In the image space constructed by the morphological image data, a morphological image data extracting step for extracting the region of the target tissue;
Dividing the image space that region of the target tissue is extracted in and out of the area before Symbol target tissue, so that the relatively high predetermined signal value at each pixel in the area, whereas the outside the region A morphological image data binarization step for binarizing the signal value so that the pixel has a relatively low predetermined signal value ;
A registration processing step of performing a registration processing for aligning the position of the target tissue region in the binarized morphological image data with the position of the target tissue region in the functional image data is performed in this order. Is.

Furthermore, a medical image registration processing program according to the present invention includes:
Registration process of medical image in which registration processing of morphological image data carrying the morphology information of the target tissue of the specific subject and functional image data carrying the functional information of the target tissue of the specific subject is executed by a computer A program,
In the image space constructed by the morphological image data, a morphological image data extracting step for extracting the region of the target tissue;
The image space from which the target tissue region is extracted is divided into the target tissue region inside and outside, and each pixel inside the region has a relatively high predetermined signal value, A morphological image data binarization step for binarizing the signal value so that the pixel has a relatively low predetermined signal value ;
A registration processing step for performing a registration process for aligning the target tissue region in the binarized morphological image data and the target tissue region in the functional image data;
Is executed in the computer.

Claims (7)

A medical image registration processing apparatus that performs registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific subject,
Image data storage means for storing the morphological image data and the functional image data;
Morphological image data extracting means for extracting morphological image data of a desired target region from the morphological image data;
The extracted morphological image data is divided inside and outside the desired target area, and each pixel within the area has a relatively high predetermined signal value, while each pixel outside the area is relatively low. Morphological image data binarization means for performing binarization processing of signal values so as to obtain predetermined signal values;
Registration processing means for performing registration processing between the binarized morphological image data and the functional image data;
A medical image registration processing apparatus comprising:   The medical image registration processing apparatus according to claim 1, wherein the morphological image data extraction unit performs a masking process on one of the inside and outside of the desired target area.   The morphological image data binarizing means sets each pixel of the image data in the desired target area to a signal value that is maximum in the image data of the functional image, and sets each pixel of the image data outside the desired target area. 3. The medical image registration processing apparatus according to claim 1, wherein each pixel is replaced with 0.   The registration processing means, when aligning the position of the morphological image data and the functional image data, a rigid registration processing for performing relative rotational movement and parallel movement of the two images, and the detected both images The medical image resist according to any one of claims 1 to 3, wherein a non-rigid registration process is performed to correct the image by curving the shape of the image from the feature points of the medical image. Equipment. The morphological image data is image data obtained by a CT apparatus or an MRI apparatus,
The functional image data is image data obtained by contrast medium administration in a CT device, image data obtained by contrast agent administration or diffusion weight enhancement in an MRI device, or SPECT (single radiation weighted tomographic image acquisition device) or PET ( 5. The medical image registration processing apparatus according to claim 1, wherein the image data is obtained by a positron radiation-weighted tomographic image acquisition apparatus. A medical image registration processing method for performing registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific subject,
A morphological image data extracting step of extracting image data of a desired target region from the morphological image data;
A morphological image data binarization step for binarizing the extracted morphological image data depending on whether the target area is inside or outside;
A medical image registration processing method, wherein a registration processing step of performing registration processing between the binarized morphological image data and the functional image data is performed in this order. A medical image registration processing program for causing a computer to execute registration processing between morphological image data carrying morphological information of a specific subject and functional image data carrying functional information of the specific anatomy,
A morphological image data extracting step of extracting image data of a desired target region from the morphological image data;
A morphological image data binarization step for binarizing the extracted morphological image data depending on whether the target area is inside or outside;
A registration processing step for performing registration processing between the binarized morphological image data and the functional image data;
Is executed by the computer. A medical image registration processing program.