KR102148600B1 - Alignment method and apparatus for stack of image based on dynamic programming - Google Patents

Abstract

In the image stack matching method using the dynamic programming method, the image processing apparatus receives a first image stack and a second image stack each including images having a specific order, and the image processing apparatus receives images of the first image stack and Calculating a similarity between all image pairs formed by the images of the second image stack, the image processing apparatus selecting one image and the second image while selecting images from the first image stack in the order Determining a maximum value of a similarity measure for each pair of images in the stack, and the image processing apparatus comprising an image pair of the last image of the first image stack and the image of the second image stack, and the second image stack And matching the first image stack and the second image stack in reverse order starting from the image pair having the largest similarity measure among the image pairs formed by the last image of and the image of the first image stack.

Description

Image stack matching method and image stack matching device using dynamic programming {ALIGNMENT METHOD AND APPARATUS FOR STACK OF IMAGE BASED ON DYNAMIC PROGRAMMING}

The technique described below relates to a matching technique for an image stack.

Medical images such as MRI and CT are composed of a kind of image stack. That is, medical images such as MRI and CT are composed of a plurality of frames. In the process of patient treatment, in general, the time point of photographing is changed or medical images are photographed at regular intervals. Therefore, a plurality of medical images (multiple image stacks) of the same patient have different starting positions or different at the time of capturing.

Korean Patent Registration No. 10-1825719

The plurality of medical images are not automatically aligned based on a specific time point or location. Therefore, the medical staff should check the frames constituting the medical images individually to check the photographing results of the same area in the plurality of medical images.

The technology described below is intended to provide a technique for matching a plurality of images composed of a plurality of image stacks, such as a medical image.

In the image stack matching method using the dynamic programming method, the image processing apparatus receives a first image stack and a second image stack each including images having a specific order, and the image processing apparatus receives images of the first image stack and Calculating a similarity between all image pairs formed by the images of the second image stack, the image processing apparatus selecting one image and the second image while selecting images from the first image stack in the order Determining a maximum value of a similarity measure for each pair of images in the stack, and the image processing apparatus comprising an image pair of the last image of the first image stack and the image of the second image stack, and the second image stack And matching the first image stack and the second image stack in reverse order starting from the image pair having the largest similarity measure among the image pairs formed by the last image of and the image of the first image stack.

The image stack matching device using the dynamic programming method is an input device that receives a first image stack and a second image stack each including images having a series of order, calculates the similarity between two images, and calculates the similarity between the two images, based on the dynamic programming method. A storage device storing a program for calculating a maximum value of a similarity measure between any one image constituting the first image stack and any one image pair constituting the second image stack, and the first image using the program Calculating the similarity between the images of the stack and all pairs of images formed by the images of the second image stack, selecting an image from the first image stack in the order, and selecting any one image and the second image Determine the maximum value of the similarity measure for each pair of images in the stack, and select an image pair with the largest similarity measure among all possible image pairs formed by the last image of the first image stack and the last image of the second image stack. It includes a computing device that matches the first image stack and the second image stack.

In the technique described below, a plurality of medical images are automatically aligned or matched so that a plurality of medical images are composed of images of corresponding regions in a certain order. The technology described below provides a matched medical image so that medical personnel can conveniently and accurately analyze the medical image.

1 is an example of a system for processing a medical image.
2 is an example of matching a plurality of image stacks and image stacks.
3 is an example of matching two image stacks.
4 is an example of a flowchart of a process of matching image stacks using a dynamic programming method.
5 is an example of a process of matching image pairs in an image stack.
6 is an example of a process of matching image pairs based on a similarity measure.
7 is an example of a pseudo code for a process of matching image pairs in an image stack.
8 is an example of a configuration of an image processing apparatus.

The technology to be described below may be modified in various ways and may have various embodiments, and specific embodiments will be illustrated in the drawings and described in detail. However, this is not intended to limit the technology to be described below with respect to a specific embodiment, and it should be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the technology described below.

Terms such as 1st, 2nd, A, B, etc. may be used to describe various components, but the components are not limited by the above terms, only for the purpose of distinguishing one component from other components. Is only used. For example, without departing from the scope of the rights of the technology described below, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component. The term and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

In terms of the terms used in the present specification, expressions in the singular should be understood as including plural expressions unless clearly interpreted differently in context, and terms such as "includes" are specified features, numbers, steps, actions, and components. It is to be understood that the presence or addition of one or more other features or numbers, step-acting components, parts or combinations thereof is not meant to imply the presence of, parts, or combinations thereof.

Prior to the detailed description of the drawings, it is intended to clarify that the division of the constituent parts in the present specification is merely divided by the main function that each constituent part is responsible for. That is, two or more constituent parts to be described below may be combined into one constituent part, or one constituent part may be divided into two or more according to more subdivided functions. In addition, each of the constituent units to be described below may additionally perform some or all of the functions of other constituent units in addition to its own main function, and some of the main functions of each constituent unit are different. It goes without saying that it may be performed exclusively by.

In addition, in performing the method or operation method, each of the processes constituting the method may occur differently from the specified order unless a specific order is clearly stated in the context. That is, each process may occur in the same order as the specified order, may be performed substantially simultaneously, or may be performed in the reverse order.

Terms used in the following description are first defined.

An image or an image refers to a digital image including certain information or objects.

An image stack is a set of multiple images. The image stack consists of a plurality of images arranged in a certain order. For example, medical images such as CT and MRI are generated while continuously capturing a certain 2D plane image. Therefore, such medical images constitute an image stack.

The image processing device is a device that constantly processes images. The image processing device corresponds to a device capable of processing and calculating certain data. For example, the image processing device may be implemented as a device such as a PC, a smart device, or a server.

The image processing apparatus can basically determine the degree of similarity for two images. The degree of similarity means the degree to which two different images are similar. The degree of similarity can be determined by various criteria or in various ways. In the following description, the method of determining the similarity of the image by the image processing apparatus is not limited. For example, the image processing apparatus may determine the degree of similarity by comparing all pixels of a frame by location. Alternatively, the image processing apparatus may determine a similarity based on the distribution of pixel values of the frame. Alternatively, the image processing apparatus may extract feature information from the frame and determine the degree of similarity to the extent that the extracted feature information matches.

Image matching or image stack matching refers to a process of determining a corresponding pair between individual images constituting each image stack targeting a plurality of image stacks (basically two image stacks). This may be referred to as image alignment or image stack alignment.

The technique described below is a technique of image matching for a plurality of image stacks. Hereinafter, it will be described that the image processing apparatus matches a plurality of image stacks. The technique described below is a technique for matching a plurality of image stacks regardless of the type of image. However, for convenience of explanation, the description will focus on medical images. In addition, the technique described below may match an image with a plurality of image stacks. However, for the convenience of explanation, the following description will be made focusing on two image stacks.

1 is an example of a system 100 for processing a medical image. The system 100 includes a medical imaging device 110, an image processing device 150, an image DB 160, and a user terminal 180. The system 100 of FIG. 1 corresponds to a service system that provides aligned images to users.

The medical imaging device 110 is a device that generates a medical image in the form of an image stack. It is assumed that the medical staff generates a plurality of medical images for the same subject using the medical imaging apparatus 110. That is, the medical imaging apparatus 110 generates at least two medical images (image stacks).

The image processing apparatus 150 receives and processes two image stacks generated by the medical imaging apparatus 110. The image processing apparatus 150 matches images of two image stacks. Meanwhile, in the case of a medical image, it is generated by scanning a certain point on the patient's body. Therefore, in the medical image, consecutive images of body parts are stacked in one direction. That is, the medical image is composed of images having a certain direction (order) in the direction of the body part. Therefore, it is assumed that the image stack is composed of images having a specific order at least on a certain basis. 1 illustrates an image processing apparatus in the form of a PC and a server as an example.

The image DB 160 stores two image stacks matched by the image processing apparatus 150 or stores information obtained by matching two image stacks. It is assumed that each of the two image stacks has a different identifier, and that individual images included in each image stack also have a separate identifier. The image DB 160 may store information matching the identifier of the image.

The user terminal 180 provides information on the matched image stack. The user terminal 180 may output the matched image. The medical staff may perform treatment or diagnosis by comparing images matched with each other through the user terminal 180.

2 is an example of matching a plurality of image stacks and image stacks. 2 shows an image stack such as a medical image as an example. Figure 2(A) shows one image stack A. 2(A) shows an image stack A composed of eight (A1 to A8) images as an example. 2(B) shows one image stack B. FIG. 2(B) shows an image stack B composed of eight (B1 to B8) images as an example. It is assumed that image stack A and image stack B are medical images of the same subject. 2 is a cross-sectional image of the abdomen. The medical staff must compare image stack A and image stack B taken at different times to determine if there are any differences after the passage of time. At this time, the medical staff subjectively determines which image of image stack A corresponds to which image of image stack B.

2(C) is a result of matching the image stack A and the image stack B. If the matched result is provided to the medical staff, the medical staff can more easily analyze the medical image. Meanwhile, each image constituting one image stack may not be matched to an image constituting another image stack.

3 is an example of matching two image stacks. 3 corresponds to the result of matching two image stacks. In FIG. 3, A and B denote different image stacks. In FIG. 3, images arranged in the same vertical column mean matching with each other. 3(A) is an example in which A1 to An are matched to B1 to Bn, respectively. B contains more images than A. 3(B) shows a case where the intervals between images are not the same. In this case, an image that is not matched in the middle as shown in FIG. 3(B) may also occur. For example, A2 has no matching image in image stack B.

The image matching technique described below is for finding how to match two image stacks in the image stack A and the image stack B to increase the average similarity of the image. The image matching technique described below does not determine the degree of similarity with respect to the number of all sortable cases, and performs image matching with lower complexity based on the dynamic programming method.

4 is an example of a flowchart of a process 200 of matching image stacks using dynamic programming. It is assumed that the image stack is composed of images in a certain order. It is assumed that image stack A = {A1, A2,...,An} is composed and image stack B = {B1, B2,...,Bm} is composed.

The image processing apparatus first calculates a similarity between images Ai (1≤i≤n) of the image stack A and each images Bj (1≤j≤m) of the image stack B (210). The image processing apparatus does not necessarily have to first calculate individual similarities for all image pairs, but may sequentially calculate similarities for necessary image pairs. The image processing apparatus may store the calculated similarity in a storage medium.

The image processing apparatus sequentially determines the maximum value of the average of similarities for all pairs of Ai and Bj using the calculated (retained) similarity (220). In this process, the maximum value of the similarity average for the image pair Ai and Bj is determined by using the maximum value of the similarity and similarity average for a specific image pair already calculated. The process of determining the maximum value of the similarity average will be described later.

The image processing apparatus determines a target pair having the largest similarity average determined in step 220 from all image pairs constituting the image stack A or the last image of the image stack B (230). Here, all image pairs include both An and a pair leading to one Bj of the image stack B and a pair formed by Bm and any one Ai of the image stack A.

The image processing apparatus may perform matching between images of the image stack by inversely tracking an index pair that determines the average maximum value of the similarity for the target pair (240).

A process of matching image pairs in an image stack based on the maximum value of the average of similarity will be described. 5 is an example of a process of matching image pairs in an image stack. 5(A) shows an image stack for image pair matching. 5(A) shows an image stack A having 5 images and an image stack B having 6 images as an example. In Fig. 5A, the solid lines connecting the images indicate matching images.

A process of matching images based on the image stack A and the image stack B of FIG. 5A will be described. The image processing apparatus may sequentially perform a matching process based on the image stack A.

The process of determining the maximum similarity average value for an image pair including A2 will be described. (i) First, the image processing apparatus calculates the similarity of each image (B1, B2, ..., B5) of A1 and image stack B. (ii) The image processing apparatus calculates the similarity of each image (B1, B2, ..., B5) of A2 and image stack B. (iii) The image processing apparatus calculates a similarity average or a maximum value of the similarity average for each image of A2 and image stack B. ① When A2 and B1 are matched, A1 is not matched. This is because each image stack is based on images that are aligned with a certain standard. Therefore, the similarity average or the maximum value of the similarity average between A2 and B1 is the same as the similarity between A2 and B1. ② When A2 and B2 match, the image processing apparatus calculates the average of "Similarity between A2 and B2" and "Similarity between A1 and B1". ③ When A2 and B3 are matched in the same process, the image processing device uses {(average of "Similarity between A2 and B3" and "Similarity between A1 and B2") and ("Similarity between A2 and B3" and "A1 and B1"). The maximum value is determined among the average)}. The determined value is the maximum value of the average similarity between A2 and B3. Thereafter, the maximum value of the average of similarity when A2 and B4 are matched, when A2 and B5 are matched, and when A2 and B6 are matched is determined in the same process. In the end, the process of determining the average maximum similarity for an image pair is determined based on the similarity for the image pair calculated in advance.

(iv) The average maximum value of the similarity for the image pair including A3 is performed in a similar process. The image processing apparatus calculates the similarity of each image (B1, B2, ..., B5) of A3 and image stack B. Similarity calculations for individual image pairs may be sequentially performed as described above, or similarity calculations for all pairs may be performed in advance. ① When A3 and B1 are matched, there is no target for matching A2 and A1. Therefore, the similarity average or the maximum value of the similarity average between A3 and B1 is the same as the similarity between A3 and B1. ② When A3 and B2 are matched, A2 can only match B1, and there is no target for A1 to match. Therefore, the maximum value of the average of similarity between A3 and B2 is the same as the average of "similarity between A3 and B2" and "similarity between A2 and B1".

③ When A3 and B3 are matched, the image processing apparatus is used as (("Similarity between A3 and B3" and "Similarity between A2 and B2" and average of "Similarity between A1 and B1") and ("Similarity between A3 and B3") And the average of the "similarity of A2 and B1")}. In this case, the image processing apparatus may not calculate the average of the similarity for all cases, and may calculate the average of the similarity by using the previously calculated result. The image processing apparatus may determine the maximum value of the average of similarity for the pair of images currently matched by using the maximum value of the average of similarity for the specific image pair previously determined (dynamic programming method). The maximum value of the similarity average can be calculated as in Equation 1 below. Equation 1 represents the maximum value F _i,j of the average of similarity for the pair of image Ai and image Bj.

F _i,j may be determined as a maximum value among the average of similarity for a set of a plurality of image pairs. At this time, the set with the maximum similarity average among the set of image pairs is called a target set. The target set consists of a number of image pairs. The number of image pairs constituting the target set is called n _i,j .

s _i,j means the similarity between image i and image j. F _{i - 1,k} means the maximum value of the average of similarity between image Ai-1 and image Bk. n _{i - 1,k} is the number of image pairs constituting the target set of F _{i - 1,k} .

At the time of calculating F _i,j , F _{i - 1,k} has already been calculated. Therefore, this point, F _i-1, n _i, and the target set of _{k -} is _{1, k} is also determined condition. When i≥2, n _{i - 1,k} is for calculating a similarity average to determine F _i,j in consideration of the number of image pairs constituting the target set of F _{i - 1,k} .

Δ(delta) may determine a range for finding a matched image in the image stack B when a matching image is found in the image stack B based on the image in the image stack A. Image stack A and image stack B may have different numbers of images. Δ corresponds to the maximum number of images in image stack B that do not match the image in image stack A. Looking at the matched result as a basis, it can be seen that the image stack A has a gap. FIG. 5(B) illustrates a result of matching image stack A and image stack B of FIG. 5(A). Referring to Fig. 5B, there is one gap between the image A1 and the image A2. That is, FIG. 5 is a case where Δ ≥ 1. In fact, various types are possible depending on the type and number of images included in image stack A and image stack B. However, for convenience of explanation, it is assumed that each image of image stack A matches any one image of image stack B.

Using Equation 1, the image processing apparatus may determine the maximum values F ₃ and ₃ of the average of the similarity when A3 and B3 are matched. Afterwards, the maximum value of the similarity average F _3,4 when A3 and B4 are matched in the same way, the maximum value of the similarity average F _3,5 when A3 and B5 are matched, and the similarity average when A3 and B6 are matched The maximum value of F _3,6 can be determined.

In addition, the image processing apparatus may perform a similar process (dynamic programming method) for the similarity average maximum value for the image pair including A4 and the similarity average maximum value for the image pair including A5 in the same manner as described above.

Upon completion of this process, the image processing apparatus obtains the average maximum value of similarity for all image pairs in the image stack.

Finally, the image processing apparatus determines the largest value (maximum value) from the average of the similarity for the pair of the last image of each image stack. Referring to Fig. 5(A), the image processing apparatus includes ((A5,B1), (A5,B2), (A5,B3), (A5,B4), (A5,B5), (A5,B6) , (A1,B6), (A2,B6), (A3,B6), (A4,B6)} to determine the maximum value of the mean of similarity for the image pairs. Among the image pairs, the pair having the maximum value of the average of the greatest similarity is called a target pair. Now, the image processing apparatus performs image matching between the image stack A and the image stack B by inverting the image pair based on the target pair. Through this process, a matching result as shown in FIG. 5(B) may be calculated.

6 is an example of a process of matching image pairs based on a similarity measure. In FIG. 5, it is assumed that a similarity measure for possible image pairs between the image stack A and the image stack B is determined for the image stack of FIG. 5A. It is assumed that the similarity measure is the maximum value of the above-described similarity average. Other values may be used for the similarity measure. This will be described later.

(i) The image processing apparatus finally determines the maximum value of the average of the similarity for the pair of the last image of each image stack. Assume that the target pair is (A5, B6). The target pair represents the matched image pair for the last image. 6 shows the target pair at the bottom. (ii) The image processing apparatus determines an image pair whose similarity measure is the maximum value in all cases in which the A4 image located immediately before A5 is paired with the image of the image stack B (S1). At this time, in the image stack B, the index j ₄ of the image in which the similarity measure to the A4 image is the maximum value is determined. _i j denotes the index of the stack image B form a pair of images A _i and the target stack A. (iii) The image processing apparatus determines an image pair whose similarity measure is the maximum value in all cases in which the A3 image located immediately before A4 is paired with the image of the image stack B (S2). At this time, in the image stack B, the index j ₃ of the image whose similarity measure to the A3 image is the maximum value is determined. (iv) The image processing apparatus determines an image pair whose similarity measure is the maximum value in all cases in which the A2 image located immediately before A3 is paired with the image of the image stack B (S3). At this time, in the image stack B, the index j ₂ of the image whose similarity measure to the A2 image is the maximum value is determined. (v) The image processing apparatus determines an image pair whose similarity measure is the maximum value in all cases in which the A1 image located immediately before A2 is paired with the image of the image stack B (S4). At this time, in the image stack B, the index j ₁ of the image in which the similarity measure to the A1 image has the maximum value is determined.

Following this process, the matching of the images of the image stack A and the image stack B is terminated. (A1, j ₁ ), (A2, j ₂ ), (A3, j ₃ ), (A4, j ₄ ), (A5, j ₅ (=B6)) are matched results. 5(B) is an example showing the result of matching the image stack of FIG. 5(A).

The above-described image stack matching process will be generalized and described. 7 is an example of a pseudo code for a process of matching image pairs in an image stack. 7 is an example of using the maximum value of the average of the similarity described above. The algorithm name is Stack Alignment. Basically, you need image stack A and image stack B. n ₁ is the length of image stack A (= number of images). n ₂ is the length of image stack B (= number of images). F is a matrix of size n ₁ × n ₂ and its initial value is 0. F stores the maximum value of the average of the similarity of the image pair. F[i,j] is the maximum value of the average of similarity for the pair of image i of image stack A and image j of image stack B. N is a matrix of size n ₁ × n ₂ , and the initial value is 0. N stores the number of candidate image pairs used to determine the maximum value of the average of similarity. N is for recording how many pairs were averaged when calculating the similarity average. N[i,j] is the number of candidate image pairs used to determine F[i,j]. S(i,j) corresponds to the calculated similarity between image i of image stack A and image j of image stack B.

Lines 5 through 17 in the pseudo code are the process of calculating the maximum value of the average of similarity for all pairs that the image of image stack A and image of image stack B can compose. As described above, F[1,j] or F[i,1], which is the maximum value of the average similarity for the first index, is equal to S(1,j) and S(i,1), respectively. In this case, N[i,j] becomes 1 (line 8 to line 10). The maximum value of the average of similarity for the pair of images having the index other than the first index is determined in line 12. Line 12 is a process of determining the maximum similarity average value for the current F[i,j] using the maximum value of the previously stored similarity average. Lines 13 and 14 are processes of accumulating and storing the number of image pairs used to determine the maximum value of the average of similarity.

Lines 18 to 25 in the pseudo code correspond to a process of matching the image of the image stack A with the image of the image stack B by using the result of calculating the average maximum value of the similarity for the image pair. Line 18 is a process of selecting an image pair having the highest similarity average from the image pair having the last index. First, A[I] and B[J] are matched. M[I] is a matrix that stores the index (J) of the image matching A[I] in the image stack B. M may have the same length as image stack A. Now, starting from the target pair selected from the image pair with the last index determined, the index of the image stack A is decreased by one, and the index of the image with the maximum similarity average when the image stack B matches A[i] is found. The process of matching B[J] determined for A[i] is repeated. Finally, when a matching image for A[1] is found, the process of matching the image of image stack A with the image of image stack B ends.

In the above description, the process of matching the image stack A and the image stack B has been described. An example of matching image stack B with image stack A among two image stacks has been described. Of course, one of the two image stacks can be used to match another image stack. The image stack as a reference can be arbitrarily selected.

The type of image stack and whether the image stack is actually matched may vary, and accordingly, a specific process (algorithm) for image stack matching may be slightly modified.

(1) Image stack A and image stack B may have different numbers of images. 3 and 5 also illustrate a case where the number of images of the image stack A and the image stack B are different from each other. In this case, it may be desirable to determine whether to match another image stack based on an image stack having a small number of images. For convenience of explanation, it is assumed that image stack A includes fewer images than image stack B.

(2) Each image of image stack A may match any one of images of image stack B. When the number of images in image stack B is larger than the number of images in image stack A, image stack B includes images that do not match image stack A. In FIGS. 3(A) and 5(B), an unmatched image of the image stack B is shown.

(3) Furthermore, the image stack A may include an image that does not match the image of the image stack B. Referring to FIG. 3(B), a case in which image A2 of image stack A does not have a matched image in image stack B is illustrated. As such, when there is an image that does not match in image stack A, the case should be considered. In this case, the water supply code shown in FIG. 7 needs to be partially modified. In a specific A[k], each image of the image B and the similarity S are calculated. If the similarity S is less than or equal to the reference value for all possible pairs, it is not necessary to calculate the average maximum value of the similarity for a specific A[k]. Alternatively, it may not be determined whether to match A[k] during the matching process. For example, after determining whether to match A[k+1] in the matching process, it is possible to determine whether to match A[k-1] without determining whether to match A[k].

In addition, an example in which the maximum value of the average of the similarity of an image pair is used as a reference has been described in FIG. 7 and the like. However, this is only an example, and a measure other than the average of similarity (similarity scale) can be used. The similarity measure may have various forms that can be calculated based on the similarity of an image pair. As for the similarity measure, a value indicating similarity information of an image pair is sufficient. For example, the maximum value of the similarity measure may be determined as in Equation 2 below. In the capital code of FIG. 7, line 12 may be replaced by Equation 2 below. In this case, since the similarity average is not calculated, there is no need to use a matrix or variable required for calculating the average value (N and lines 13 to 14).

Of course, the similarity measure or the maximum value of the similarity measure may be determined by an equation or function different from Equation 2.

8 is an example of a configuration of the image processing apparatus 300. The image processing apparatus 300 corresponds to the image processing apparatus 150 of FIG. 1.

The image processing apparatus 300 is a device that aligns or matches images of the above-described image stack. The image processing apparatus 300 may be physically implemented in various forms. For example, the image processing apparatus 300 may have a form such as a computer device such as a PC, a server of a network, and a chipset dedicated to image processing. The computer device may include a mobile device such as a smart device.

The image processing device 300 includes a storage device 310, a memory 320, an operation device 330, an interface device 340, and a communication device 350.

The storage device 310 stores a program or source code for image matching. For example, the storage device 310 may store a program for calculating a similarity between two images. The storage device 310 may store a program that calculates a maximum value of an average of similarity for an image pair of two image stacks based on a dynamic programming method. Alternatively, a program that calculates the maximum value of the similarity measure may be stored in an image pair of two image stacks based on the dynamic programming method. In addition, the storage device 310 may store two input image stacks.

The memory 320 may store data generated while the image processing apparatus 300 matches the image stack and some images of the image stack.

The interface device 340 is a device that receives certain commands and data from the outside. The interface device 340 may receive images constituting an image stack from an input device physically connected or an external storage device. The interface device 340 may receive various programs or commands for image processing.

The communication device 350 refers to a configuration that receives and transmits certain information through a wired or wireless network. The communication device 350 may receive an image of an image stack from an external object. The communication device 350 may receive programs or parameters for matching image stacks. Furthermore, the communication device 350 may transmit a result of matching two image stacks to an external object.

The communication device 350 to the interface device 340 are devices that receive certain data or commands from the outside. The communication device 350 to the interface device 340 may be referred to as an input device.

The computing device 330 performs matching on an image stack using an image and a program stored in the storage device 310. The computing device 330 may perform matching on two image stacks through the above-described process. For example, the computing device 330 may calculate a similarity between the first images of the first image stack and all pairs of images formed by the second images of the second image stack. The computing device 330 may determine a maximum value of an average of similarity or a maximum value of a similarity measure for a pair of the selected one image and each of the second images while selecting the first images in order. The computing device 330 includes a maximum value of the average of the similarity (or the maximum value of the similarity measure) for each pair of the last image of the first images and the pair of the second images, and a pair of the last image of the second images and each of the first images. A target image pair having a weighted larger value (maximum value) among the similarity average (or similarity measure) for can be determined. The computing device 330 may determine a pair of first images that match the second images based on the target image pair. The computing device 330 may be a device such as a processor, an AP, or a chip in which a program is embedded that processes data and processes certain operations.

In addition, the image stack matching method as described above may be implemented as a program (or application) including an executable algorithm that can be executed on a computer. The program may be provided by being stored in a non-transitory computer readable medium.

The non-transitory readable medium refers to a medium that stores data semi-permanently and can be read by a device, not a medium that stores data for a short moment, such as a register, cache, or memory. Specifically, the above-described various applications or programs may be provided by being stored in a non-transitory readable medium such as a CD, DVD, hard disk, Blu-ray disk, USB, memory card, and ROM.

The present embodiment and the accompanying drawings are merely illustrative of some of the technical ideas included in the above-described technology, and those skilled in the art will be able to easily within the scope of the technical ideas included in the specification and drawings It will be apparent that all of the modified examples and specific embodiments that can be inferred are included in the scope of the rights of the above-described technology.

100: System for processing medical images
110: medical imaging device
150: image processing device
160: video DB
180: user terminal
300: image processing device
310: storage device
320: memory
330: operation device
340: interface device
350: communication device

Claims (13)

Receiving, by the image processing apparatus, a first image stack and a second image stack each including images having a specific order;
Calculating, by the image processing apparatus, a degree of similarity between the images of the first image stack and all image pairs formed by the images of the second image stack;
The image processing apparatus and the images of the first image stack according to the order
Determining a maximum value of an average of similarity for image pairs formed by the images of the second image stack; And
The image processing apparatus is the average of the similarity among the image pair formed by the last image of the first image stack and the image of the second image stack, and the image pair formed by the last image of the second image stack and the image of the first image stack. Matching the first image stack and the second image stack in reverse order based on the maximum value of the average of similarity starting from the target image pair having the largest maximum value,
The determining of the maximum value of the average of the similarity may include, for a pair of the first image of the first image stack and the second image of the second image stack, of the first image and the second image. A first pair and at least one second pair consisting of all images located before the first image in the first image stack and any one image located before the second image in the second image stack based on the order An image stack matching method using a dynamic programming method in which an average of the similarity of an image pair belonging to the image pair set among the configured image pair set is determined as a maximum value. delete delete The method of claim 1,
The image processing apparatus is an image stack matching method using a dynamic programming method in which the maximum value of the average of similarity for the target image pair is determined by using the maximum value of the average of similarity for the image pair stored before determining the maximum value of the average of similarity for the target image pair. . delete The method of claim 1,
After the image processing apparatus determines the target image pair,
The first image stack and the second image stack are determined by repeating the process of determining the maximum value of the average similarity between each image located before the image constituting the target image pair in the first image stack and the image of the second image stack. Image stack matching method using matching dynamic programming method. The method of claim 6,
When the similarity between the at least one image of the first image stack and the image of the second image stack is less than or equal to a reference value, the image processing apparatus is configured to match the first image stack with the second image stack. An image stack matching method using a dynamic programming method that does not determine whether an image is matched. An input device that receives a first image stack and a second image stack each including images having a series of sequences;
Calculates the similarity between two images, and calculates the maximum value of the average of similarity between any one image constituting the first image stack and any one image pair constituting the second image stack based on dynamic programming. A storage device for storing a program; And
The program is used to calculate the similarity between the images of the first image stack and all pairs of images formed by the images of the second image stack, and select images from the first image stack according to the order. Determine a maximum value of an average of similarity for an image pair formed by any one image and each of the images of the second image stack,
The target having the largest similarity average among the image pairs formed by the last image of the first image stack and the image of the second image stack, and the image pair formed by the last image of the second image stack and the image of the first image stack Starting from an image pair, comprising a computing device for matching the first image stack and the second image stack in reverse order based on the maximum value of the average of the similarity,
For a pair of any one of the first image of the first image stack and any one of the second image of the second image stack, the computing device is configured to perform a first pair of the first image and the second image, and the second image. 1 Among a set of image pairs consisting of all images located before the first image in the image stack and at least one second pair of any one image located before the second image in the second image stack based on the order An image stack matching apparatus using a dynamic programming method for determining a maximum value of the similarity average as a value having the maximum similarity average of an image pair belonging to the image pair set. delete The method of claim 8,
The computing device is
An image stack matching device using a dynamic programming method for determining the maximum value of the average of similarity for the target image pair by using the maximum value of the average of similarity for the image pair stored before determining the maximum value of the average of similarity for the target image pair. The method of claim 8,
After determining the target image pair, the computing device repeats the process of determining the maximum value of the average of similarity between each image located before the image constituting the target image pair in the first image stack and the image of the second image stack. An image stack matching apparatus using a dynamic programming method for matching the first image stack and the second image stack. The method of claim 11,
When the similarity between the at least one image of the first image stack and the image of the second image stack is less than or equal to a reference value, the processing unit matches the first image stack with the second image stack. An image stack matching device using a dynamic programming method that does not determine whether or not to match. A computer-readable recording medium in which a program for executing the image stack matching method using the dynamic programming method according to any one of claims 1, 4 and 6 to 7 is recorded on a computer.

Images