An image registration method and apparatus

Abstract

This invention provides an image registration method and apparatus, relating to the field of image processing technology. The method includes: obtaining multiple CT images obtained by performing DRR processing on CT data based on a preset digital image reconstruction DRR parameter set, and obtaining an X-ray image; determining a first image among the CT images that has the highest similarity to the X-ray image; adjusting a target DRR parameter set; performing DRR processing on the CT data based on the adjusted target DRR parameter set to obtain a second image; calculating the similarity between the second image and the X-ray image; if the similarity does not meet a preset similarity condition, returning to the step of adjusting the target DRR parameter set; if the similarity meets the similarity condition, determining the second image as the result of CT data and X-ray image registration. The solution provided by this invention can achieve image registration of CT images and X-ray images.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to an image registration method and apparatus. Background Technology

[0002] Image registration compares the similarity of two images to determine whether they contain the same object and were acquired under different conditions. These conditions could include different image acquisition devices, different shooting times, or different shooting angles.

[0003] In medical diagnostic scenarios, it is typically necessary to acquire CT data and X-ray images of the subject. For ease of observation and comparison, the CT data needs to be converted into images; for simplicity, these converted images are referred to as CT images. The CT images are then registered with the X-ray images to find the CT image that matches the X-ray image. Therefore, a scheme for image registration of CT and X-ray images is needed. Summary of the Invention

[0004] The purpose of this invention is to provide an image registration method and apparatus to achieve image registration of CT images and X-ray images. The specific technical solution is as follows:

[0005] In a first aspect, embodiments of the present invention provide an image registration method, the method comprising:

[0006] Multiple CT images are obtained by performing DRR processing on CT data based on a preset set of digital image reconstruction DRR parameters, and X-ray images are obtained. Each CT image corresponds to a preset set of DRR parameters.

[0007] Identify the first image in the CT images that has the greatest similarity to the X-ray image;

[0008] The target DRR parameter set is adjusted, and the CT data is processed by DRR based on the adjusted target DRR parameter set to obtain a second image. The initial value of the target DRR parameter set is the DRR parameter set corresponding to the first image.

[0009] Calculate the similarity between the second image and the X-ray image;

[0010] If the similarity does not meet the preset similarity conditions, return to the step of adjusting the target DRR parameter group;

[0011] If the similarity satisfies the similarity condition, the second image is determined as the result of the registration of the CT data and the X-ray image.

[0012] In one embodiment of the present invention, determining the first image with the highest similarity to the X-ray image in the CT image includes:

[0013] Based on the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, the first DRR parameter group is calculated according to the preset first parameter selection algorithm, and the second DRR parameter group is calculated according to the preset second parameter selection algorithm. The third image is the CT image that appears first in the CT image sequence, and the fourth image is the CT image that appears last in the CT image sequence. The CT image sequence is a sequence obtained by sorting multiple CT images according to the sorting priority of each parameter in the DRR parameter group and the value of each parameter in each preset DRR parameter group.

[0014] Determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

[0015] If so, calculate the mean of each parameter in the first DRR parameter group and the second DRR parameter group to obtain the third DRR parameter group; determine the preset DRR parameter group that is closest to the third DRR parameter group, and determine the CT image corresponding to the determined DRR parameter group as the first image with the greatest similarity to the X-ray image in the CT image;

[0016] If not, then determine the first CT image corresponding to the preset DRR parameter group closest to the first DRR parameter group and the second CT image corresponding to the preset DRR parameter group closest to the second DRR parameter group; determine the first image among multiple CT images according to the first similarity and the second similarity, wherein the first similarity is: the similarity between the first CT image and the X-ray image, and the second similarity is: the similarity between the second CT image and the X-ray image.

[0017] In one embodiment of the present invention, determining the first image among multiple CT images based on a first similarity and a second similarity includes:

[0018] Determine whether the first similarity score is less than the second similarity score;

[0019] If so, the fourth image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group, and the second DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the second parameter selection algorithm.

[0020] If not, the third image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the first DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the first parameter selection algorithm.

[0021] Return to the step of determining whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

[0022] In one embodiment of the present invention,

[0023] The algorithm for selecting the first parameter is: r1 = a + 0.382(ba)

[0024] Where r1, a, and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0025] The algorithm for selecting the second parameter is: r2 = a + 0.618(ba)

[0026] Where r2, a, and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0027] In one embodiment of the present invention, adjusting the target DRR parameter set includes:

[0028] Using the target DRR parameter set as the initial parameters for the covariance evolution algorithm, multiple sets of first-sample DRR parameter sets are generated.

[0029] Multiple sets of first-sample DRR parameter groups are processed by Gaussian distribution to obtain multiple sets of second-sample DRR parameter groups;

[0030] Calculate the average value of multiple sets of second-sample DRR parameter groups to obtain the average DRR parameter group;

[0031] The target DRR parameter set is adjusted to the average DRR parameter set.

[0032] In one embodiment of the present invention,

[0033] The CT image is: an image with a preset resolution obtained by down-resolution processing of a first initial CT image; the first initial CT image is: an image obtained by performing clarity enhancement processing on a second initial CT image using wavelet transform; the second initial CT image is: an image obtained by performing DRR processing on the CT data based on a preset DRR parameter set.

[0034] The step of determining the first image in the CT image that has the greatest similarity to the X-ray image includes:

[0035] The X-ray image is enhanced with wavelet transform.

[0036] The resolution of the enhanced X-ray image is reduced to obtain a fifth image with a preset resolution.

[0037] The first image with the highest similarity to the fifth image among the CT images is determined.

[0038] In one embodiment of the present invention, the DRR parameter set includes at least one of the following parameters:

[0039] The rotation angle, translation parameters, imaging angle, and distance between the virtual X-ray source and the imaging plane of the DRR virtual device, wherein the DRR virtual device is a virtual device used for DRR processing of CT data.

[0040] Secondly, embodiments of the present invention provide an image registration apparatus, the apparatus comprising:

[0041] The image acquisition module is used to acquire multiple CT images obtained by DRR processing of CT data based on a preset set of digital image reconstruction DRR parameters, and to acquire X-ray images. Each CT image corresponds to a preset set of DRR parameters.

[0042] The first image determination module is used to determine the first image in the CT image that has the greatest similarity to the X-ray image;

[0043] The second image acquisition module is used to adjust the target DRR parameter set, and perform DRR processing on the CT data based on the adjusted target DRR parameter set to obtain the second image. The initial value of the target DRR parameter set is the DRR parameter set corresponding to the first image.

[0044] The similarity calculation module is used to calculate the similarity between the second image and the X-ray image;

[0045] The similarity judgment module is used to determine whether the similarity meets the preset similarity conditions. If not, the second image acquisition module is triggered; if yes, the registration result determination module is triggered.

[0046] The registration result determination module is used to determine the second image as the result of the registration of the CT data and the X-ray image.

[0047] In one embodiment of the present invention, the first image determination module includes:

[0048] The parameter group calculation submodule is used to calculate the first DRR parameter group according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and to calculate the second DRR parameter group according to the preset first parameter selection algorithm. The third image is the CT image that is ranked first in the CT image sequence, and the fourth image is the CT image that is ranked last in the CT image sequence. The CT image sequence is a sequence obtained by sorting multiple CT images according to the sorting priority of each parameter in the DRR parameter group and the value of each parameter in each preset DRR parameter group.

[0049] The difference judgment submodule is used to determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold. If yes, the first determination submodule is triggered; if no, the second determination submodule is triggered.

[0050] The first determining submodule is used to calculate the mean value of each parameter in the first DRR parameter group and the second DRR parameter group to obtain the third DRR parameter group; determine the preset DRR parameter group that is closest to the third DRR parameter group; and determine the CT image corresponding to the determined DRR parameter group as the first image with the greatest similarity to the X-ray image in the CT image.

[0051] The second determining submodule is used to determine a first CT image corresponding to a preset DRR parameter group that is closest to the first DRR parameter group, and a second CT image corresponding to a preset DRR parameter group that is closest to the second DRR parameter group; and to determine the first image among multiple CT images based on a first similarity and a second similarity, wherein the first similarity is the similarity between the first CT image and the X-ray image, and the second similarity is the similarity between the second CT image and the X-ray image.

[0052] In one embodiment of the present invention, the second determining submodule is specifically used for:

[0053] Determine the first CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the second CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group;

[0054] Determine whether the first similarity is less than the second similarity;

[0055] If so, the fourth image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group, and the second DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the second parameter selection algorithm.

[0056] If not, the third image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the first DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the first parameter selection algorithm.

[0057] Return to the step of determining whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

[0058] In one embodiment of the present invention,

[0059] The algorithm for selecting the first parameter is: r1 = a + 0.382(ba)

[0060] Where r1, a, and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0061] The algorithm for selecting the second parameter is: r2 = a + 0.618(ba)

[0062] Where r2, a, and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0063] In one embodiment of the present invention, the second image acquisition module is specifically used for:

[0064] Using the target DRR parameter set as the initial parameters for the covariance evolution algorithm, multiple sets of first-sample DRR parameter sets are generated.

[0065] Multiple sets of first-sample DRR parameter groups are processed by Gaussian distribution to obtain multiple sets of second-sample DRR parameter groups;

[0066] Calculate the average value of multiple sets of second-sample DRR parameter groups to obtain the average DRR parameter group;

[0067] Adjust the target DRR parameter set to the average DRR parameter set;

[0068] Based on the adjusted target DRR parameter set, the CT data is subjected to DRR processing to obtain a second image.

[0069] In one embodiment of the present invention,

[0070] The CT image is: an image with a preset resolution obtained by down-resolution processing of a first initial CT image; the first initial CT image is: an image obtained by performing clarity enhancement processing on a second initial CT image using wavelet transform; the second initial CT image is: an image obtained by performing DRR processing on the CT data based on a preset DRR parameter set.

[0071] The first image determination module is specifically used for:

[0072] The X-ray image is enhanced with wavelet transform.

[0073] The resolution of the enhanced X-ray image is reduced to obtain a fifth image with a preset resolution.

[0074] The first image with the highest similarity to the fifth image among the CT images is determined.

[0075] In one embodiment of the present invention, the DRR parameter set includes at least one of the following parameters:

[0076] The rotation angle, translation parameters, imaging angle, and distance between the virtual X-ray source and the imaging plane of the DRR virtual device, wherein the DRR virtual device is a virtual device used for DRR processing of CT data.

[0077] Thirdly, embodiments of the present invention provide an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0078] Memory, used to store computer programs;

[0079] When a processor executes a program stored in memory, it implements the image registration method steps described in any of the first aspects above.

[0080] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the image registration method steps described in any of the first aspects above.

[0081] Beneficial effects of the embodiments of the present invention:

[0082] In the image registration scheme provided by this invention, based on a preset digital image reconstruction DRR parameter set, multiple CT images obtained by DRR processing of CT data are pre-obtained, and an X-ray image is also obtained, with each CT image corresponding to a preset DRR parameter set. A first image with the highest similarity to the X-ray image is determined from the multiple CT images. The DRR parameter set corresponding to the first image is used as the initial value of the target DRR parameter set, and the target DRR parameter set is adjusted. Based on the adjusted target DRR parameter set, DRR processing is performed on the CT data to obtain a second image. The similarity between the second image and the X-ray image is calculated. If the similarity does not meet a preset similarity condition, the process returns to adjusting the target DRR parameter set; if the similarity meets the similarity condition, the second image is determined as the result of CT data and X-ray image registration. Therefore, by applying the scheme provided by this invention, image registration of CT images and X-ray images can be achieved. Attached Figure Description

[0083] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0084] Figure 1a A flowchart illustrating the first image registration method is provided for embodiments of the present invention;

[0085] Figure 1b A schematic diagram comparing an X-ray image and a second image provided in an embodiment of the present invention;

[0086] Figure 2 A flowchart illustrating a second image registration method is provided for embodiments of the present invention;

[0087] Figure 3 A flowchart illustrating a third image registration method is provided for embodiments of the present invention.

[0088] Figure 4 A flowchart illustrating the fourth image registration method provided for embodiments of the present invention;

[0089] Figure 5 A flowchart illustrating the fifth image registration method provided in this embodiment of the invention;

[0090] Figure 6 A schematic diagram of the structure of a first image registration device is provided for embodiments of the present invention;

[0091] Figure 7 A schematic diagram of the structure of a second image registration device is provided for embodiments of the present invention;

[0092] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0093] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art based on the present invention are within the scope of protection of the present invention.

[0094] See Figure 1a The flowchart of the first image registration method is provided. The method includes the following steps S101-S106.

[0095] Step S101: Obtain multiple CT images obtained by pre-processing CT data with DRR based on a preset DRR parameter set, and obtain X-ray images. Each CT image corresponds to a preset DRR parameter set.

[0096] The aforementioned DRR refers to Digital Reconstructed Radiograph.

[0097] The CT data and X-ray images mentioned above are for the same object. This object can be a human, animal, etc.

[0098] The CT data of the object contains the object's three-dimensional vital signs information. The CT data is processed by DRR to obtain CT images. This process makes the object's three-dimensional vital signs information appear as a two-dimensional image. Therefore, the CT images contain vital signs information obtained from observing the object from a specific angle, which is determined by a preset DRR parameter set.

[0099] Because DRR parameter sets need to be set during the process of obtaining CT images from CT data, different CT images can be obtained by setting different DRR parameter sets. Therefore, DRR processing of CT data based on a preset DRR parameter set can produce one CT image, while DRR processing of CT data based on multiple preset DRR parameter sets can produce multiple CT images, with each CT image corresponding to a preset DRR parameter set.

[0100] In this step, each CT image is a pre-generated image before the start of image registration.

[0101] In one embodiment of the present invention, the DRR parameter set includes at least one of the following parameters:

[0102] The rotation angle, translation parameters, imaging angle, and distance between the virtual X-ray source and the imaging plane of the DRR virtual device are defined as follows: The DRR virtual device is a virtual device used to perform DRR processing on CT data.

[0103] Step S102: Determine the first image in the CT images that has the highest similarity to the X-ray images.

[0104] By calculating the similarity between CT images and X-ray images, a score can be obtained to characterize the degree of similarity between the two images. The higher the score, the greater the similarity between the CT image and the X-ray image. The CT image with the highest score is the first image with the greatest similarity to the X-ray image.

[0105] There are various methods for similarity calculation, such as histogram matching and matrix factorization. A similarity function can also be constructed to evaluate the similarity between two images, and then the similarity between CT images and X-ray images can be calculated using the similarity function.

[0106] Step S103: Adjust the target DRR parameter set. Based on the adjusted target DRR parameter set, perform DRR processing on the CT data to obtain the second image. The initial value of the target DRR parameter set is the DRR parameter set corresponding to the first image.

[0107] The above-mentioned adjustment of the target DRR parameter group can be based on preset adjustment rules. For example, if the target DRR parameter group contains angle-related parameters, the adjustment of the target DRR parameter group can be to lower the parameter by 1 degree.

[0108] The aforementioned adjustment of the target DRR parameter set can also be based on a preset adjustment algorithm, such as using an evolutionary algorithm to adjust each parameter in the target DRR parameter set. This evolutionary algorithm could be a covariance evolutionary algorithm.

[0109] The initial value of the target DRR parameter set is the DRR parameter set corresponding to the first image. That is, when the target DRR parameter set is adjusted for the first time, it is adjusted based on the DRR parameter set corresponding to the first image to obtain the adjusted target DRR parameter set; when the target DRR parameter set is adjusted again, it is adjusted based on the previously obtained adjusted target DRR parameter set.

[0110] Step S104: Calculate the similarity between the second image and the X-ray image.

[0111] As can be seen from step S102, the similarity between the second image and the X-ray image is calculated, that is, the score representing the degree of similarity between the second image and the X-ray image is calculated. This score can be obtained by histogram matching or matrix factorization and other methods.

[0112] Step S105: Determine whether the similarity meets the preset similarity conditions. If not, return to step S103 to adjust the target DRR parameter group. If yes, proceed to step S106.

[0113] In one implementation, since the aforementioned similarity is converted into a score to characterize the degree of similarity, the aforementioned preset similarity condition can be that the similarity is greater than or equal to a score threshold.

[0114] If the similarity score between the second image and the X-ray image is less than the set score threshold, it means that the similarity does not meet the preset similarity condition; if the similarity score between the second image and the X-ray image is greater than or equal to the set score threshold, it means that the similarity meets the preset similarity condition.

[0115] For example, if the score threshold is set to 0.6, and the calculated score is less than 0.6, it means that the similarity does not meet the preset similarity condition. In this case, return to step S103 and readjust the target DRR parameter group. If the calculated score is greater than or equal to 0.6, it means that the similarity meets the preset similarity condition. In this case, execute step S106.

[0116] In another implementation, since step S103 adjusts the target DRR parameter set based on the previously adjusted target DRR parameter set, the similarity between the second image and the X-ray image corresponding to the target DRR parameter set will initially increase as the number of adjustments increases. This indicates that the second image and the X-ray image are becoming increasingly similar. When the target DRR parameter set is adjusted to a suitable parameter set, the similarity between the second image and the X-ray image will reach its maximum value. Further adjustments to the target DRR parameter set at this point will decrease the similarity between the second image and the X-ray image, indicating that the second image is becoming increasingly dissimilar to the X-ray image. Therefore, the preset similarity condition can be that the similarity between the second image and the X-ray image after the previous parameter set adjustment is greater than the similarity between the current second image and the X-ray image, and also greater than the similarity between the second image and the X-ray image after the previous parameter set adjustment.

[0117] For example, if the similarity between the second image corresponding to the target DRR parameter group and the X-ray image is 0.5 after the previous parameter group adjustment, the similarity between the second image corresponding to the target DRR parameter group and the X-ray image is 0.6 after the previous parameter group adjustment, and the similarity between the current second image and the X-ray image is 0.4, then it means that the similarity meets the preset similarity condition.

[0118] Step S106: Determine the second image as the result of CT data and X-ray image registration.

[0119] When the similarity meets the preset similarity conditions, it means that the second image has met the requirements for image registration. Therefore, the second image can be identified as the result of the registration of CT data and X-ray images.

[0120] It should be noted that if the preset similarity condition is the same as the similarity condition mentioned in the second implementation method in step S105, then in this step, the second image corresponding to the target DRR parameter group after the last parameter group adjustment is determined as the result of CT data and X-ray image registration.

[0121] See Figure 1b The diagram shows a comparison between an X-ray image and a second image. Figure 1b In the image, a1, a2, a3, and a4 are the four acquired X-ray images, and b1, b2, b3, and b4 are the four second images obtained by image registration based on a1, a2, a3, and a4.

[0122] Therefore, the solution provided in this embodiment of the invention involves obtaining multiple CT images and X-ray images in advance based on a preset digital image reconstruction DRR parameter set, and each CT image corresponds to a preset DRR parameter set. A first image with the highest similarity to the X-ray image is determined from the multiple CT images. The DRR parameter set corresponding to the first image is used as the initial value of the target DRR parameter set, and the target DRR parameter set is adjusted. Based on the adjusted target DRR parameter set, the CT data is subjected to DRR processing to obtain a second image. The similarity between the second image and the X-ray image is calculated. If the similarity does not meet a preset similarity condition, the process returns to adjusting the target DRR parameter set. If the similarity meets the similarity condition, the second image is determined as the result of CT data and X-ray image registration. Therefore, by applying the solution provided in this embodiment of the invention, image registration of CT images and X-ray images can be achieved.

[0123] In the solution provided by the embodiments of the present invention, since the multiple CT images are obtained by pre-processing the CT data based on a preset DRR parameter set, that is, multiple CT images have already been obtained before image registration, this allows the pre-generated CT images to be directly used for subsequent processing during image registration, without the need to generate multiple CT images during the image registration process. Therefore, the solution provided by the embodiments of the present invention is fast and efficient for image registration.

[0124] In one embodiment of the present invention, see Figure 2 The flowchart of the second image registration method is provided. Compared with the embodiment shown in Figure 1 above, in this embodiment, the above step S102 determines the first image with the greatest similarity between the CT image and the X-ray image, including the following steps S102A-S102D.

[0125] Step S102A: Based on the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, calculate the first DRR parameter group according to the preset first parameter selection algorithm, and calculate the second DRR parameter group according to the preset second parameter selection algorithm.

[0126] The third image is the CT image that appears first in the CT image sequence, and the fourth image is the CT image that appears last in the CT image sequence. The CT image sequence is a sequence obtained by sorting multiple CT images according to the sorting priority of each parameter in the DRR parameter group and the preset values ​​of each parameter in each DRR parameter group.

[0127] The following example illustrates the process of sorting multiple CT images.

[0128] For example, a DRR parameter set contains two parameters, m and n, where m takes the value of m1 or m2 and n takes the value of n1 or n2, and m1 is less than m2 and n1 is less than n2. Then, the preset DRR parameter sets are DRR parameter set 1 containing m1 and n1, DRR parameter set 2 containing m1 and n2, DRR parameter set 3 containing m2 and n1, and DRR parameter set 4 containing m2 and n2. Based on these four DRR parameter sets, DRR processing is performed on the CT data to obtain four CT images, which are called image 1, image 2, image 3, and image 4.

[0129] If the sorting priority of m is specified to be higher than that of n, meaning the value of m is considered first during sorting, and the sorting is done in ascending order, then because m1 is less than m2, images 1 and 2 will be ranked higher than images 3 and 4. However, if the parameters m are equal, then the value of n, which has a lower priority, will be considered, and the sorting will be done in ascending order. In this case, image 1 will be ranked higher than image 2, and image 3 will be ranked higher than image 4. Therefore, the CT image sequence is: Image 1, Image 2, Image 3, Image 4. In this sequence, the third image is image 1, and the fourth image is image 4.

[0130] Similarly, if the sorting priority of n is specified to be higher than that of m, and the images are sorted in ascending order of their numerical values, the resulting CT image sequence will be image 1, image 3, image 2, and image 4. In this case, the third image is image 1, and the fourth image is image 4.

[0131] In one scenario, the above first parameter selection algorithm can be expressed as: r1 = a + 0.382(ba)

[0132] Where r1, a, and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0133] For example, when the DRR parameters include the rotation angle and translation parameters of the DRR virtual device, r1, a, and b can represent the rotation angles in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0134] r1, a, and b can also represent the translation parameters in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0135] The algorithm for selecting the second parameter described above can be expressed as: r² = a + 0.618(ba)

[0136] Where r2, a, and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0137] In another scenario, the first parameter selection algorithm described above can be expressed as: r1 = 0.75a + 0.25b

[0138] Where r1, a, and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0139] The algorithm for selecting the second parameter described above can be expressed as: r² = 0.25a + 0.75b

[0140] Where r2, a, and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0141] Step S102B: Determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold. If yes, proceed to step S102C; if no, proceed to step S102D.

[0142] The differences between the first DRR parameter group and the second DRR parameter group can be given in a variety of implementations.

[0143] In one implementation, the difference between the first DRR parameter group and the second DRR parameter group may include the difference between each parameter in the first DRR parameter group and the second DRR parameter group. When the difference between each parameter is less than the corresponding preset threshold, it is determined that the difference between the first DRR parameter group and the second DRR parameter group is less than the preset threshold; when the difference between any parameter in the first DRR parameter group and the second DRR parameter group is greater than the corresponding preset threshold, it is determined that the difference between the first DRR parameter group and the second DRR parameter group is greater than the preset threshold.

[0144] The differences between all the above parameters are less than their corresponding preset thresholds. In other words, each parameter in the DRR parameter group corresponds to a preset threshold. For example, if the DRR parameter group includes the rotation angle and translation parameters of the DRR virtual device, then the rotation angle corresponds to a rotation angle threshold, and the translation parameter corresponds to a translation parameter threshold. In this case, the difference between the first and second DRR parameter groups can only be determined to be less than the preset threshold if the difference between the rotation angles in the first and second DRR parameter groups is less than the rotation angle threshold, and the difference between the translation parameters is less than the translation parameter threshold.

[0145] In another implementation, the difference between the first DRR parameter group and the second DRR parameter group can be the difference between the mean values ​​determined by each parameter in the first DRR parameter group and the second DRR parameter group. When the difference is less than a preset threshold, it is determined that the difference between the first DRR parameter group and the second DRR parameter group is less than the preset threshold; when the difference is greater than the preset threshold, it is determined that the difference between the first DRR parameter group and the second DRR parameter group is greater than the preset threshold.

[0146] Step S102C: Calculate the mean value of each parameter in the first DRR parameter group and the second DRR parameter group to obtain the third DRR parameter group; determine the preset DRR parameter group that is closest to the third DRR parameter group, and determine the CT image corresponding to the determined DRR parameter group as the first image with the greatest similarity to the X-ray image in the CT image.

[0147] For example, in the example of step S102A above, assuming that the value of m also includes m3 and m4, the value of n also includes n3 and n4, and the first DRR parameter group contains the parameters m3 and n3, the second DRR parameter group contains the parameters m4 and n4, then the parameters contained in the third DRR parameter group are... and At this time, if m1 in parameter m is the same as... The closest, in parameter n, n2 and If the closest is determined, then the preset DRR parameter group that is closest to the third DRR parameter group is parameter group 2 containing m1 and n2. The CT image corresponding to the determined DRR parameter group is image 2. Therefore, image 2 is determined as the first image.

[0148] Step S102D: Determine the first CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the second CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group; determine the first image among multiple CT images based on the first similarity and the second similarity.

[0149] The first similarity is the similarity between the first CT image and the X-ray image, and the second similarity is the similarity between the second CT image and the X-ray image.

[0150] In a CT image sequence, CT images are ordered according to their corresponding DRR parameter sets following a certain pattern. The similarity between CT and X-ray images in the sequence also follows a specific pattern. Typically, the first and last CT images in the sequence are not the images with the highest similarity to the X-ray images; rather, the CT images at a certain point in the sequence have the highest similarity. Therefore, as CT images are ranked further down the sequence, the similarity between them initially increases and then decreases. Thus, based on the relationship between the first and second similarity scores, the range of the first image within the CT sequence can be continuously narrowed down, ultimately determining the first image.

[0151] Therefore, in the solution provided by the embodiments of the present invention, a CT image sequence is obtained by sorting multiple CT images, and then a first DRR parameter group and a second DRR parameter group are calculated according to the first parameter selection algorithm and the second parameter selection algorithm. The first image is determined by using the difference between the first DRR parameter group and the second DRR parameter group. Applying the solution provided by the embodiments of the present invention, it is not necessary to calculate the similarity between all CT images and X-ray images; only a portion of the CT images needs to be selected from multiple CT images to determine the first image, thereby improving the efficiency of image registration.

[0152] In one embodiment of the present invention, see Figure 3A flowchart illustrating the third image registration method is provided, which is consistent with the above. Figure 2 Compared to the illustrated embodiment, in this embodiment, step S102D determines the first CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group and the second CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group; the first image is determined among multiple CT images based on the first similarity and the second similarity, including the following steps S102D1-S102D4.

[0153] Step S102D1: Determine the first CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the second CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group.

[0154] The aforementioned preset DRR parameter set that is closest to the first DRR parameter set can be understood as the one where the difference between the same parameter in the first DRR parameter set and the preset DRR parameter set is the smallest. For example, the preset DRR parameter set that is closest to the first DRR parameter set can be determined by comparing the difference of a specific parameter in the DRR parameter set.

[0155] Alternatively, the differences between multiple parameters in the DRR parameter group can be compared, and the preset DRR parameter group that is closest to the first DRR parameter group can be determined by comprehensively considering the differences of each parameter.

[0156] For example, if the value of a parameter in the first DRR parameter group is 8, and the value of the same parameter in the preset DRR parameter group is 1, 4, 7 or 10, then the preset DRR parameter group that includes the parameter with the value 7 is the preset DRR parameter group that is closest to the first DRR parameter group.

[0157] For example, if a DRR parameter set contains three parameters, and there are two preset DRR parameter sets, referred to as the first preset DRR parameter set and the second preset DRR parameter set, the differences between the parameters in the first preset DRR parameter set and the first DRR parameter set are 0.4, 1, and 0.6, respectively, and the differences between the parameters in the second preset DRR parameter set and the first DRR parameter set are 0.5, 0.9, and 0.7, respectively. Considering all the differences, we can consider the number of parameters with the smallest differences. Since two of the parameter differences in the first preset DRR parameter set are smaller than the differences in the parameters in the second preset DRR parameter set, it can be determined that the first preset DRR parameter set is the preset DRR parameter set that is closest to the first DRR parameter set.

[0158] Since the parameters in the DRR parameter set obtained according to the parameter selection algorithm may differ from the corresponding parameter values ​​in the preset DRR parameter set, if DRR processing is performed on CT data to obtain CT images based on the DRR parameter set obtained according to the parameter selection algorithm, each DRR processing to obtain a CT image will take a certain amount of time, thus reducing the efficiency of image registration. Selecting the CT image corresponding to the preset DRR parameter set that is closest to the first DRR parameter set as the first CT image determined in this step, and selecting the CT image corresponding to the preset DRR parameter set that is closest to the second DRR parameter set as the second CT image determined in this step, only requires comparing the numerical relationships of the parameters in the DRR parameter set, and the time spent is far less than the time spent performing DRR processing to obtain CT images.

[0159] Step S102D2: Determine whether the first similarity is less than the second similarity. If yes, proceed to step S102D3; if no, proceed to step S102D4.

[0160] The first similarity is the similarity between the first CT image and the X-ray image, and the second similarity is the similarity between the second CT image and the X-ray image. If the first similarity is less than the second similarity, then step S012D3 is executed; if the first similarity is greater than or equal to the second similarity, then step S102D4 is executed.

[0161] Step S102D3: Set the fourth image to be the CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group. Set the second DRR parameter group to be the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the second parameter selection algorithm. Then return to step S102B.

[0162] As can be seen from step S102A above, the fourth image is the CT image that is ranked last in the CT image sequence.

[0163] In this step, the fourth image is set as the CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group. That is, the second CT image is taken as the last CT image in the CT image sequence, thereby reducing the range of the CT image sequence. The second DRR parameter group is recalculated based on the reduced range of the CT image sequence, and then the process returns to step S102B to determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold. If the difference between the first DRR parameter group and the second DRR parameter group is still greater than the preset threshold, and the first similarity is still less than the second similarity, the range of the CT image sequence is reduced again, and the second DRR parameter group is recalculated again, until the difference between the first DRR parameter group and the second DRR parameter group is less than the preset threshold.

[0164] Step S102D4: Set the third image to be the CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group. Set the first DRR parameter group to be the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the first parameter selection algorithm. Then return to step S102B.

[0165] Similarly to step S102D3 above, the third image is the CT image that appears first in the CT image sequence.

[0166] In this step, the first CT image is taken as the first CT image in the CT image sequence, thereby reducing the range of the CT image sequence. The first DRR parameter group is recalculated based on the reduced range of the CT image sequence. Then, the process returns to step S102B to determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold. If the difference between the first DRR parameter group and the second DRR parameter group is still greater than the preset threshold, and the first similarity is still greater than the second similarity, the range of the CT image sequence is reduced again, and the first DRR parameter group is recalculated again, until the difference between the first DRR parameter group and the second DRR parameter group is less than the preset threshold.

[0167] Therefore, in the solution provided by the embodiments of the present invention, by judging the relationship between the first similarity and the second similarity, the third image is set as the CT image corresponding to the preset DRR parameter group closest to the first DRR parameter group, or the fourth image is set as the CT image corresponding to the preset DRR parameter group closest to the second DRR parameter group. This narrows down the range of the CT image sequence. Based on the DRR parameter groups corresponding to the third and fourth images after the settings, the first or second DRR parameter group is recalculated, and the step of judging whether the difference between the first and second DRR parameter groups is less than a preset threshold is returned until the difference between the first and second DRR parameter groups is less than the preset threshold. Therefore, applying the solution provided by the embodiments of the present invention can quickly determine the first and second DRR parameter groups with differences less than the preset threshold, improving the efficiency of image registration.

[0168] In one embodiment of the present invention, see Figure 4 The document provides a flowchart of a fourth image registration method. Compared with the embodiment shown in Figure 1 above, in this embodiment, step S103 adjusts the target DRR parameter group, and DRR processing is performed on the CT data based on the adjusted target DRR parameter group to obtain a second image, including the following steps S103A-S103E.

[0169] Step S103A: Using the target DRR parameter set as the initial parameters for the covariance evolution algorithm, generate multiple sets of first sampled DRR parameter sets.

[0170] Using the covariance evolution algorithm, multiple DRR parameter sets can be randomly generated within the parameter range specified by the algorithm, based on the covariance and variance information of the previously generated first sample DRR parameter sets, with the target DRR parameter set as the benchmark. Each randomly generated DRR parameter set is a first sample DRR parameter set.

[0171] Step S103B: Perform Gaussian distribution processing on multiple sets of first-sample DRR parameter groups to obtain multiple sets of second-sample DRR parameter groups.

[0172] Since the multiple sets of first sampled DRR parameter sets generated in step S103A above may not follow a Gaussian distribution, performing Gaussian distribution processing on the multiple sets of first sampled DRR parameter sets can make the processed DRR parameter sets follow a Gaussian distribution.

[0173] Step S103C: Calculate the average value of multiple sets of second-sample DRR parameter groups to obtain the average DRR parameter group.

[0174] For example, if the DRR parameter set includes parameters x and y, step S103B above yields three second-sample DRR parameter sets: sample parameter set 1 containing x1 and y1, sample parameter set 2 containing x2 and y2, and sample parameter set 3 containing x3 and y3. Then, the average number of parameters included in the DRR parameter set is... and

[0175] Step S103D: Adjust the target DRR parameter set to the average DRR parameter set.

[0176] After calculating the average DRR parameter set in step S103C above, the values ​​of each parameter in the target DRR parameter set are adjusted to the values ​​of the corresponding parameters in the average DRR parameter set.

[0177] Step S103E: Based on the adjusted target DRR parameter set, perform DRR processing on the CT data to obtain the second image.

[0178] As can be seen from step S101 above, in the process of obtaining a CT image by performing DRR processing on CT data, it is necessary to set the DRR parameter group. In this step, the DRR parameter group is set to the adjusted target DRR parameter group, and the CT data is processed by DRR. The resulting CT image is the second image.

[0179] Furthermore, when performing image registration using the scheme provided in this embodiment of the invention, step S105 can be adjusted to determine whether the variance corresponding to the second image is less than a preset variance. If not, return to step S103 to adjust the target DRR parameter group; if yes, execute step S106.

[0180] Therefore, in the solution provided by this embodiment of the invention, multiple sets of first-sample DRR parameter sets are generated using the covariance evolution algorithm. After Gaussian distribution processing of these multiple sets of first-sample DRR parameter sets, the mean of the first-sample DRR parameter sets is taken to obtain the average DRR parameter set. Then, the target DRR parameter set is adjusted, and the second image is obtained based on the adjusted target DRR parameter set. Since the covariance evolution algorithm is a global optimization algorithm with high learning capacity and robustness, it can accurately obtain the second image, ensuring that the similarity between the second image and the X-ray image meets the preset similarity condition.

[0181] In one embodiment of the present invention, the above-mentioned CT image is: an image with a preset resolution obtained by reducing the resolution of a first initial CT image; the first initial CT image is: an image obtained by performing a sharpness enhancement process on a second initial CT image using wavelet transform; and the second initial CT image is: an image obtained by performing DRR processing on CT data based on a preset DRR parameter set.

[0182] In this embodiment, see Figure 5 The flowchart of the fifth image registration method is provided. Compared with the embodiment shown in Figure 1 above, in this embodiment, the above step S102 determines the first image with the greatest similarity between the CT image and the X-ray image, including the following steps S102E-S102G.

[0183] Step S102E: Use wavelet transform to enhance the sharpness of the X-ray image.

[0184] The aforementioned sharpness enhancement process refers to eliminating irrelevant information in X-ray images and making blurry areas of the image content in the X-ray image clear.

[0185] The irrelevant information in the above X-ray images can be specks or noise in the X-ray images.

[0186] The wavelet transform method described above can use low-pass and high-pass filters to filter X-ray images, thereby removing irrelevant information from the X-ray images.

[0187] Step S102F: Reduce the resolution of the enhanced X-ray image to obtain a fifth image with a preset resolution.

[0188] The aforementioned resolution reduction process refers to lowering the resolution of X-ray images. This can be achieved, for example, by downsampling the X-ray image; it can also be achieved by pooling the X-ray image; or it can be achieved through other methods.

[0189] Step S102G: Determine the first image in the CT images that has the highest similarity to the fifth image.

[0190] As described in step S102 above, a score representing the degree of similarity between the CT image and the fifth image can be obtained by calculating the similarity between the CT image and the fifth image. The higher the score, the greater the similarity between the CT image and the fifth image. The CT image corresponding to the highest score is the first image with the greatest similarity to the fifth image.

[0191] Therefore, in the solution provided by the embodiments of the present invention, the CT image is an image obtained by performing DRR processing on CT data based on a preset DRR parameter set, and then undergoing sharpness enhancement processing and low-resolution processing. The fifth image is an image obtained by performing sharpness enhancement processing and low-resolution processing on an X-ray image. Since the processed CT image and the fifth image have higher sharpness and lower resolution than the image obtained by performing DRR processing on CT data based on a preset DRR parameter set and the X-ray image, the time spent calculating the similarity between the CT image and the fifth image is less than the time spent calculating the similarity between the image obtained by performing DRR processing on CT data based on a preset DRR parameter set and the X-ray image, thereby improving the efficiency of image registration.

[0192] Corresponding to the image registration method described above, this embodiment of the invention also provides an image registration device.

[0193] See Figure 6 A schematic diagram of the structure of a first image registration device is provided, the device comprising:

[0194] The image acquisition module 601 is used to acquire multiple CT images obtained by DRR processing of CT data based on a preset set of digital image reconstruction DRR parameters, and to acquire X-ray images. Each CT image corresponds to a preset set of DRR parameters.

[0195] The first image determination module 602 is used to determine the first image in the CT image that has the greatest similarity to the X-ray image;

[0196] The second image acquisition module 603 is used to adjust the target DRR parameter group, and perform DRR processing on the CT data based on the adjusted target DRR parameter group to obtain the second image. The initial value of the target DRR parameter group is the DRR parameter group corresponding to the first image.

[0197] Similarity calculation module 604 is used to calculate the similarity between the second image and the X-ray image;

[0198] The similarity judgment module 605 is used to determine whether the similarity meets the preset similarity conditions. If it does not, the second image acquisition module 603 is triggered. If it does, the registration result determination module 606 is triggered.

[0199] The registration result determination module 606 is used to determine the second image as the result of the registration of the CT data and the X-ray image.

[0200] Therefore, the solution provided in this embodiment of the invention involves obtaining multiple CT images and X-ray images in advance by performing DRR processing on CT data based on a preset digital image reconstruction DRR parameter set, with each CT image corresponding to a preset DRR parameter set. A first image with the highest similarity to the X-ray image is determined from the multiple CT images. The DRR parameter set corresponding to the first image is used as the initial value of the target DRR parameter set, and the target DRR parameter set is adjusted. Based on the adjusted target DRR parameter set, DRR processing is performed on the CT data to obtain a second image. The similarity between the second image and the X-ray image is calculated. If the similarity does not meet a preset similarity condition, the process returns to adjusting the target DRR parameter set; if the similarity meets the similarity condition, the second image is determined as the result of CT data and X-ray image registration. Therefore, by applying the solution provided in this embodiment of the invention, image registration of CT images and X-ray images can be achieved.

[0201] In one embodiment of the present invention, see Figure 7 A schematic diagram of the structure of a second image registration device is provided, which is similar to the one described above. Figure 6 Compared to the illustrated embodiment, in this embodiment, the first image determination module 602 includes:

[0202] The parameter group calculation submodule 602A is used to calculate the first DRR parameter group according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and to calculate the second DRR parameter group according to the preset first parameter selection algorithm. The third image is the CT image that is ranked first in the CT image sequence, and the fourth image is the CT image that is ranked last in the CT image sequence. The CT image sequence is a sequence obtained by sorting multiple CT images according to the sorting priority of each parameter in the DRR parameter group and the value of each parameter in each preset DRR parameter group.

[0203] The difference judgment submodule 602B is used to determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold. If yes, the first determination submodule 602C is triggered; if no, the second determination submodule 602D is triggered.

[0204] The first determining submodule 602C is used to calculate the mean value of each parameter in the first DRR parameter group and the second DRR parameter group to obtain the third DRR parameter group; determine the preset DRR parameter group that is closest to the third DRR parameter group; and determine the CT image corresponding to the determined DRR parameter group as the first image with the greatest similarity to the X-ray image in the CT image.

[0205] The second determining submodule 602D is used to determine a first CT image corresponding to a preset DRR parameter group that is closest to the first DRR parameter group, and a second CT image corresponding to a preset DRR parameter group that is closest to the second DRR parameter group; and to determine the first image among multiple CT images based on a first similarity and a second similarity, wherein the first similarity is the similarity between the first CT image and the X-ray image, and the second similarity is the similarity between the second CT image and the X-ray image.

[0206] Therefore, in the solution provided by the embodiments of the present invention, a CT image sequence is obtained by sorting multiple CT images, and then a first DRR parameter group and a second DRR parameter group are calculated according to the first parameter selection algorithm and the second parameter selection algorithm. The first image is determined by using the difference between the first DRR parameter group and the second DRR parameter group. Applying the solution provided by the embodiments of the present invention, it is not necessary to calculate the similarity between all CT images and X-ray images; only a portion of the CT images needs to be selected from multiple CT images to determine the first image, thereby improving the efficiency of image registration.

[0207] In one embodiment of the present invention, the second determining submodule 602D is specifically used for:

[0208] Determine the first CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the second CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group;

[0209] Determine whether the first similarity is less than the second similarity;

[0210] If so, the fourth image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the second DRR parameter group, and the second DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the second parameter selection algorithm.

[0211] If not, the third image is set to: the CT image corresponding to the preset DRR parameter group that is closest to the first DRR parameter group, and the first DRR parameter group is set to: the DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, and according to the first parameter selection algorithm.

[0212] Return to the step of the difference judgment submodule 602B to determine whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

[0213] Therefore, in the solution provided by the embodiments of the present invention, by judging the relationship between the first similarity and the second similarity, the third image is set as the CT image corresponding to the preset DRR parameter group closest to the first DRR parameter group, or the fourth image is set as the CT image corresponding to the preset DRR parameter group closest to the second DRR parameter group. This narrows down the range of the CT image sequence. Based on the DRR parameter groups corresponding to the third and fourth images after the settings, the first or second DRR parameter group is recalculated, and the step of judging whether the difference between the first and second DRR parameter groups is less than a preset threshold is returned until the difference between the first and second DRR parameter groups is less than the preset threshold. Therefore, applying the solution provided by the embodiments of the present invention can quickly determine the first and second DRR parameter groups with differences less than the preset threshold, improving the efficiency of image registration.

[0214] In one embodiment of the present invention,

[0215] The algorithm for selecting the first parameter is: r1 = a + 0.382(ba)

[0216] Where r1, a, and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0217] The algorithm for selecting the second parameter is: r2 = a + 0.618(ba)

[0218] Where r2, a, and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image, and the DRR parameter group corresponding to the fourth image, respectively.

[0219] In one embodiment of the present invention, the second image acquisition module 603 is specifically used for:

[0220] Using the target DRR parameter set as the initial parameters for the covariance evolution algorithm, multiple sets of first-sample DRR parameter sets are generated.

[0221] Multiple sets of first-sample DRR parameter groups are processed by Gaussian distribution to obtain multiple sets of second-sample DRR parameter groups;

[0222] Calculate the average value of multiple sets of second-sample DRR parameter groups to obtain the average DRR parameter group;

[0223] Adjust the target DRR parameter set to the average DRR parameter set;

[0224] Based on the adjusted target DRR parameter set, the CT data is subjected to DRR processing to obtain a second image.

[0225] Therefore, in the solution provided by this embodiment of the invention, multiple sets of first-sample DRR parameter sets are generated using the covariance evolution algorithm. After Gaussian distribution processing of these multiple sets of first-sample DRR parameter sets, the mean of the first-sample DRR parameter sets is taken to obtain the average DRR parameter set. Then, the target DRR parameter set is adjusted, and the second image is obtained based on the adjusted target DRR parameter set. Since the covariance evolution algorithm is a global optimization algorithm with high learning capacity and robustness, it can accurately obtain the second image, ensuring that the similarity between the second image and the X-ray image meets the preset similarity condition.

[0226] In one embodiment of the present invention, the CT image is: an image with a preset resolution obtained by down-resolution processing of a first initial CT image; the first initial CT image is: an image obtained by performing sharpness enhancement processing on a second initial CT image using wavelet transform; and the second initial CT image is: an image obtained by performing DRR processing on the CT data based on a preset DRR parameter set.

[0227] The first image determination module 602 is specifically used for:

[0228] The X-ray image is enhanced with wavelet transform.

[0229] The resolution of the enhanced X-ray image is reduced to obtain a fifth image with a preset resolution.

[0230] The first image with the highest similarity to the fifth image among the CT images is determined.

[0231] Therefore, in the solution provided by the embodiments of the present invention, the CT image is an image obtained by performing DRR processing on CT data based on a preset DRR parameter set, and then undergoing sharpness enhancement processing and low-resolution processing. The fifth image is an image obtained by performing sharpness enhancement processing and low-resolution processing on an X-ray image. Since the processed CT image and the fifth image have higher sharpness and lower resolution than the image obtained by performing DRR processing on CT data based on a preset DRR parameter set and the X-ray image, the time spent calculating the similarity between the CT image and the fifth image is less than the time spent calculating the similarity between the image obtained by performing DRR processing on CT data based on a preset DRR parameter set and the X-ray image, thereby improving the efficiency of image registration.

[0232] In one embodiment of the present invention, the DRR parameter set includes at least one of the following parameters:

[0233] The rotation angle, translation parameters, imaging angle, and distance between the virtual X-ray source and the imaging plane of the DRR virtual device, wherein the DRR virtual device is a virtual device used for DRR processing of CT data.

[0234] This invention also provides an electronic device, such as... Figure 8 As shown, it includes a processor 801, a communication interface 802, a memory 803, and a communication bus 804. The processor 801, communication interface 802, and memory 803 communicate with each other via the communication bus 804.

[0235] Memory 803 is used to store computer programs;

[0236] When processor 801 executes a program stored in memory 803, it performs the following steps:

[0237] Multiple CT images are obtained by performing DRR processing on CT data based on a preset set of digital image reconstruction DRR parameters, and X-ray images are obtained. Each CT image corresponds to a preset set of DRR parameters.

[0238] Identify the first image in the CT images that has the greatest similarity to the X-ray image;

[0239] The target DRR parameter set is adjusted, and the CT data is processed by DRR based on the adjusted target DRR parameter set to obtain a second image. The initial value of the target DRR parameter set is the DRR parameter set corresponding to the first image.

[0240] Calculate the similarity between the second image and the X-ray image;

[0241] If the similarity does not meet the preset similarity conditions, return to the step of adjusting the target DRR parameter group;

[0242] If the similarity satisfies the similarity condition, the second image is determined as the result of the registration of the CT data and the X-ray image.

[0243] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0244] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0245] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0246] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0247] In another embodiment of the present invention, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described image registration methods.

[0248] In another embodiment of the present invention, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to perform any of the image registration methods described above.

[0249] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0250] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0251] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, electronic devices, computer-readable storage media, and computer program products are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0252] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. An image registration method characterized by, The method comprises: obtaining a plurality of CT images obtained by performing DRR processing on CT data based on a preset digital image reconstruction DRR parameter set, and obtaining an X-ray image, each CT image corresponding to a preset DRR parameter set, the CT data and the X-ray image being data of a same object, the CT data of the object containing three-dimensional feature information of the object, the process of performing DRR processing on the CT data to obtain the CT image causing the three-dimensional feature information of the object to be presented in the form of a two-dimensional image; determining a first image with the greatest similarity between the CT image and the X-ray image; adjusting a target DRR parameter set, performing DRR processing on the CT data based on the adjusted target DRR parameter set to obtain a second image, the initial value of the target DRR parameter set being the DRR parameter set corresponding to the first image; calculating the similarity between the second image and the X-ray image; if the similarity does not satisfy a preset similarity condition, returning to the step of adjusting the target DRR parameter set; if the similarity satisfies the similarity condition, determining the second image as the registration result of the CT data and the X-ray image; the determination of the first image with the greatest similarity between the CT image and the X-ray image comprises: calculating a first DRR parameter set according to a preset first parameter selection algorithm based on the DRR parameter set corresponding to a third image and the DRR parameter set corresponding to a fourth image, and calculating a second DRR parameter set according to a preset second parameter selection algorithm, wherein the third image is the CT image with the highest ranking in a CT image sequence, the fourth image is the CT image with the lowest ranking in the CT image sequence, and the CT image sequence is obtained by ranking a plurality of CT images according to the ranking priority of each parameter in the DRR parameter set and the value of each parameter in each preset DRR parameter set; judging whether the difference between the first DRR parameter set and the second DRR parameter set is less than a preset threshold value; if yes, calculating the mean value of each parameter in the first DRR parameter set and the second DRR parameter set to obtain a third DRR parameter set, determining the preset DRR parameter set closest to the third DRR parameter set, and determining the CT image corresponding to the determined DRR parameter set as the first image with the greatest similarity between the CT image and the X-ray image; if no, determining the first CT image corresponding to the preset DRR parameter set closest to the first DRR parameter set and the second CT image corresponding to the preset DRR parameter set closest to the second DRR parameter set, and determining the first image from the plurality of CT images according to a first similarity and a second similarity, wherein the first similarity is the similarity between the first CT image and the X-ray image, and the second similarity is the similarity between the second CT image and the X-ray image.

2. The method of claim 1, wherein, the determination of the first image from the plurality of CT images according to the first similarity and the second similarity comprises: judging whether the first similarity is less than the second similarity; If yes, set the fourth image as a CT image corresponding to a preset DRR parameter group closest to the second DRR parameter group, and set the second DRR parameter group as a DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image and according to the second parameter selection algorithm; If no, set the third image as a CT image corresponding to a preset DRR parameter group closest to the first DRR parameter group, and set the first DRR parameter group as a DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image and according to the first parameter selection algorithm; Return to the step of judging whether the difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

3. The method of claim 1, wherein the first parameter selection algorithm is r1 = a + 0.382 (b-a), wherein r1, a and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image respectively. The second parameter selection algorithm is r2 = a + 0.618 (b-a), wherein r2, a and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image respectively. The adjusting of the target DRR parameter group comprises: Taking the target DRR parameter group as the initial parameter of the covariance evolution algorithm, a plurality of first sampling DRR parameter groups are generated; The plurality of first sampling DRR parameter groups are processed according to Gaussian distribution to obtain a plurality of second sampling DRR parameter groups; 4. The method of claim 1, wherein, The average value of the plurality of second sampling DRR parameter groups is calculated to obtain an average DRR parameter group; The target DRR parameter group is adjusted to the average DRR parameter group.

5. The method of any one of claims 1-4, wherein the CT image is an image of a preset resolution obtained by performing resolution reduction processing on a first initial CT image, and the first initial CT image is an image obtained by performing sharpness enhancement processing on a second initial CT image by using a wavelet transform method, and the second initial CT image is an image obtained by performing DRR processing on the CT data based on a preset DRR parameter group; The determining of the first image in the CT image with the maximum similarity to the X-ray image comprises: performing sharpness enhancement processing on the X-ray image by using a wavelet transform method; performing resolution reduction processing on the X-ray image after the sharpness enhancement to obtain a fifth image of a preset resolution; determining the first image in the CT image with the maximum similarity to the fifth image. The DRR parameter group comprises at least one of the following parameters: a rotation angle of a DRR virtual device, a translation parameter, an imaging angle, a distance between a virtual X-ray light source and an imaging surface, wherein the DRR virtual device is a virtual device used for performing DRR processing on CT data. The device comprises: ​ 6. The method according to any one of claims 1 to 4, characterized in that, ​ ​ 7. An image registration apparatus characterized by comprising: ​ The image obtaining module is configured to obtain a plurality of CT images obtained by performing DRR processing on CT data based on a preset digital image reconstruction DRR parameter set and obtain an X-ray image, each CT image corresponds to a preset DRR parameter set, the CT data and the X-ray image are data of a same object, the CT data of the object contains three-dimensional feature information of the object, and the three-dimensional feature information of the object is presented in a two-dimensional image form through the DRR processing on the CT data; The first image determining module is configured to determine a first image with the greatest similarity to the X-ray image in the CT images; The second image obtaining module is configured to adjust a target DRR parameter set, perform DRR processing on the CT data based on the adjusted target DRR parameter set, and obtain a second image, the initial value of the target DRR parameter set being a DRR parameter set corresponding to the first image; The similarity calculating module is configured to calculate the similarity between the second image and the X-ray image; The similarity judging module is configured to judge whether the similarity meets a preset similarity condition, if not, trigger the second image obtaining module, and if yes, trigger the registration result determining module; The registration result determining module is configured to determine the second image as a registration result of the CT data and the X-ray image; The first image determining module comprises: The parameter set calculating submodule is configured to calculate a first DRR parameter set according to a preset first parameter selection algorithm based on a DRR parameter set corresponding to a third image and a DRR parameter set corresponding to a fourth image, and calculate a second DRR parameter set according to a preset second parameter selection algorithm, the third image being a CT image with the highest ranking in a CT image sequence, the fourth image being a CT image with the lowest ranking in the CT image sequence, and the CT image sequence being a sequence obtained by sorting a plurality of CT images according to the ranking priority of each parameter in the DRR parameter set and the value of each parameter in each preset DRR parameter set; The difference judging submodule is configured to judge whether the difference between the first DRR parameter set and the second DRR parameter set is less than a preset threshold, if yes, trigger the first determining submodule, and if not, trigger the second determining submodule; The first determining submodule is configured to calculate the mean value of each parameter in the first DRR parameter set and the second DRR parameter set to obtain a third DRR parameter set, determine a preset DRR parameter set closest to the third DRR parameter set, and determine the CT image corresponding to the determined DRR parameter set as the first image with the greatest similarity to the X-ray image in the CT images. The second determining sub-module is configured to determine a first CT image corresponding to a preset DRR parameter group closest to the first DRR parameter group and a second CT image corresponding to a preset DRR parameter group closest to the second DRR parameter group, and determine the first image from the plurality of CT images according to a first similarity and a second similarity, wherein the first similarity is a similarity between the first CT image and the X-ray image, and the second similarity is a similarity between the second CT image and the X-ray image.

8. The apparatus of claim 7, wherein, The second determining sub-module is specifically configured to: determine a first CT image corresponding to a preset DRR parameter group closest to the first DRR parameter group and a second CT image corresponding to a preset DRR parameter group closest to the second DRR parameter group; determine whether the first similarity is less than the second similarity; if yes, set the fourth image as the CT image corresponding to the preset DRR parameter group closest to the second DRR parameter group, and set the second DRR parameter group as a DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image and according to the second parameter selection algorithm; if no, set the third image as the CT image corresponding to the preset DRR parameter group closest to the first DRR parameter group, and set the first DRR parameter group as a DRR parameter group calculated according to the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image and according to the first parameter selection algorithm; return to the step of determining whether a difference between the first DRR parameter group and the second DRR parameter group is less than a preset threshold.

9. The apparatus of claim 7, wherein the first parameter selection algorithm is r1=a+0.382(b-a), wherein r1, a and b represent the same parameter in the first DRR parameter group, the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, respectively; the second parameter selection algorithm is r2=a+0.618(b-a), wherein r2, a and b represent the same parameter in the second DRR parameter group, the DRR parameter group corresponding to the third image and the DRR parameter group corresponding to the fourth image, respectively. The second image obtaining module is specifically configured to: generate a plurality of first sampling DRR parameter groups by taking a target DRR parameter group as initial parameters of a covariance evolution algorithm; 10. The apparatus of claim 7, wherein, perform Gaussian distribution processing on the plurality of first sampling DRR parameter groups to obtain a plurality of second sampling DRR parameter groups; calculate an average value of the plurality of second sampling DRR parameter groups to obtain an average DRR parameter group; adjust the target DRR parameter group to the average DRR parameter group; perform DRR processing on the CT data based on the adjusted target DRR parameter group to obtain a second image.

11. The apparatus of any one of claims 7-10, wherein ​ ​ The CT image is an image of a preset resolution obtained by performing a resolution reduction process on a first initial CT image, the first initial CT image is an image obtained by performing a definition enhancement process on a second initial CT image in a wavelet transform manner, and the second initial CT image is an image obtained by performing a DRR process on the CT data based on a preset DRR parameter group; The first image determination module is specifically configured to: perform a definition enhancement process on the X-ray image in a wavelet transform manner; perform a resolution reduction process on the X-ray image after the definition enhancement to obtain a fifth image of a preset resolution; determine a first image with the greatest similarity to the fifth image in the CT image.

12. The device of any one of claims 7-10, wherein, The DRR parameter group includes at least one of the following parameters: a rotation angle, a translation parameter, an imaging angle, and a distance between a virtual X-ray light source and an imaging surface of a DRR virtual device, wherein the DRR virtual device is a virtual device for performing a DRR process on CT data.

13. An electronic device, comprising: The device comprises a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The memory is used to store a computer program. The processor is used to execute the program stored in the memory to implement the method steps in any one of claims 1-6.

14. A computer-readable storage medium, characterized in that, The computer program stored in the computer readable storage medium is executed by the processor to implement the method steps in any one of claims 1-6.

Images