CN115953440A

CN115953440A - Medical image registration method and device, storage medium and electronic equipment

Info

Publication number: CN115953440A
Application number: CN202310228102.8A
Authority: CN
Inventors: 周琦超; 肖远彪
Original assignee: Manteia Data Technology Co ltd In Xiamen Area Of Fujian Pilot Free Trade Zone
Current assignee: Manteia Data Technology Co ltd In Xiamen Area Of Fujian Pilot Free Trade Zone
Priority date: 2023-03-10
Filing date: 2023-03-10
Publication date: 2023-04-11
Anticipated expiration: 2043-03-10
Also published as: CN115953440B

Abstract

The application discloses a registration method and a registration device for medical images, a storage medium and electronic equipment, which relate to the technical field of image processing, and the method comprises the following steps: performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information. By the method and the device, the problem of low accuracy of image registration in the related art is solved.

Description

Medical image registration method and device, storage medium and electronic equipment

Technical Field

The present application relates to the field of image processing technologies, and in particular, to a registration method and apparatus for medical images, a storage medium, and an electronic device.

Background

The medical image registration technology is the basis of medical image processing, and plays an important role in radiation treatment planning, tumor diagnosis, surgical guidance, treatment tracking and the like. Generally, image registration can be divided into rigid body registration and deformation registration (non-rigid body registration), however, in practical applications, rigid body registration cannot satisfy most highly nonlinear and complex deformation human tissue registration. Therefore, in practical application, a deformation registration algorithm having more degrees of freedom and capable of realizing nonlinear white space deformation and obtaining nonlinear transformation parameters or displacement fields in an optimized manner is particularly important.

The current deformation registration algorithm mainly carries out global deformation on medical images, the deformation effect of the algorithm on organ regions and target regions which are more important in radiotherapy planning and surgical guidance is poor, folding conditions can occur, and the requirements on deformation of the organ regions and the target regions in practical application cannot be met.

Aiming at the problem that the accuracy of image registration is low because deformation registration during image registration depends on the global deformation of medical images in the related art, an effective solution is not provided at present.

Disclosure of Invention

The present application mainly aims to provide a registration method and apparatus for medical images, a storage medium, and an electronic device, so as to solve the problem that the accuracy of image registration is low because deformation registration during image registration depends on global deformation of medical images in the related art.

To achieve the above object, according to one aspect of the present application, there is provided a registration method of a medical image. The method comprises the following steps: acquiring medical image information, wherein the medical image information at least comprises: a reference medical image, first organ delineation information of the reference medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

Further, after acquiring the medical image information, the method further comprises: carrying out rigid registration on the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and carrying out rigid registration on second organ delineation information of the medical image to be registered according to first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, including: performing resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

Further, the rigid registration of the medical image to be registered according to the reference medical image to obtain a first medical image to be registered includes: acquiring a conversion matrix between the reference medical image and the medical image to be registered; and carrying out rigid registration on the medical image to be registered according to the transformation matrix to obtain the first medical image to be registered.

Further, the step of performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered includes: respectively performing Gaussian filtering on the reference medical image and the medical image to be registered to obtain a filtered reference medical image and a filtered medical image to be registered; and respectively performing down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain the N layers of reference medical images and the N layers of medical images to be registered.

Further, determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information, and the N layers of second target organ delineation information includes: setting an initial deformation field of a first layer for a first layer of reference medical image and a first layer of medical image to be registered, and obtaining a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, and second target organ delineation information corresponding to the first layer of medical image to be registered and the first layer of medical image to be registered; iteratively updating the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of a first layer; for the M layer of parameter medical image and the M layer of medical image to be registered, determining the deformation field of the M-1 layer as the initial deformation field of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N; and performing up-sampling processing on the deformation field of the Nth layer, and determining the processed deformation field of the Nth layer as the target deformation field.

Further, obtaining a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered, and second target organ delineation information corresponding to the first layer of medical image to be registered includes: deforming the first layer medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer medical image to be registered; calculating according to the deformed pixels in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function; calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; determining the target constraint function from the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function.

Further, calculating according to the deformed pixel points in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function includes: calculating the sum of first square differences of pixel points in the deformed first layer medical image to be registered, and calculating according to the sum of the first square differences and noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; calculating the sum of second mean square differences of pixel points in the first layer of reference medical image, and calculating according to the sum of the second mean square differences and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and calculating according to the first self-similarity function and the second self-similarity function to obtain the first constraint function.

Further, calculating according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered, and obtaining a second constraint function includes: acquiring first target organ delineation information of each organ in the first layer of reference medical image, and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; similarity calculation is carried out on the first target organ delineation information of each organ in the first layer of reference medical image and the second target organ delineation information of each organ in the deformed first layer of medical image to be registered, and a similarity constraint value corresponding to each organ is obtained; and acquiring the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value of each organ, the total volume value and the similarity constraint value corresponding to each organ to obtain the second constraint function.

Further, determining the target constraint function from the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function comprises: respectively setting weight values for the first constraint function, the second constraint function, the third constraint function and the fourth constraint function to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function; and calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

In order to achieve the above object, according to another aspect of the present application, there is provided a registration apparatus for medical images. The device includes: an obtaining unit, configured to obtain medical image information, wherein the medical image information at least includes: a reference medical image, first organ delineation information of the reference medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; the grading unit is used for carrying out resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, carrying out resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; the determining unit is used for determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and the first registration unit is used for registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

Further, after acquiring the medical image information, the apparatus further comprises: the second registration unit is used for carrying out rigid registration on the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and carrying out rigid registration on second organ delineation information of the medical image to be registered according to first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; the classification unit includes: the first grading subunit is used for carrying out resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and the second grading subunit is used for carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

Further, the second registration unit includes: an obtaining subunit, configured to obtain a conversion matrix between the reference medical image and the medical image to be registered; and the first registration subunit is used for performing rigid registration on the medical image to be registered according to the transformation matrix to obtain the first medical image to be registered.

Further, the classification unit includes: the filtering subunit is configured to perform gaussian filtering on the reference medical image and the medical image to be registered respectively to obtain a filtered reference medical image and a filtered medical image to be registered; and the down-sampling sub-unit is used for respectively performing down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain the N layers of reference medical images and the N layers of medical images to be registered.

Further, the determination unit includes: a setting subunit, configured to set an initial deformation field of a first layer for a first layer of reference medical image and a first layer of medical image to be registered, and obtain a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, and second target organ delineation information corresponding to the first layer of medical image to be registered and the first layer of medical image to be registered; the iteration subunit is used for carrying out iteration updating on the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of a first layer; the determining subunit is used for determining the deformation field of the M-1 layer as the initial deformation field of the M layer for the parameter medical image of the M layer and the medical image to be registered of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N; and the up-sampling subunit is used for performing up-sampling processing on the deformation field of the Nth layer and determining the processed deformation field of the Nth layer as the target deformation field.

Further, the setting subunit includes: the deformation module is used for deforming the first layer medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer medical image to be registered; the first calculation module is used for calculating according to the deformed pixel points in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function; the second calculation module is used for calculating according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; the first processing module is used for obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; the second processing module is used for obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; a determining module, configured to determine the target constraint function according to the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function.

Further, the first calculation module includes: the first calculation submodule is used for calculating the sum of first square deviations of pixel points in the deformed first layer medical image to be registered and calculating according to the sum of the first square deviations and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; the second calculation submodule is used for calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and the third calculation submodule is used for calculating according to the first self-similarity function and the second self-similarity function to obtain the first constraint function.

Further, the second calculation module includes: the acquisition submodule is used for acquiring first target organ delineation information of each organ in the first layer of reference medical image and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; a fourth calculation submodule, configured to perform similarity calculation on the first target organ delineation information of each organ in the first layer of reference medical image and the second target organ delineation information of each organ in the deformed first layer of medical image to be registered, so as to obtain a similarity constraint value corresponding to each organ; and the fifth calculation submodule is used for acquiring the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value of each organ, the total volume value and the similarity constraint value corresponding to each organ to obtain the second constraint function.

Further, the determining module includes: a setting submodule, configured to set weight values for the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function, respectively, to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function, and a fourth weight value corresponding to the fourth constraint function; a sixth calculating sub-module, configured to calculate according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value, and the fourth weight value to obtain the target constraint function.

In order to achieve the above object, according to another aspect of the present application, there is also provided a computer-readable storage medium storing a program, wherein when the program is executed, a device on which the storage medium is located is controlled to perform the registration method of the medical image according to any one of the above items.

To achieve the above object, according to one aspect of the present application, there is provided an electronic device comprising one or more processors and a memory for storing the one or more processors to implement the registration method of the medical image as described in any one of the above.

By the application, the following steps are adopted: acquiring medical image information, wherein the medical image information at least comprises: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information, so that the problem of low accuracy of image registration caused by the fact that deformation registration depends on global deformation of the medical image during image registration in the related technology is solved. According to the scheme, before the medical image to be registered is registered, the resolution of the image to be registered and the resolution of the reference medical image are graded, so that N layers of reference medical images and N layers of medical images to be registered are obtained. Therefore, the calculated amount of the deformation field during iteration can be effectively reduced, and then the deformation field is calculated by utilizing the organ delineation information, so that the accuracy and reasonability of organ deformation are further improved, and the effect of improving the accuracy of image registration is further achieved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application. In the drawings:

fig. 1 is a flow chart of a registration method of medical images provided according to an embodiment of the present application;

fig. 2 is a flow chart of an alternative method of registration of medical images provided according to an embodiment of the present application;

fig. 3 is a schematic diagram of a registration apparatus for medical images provided according to an embodiment of the present application;

FIG. 4 is a schematic diagram of a registration system for medical images provided in accordance with an embodiment of the present application;

fig. 5 is a schematic diagram of an electronic device provided according to an embodiment of the application.

Detailed Description

It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present application without making any creative effort shall fall within the protection scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It should be understood that the data so used may be interchanged under appropriate circumstances such that embodiments of the application described herein may be used. Moreover, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be noted that the relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for presentation, analyzed data, medical image data, etc.) referred to in the present disclosure are information and data that are authorized by the user or sufficiently authorized by each party. For example, an interface is provided between the system and the relevant user or organization, before obtaining the relevant information, an obtaining request needs to be sent to the user or organization through the interface, and after receiving the consent information fed back by the user or organization, the relevant information is obtained.

The invention is described below with reference to preferred implementation steps, and fig. 1 is a flowchart of a registration method of medical images according to an embodiment of the present application, as shown in fig. 1, the method includes the following steps:

step S101, acquiring medical image information, wherein the medical image information at least comprises: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered;

specifically, a reference medical image and a floating image (i.e., the medical image to be registered) are obtained, the delineation information of the reference medical image and the delineation information of the floating image are obtained at the same time, and the appointed corresponding delineation information or the corresponding delineation information with an adjacent target area is screened from the delineation information of the floating image and the delineation information of the reference medical image, so as to obtain the first organ delineation information and the second organ delineation information. Since the delineation information of the floating image and the reference medical image is added during the deformation field calculation in the embodiment of the application, the delineation information provided by the floating image and the reference medical image is required to be corresponding, and therefore the corresponding delineation information in the reference medical image and the medical image to be registered needs to be screened out in advance. For example, a lung delineation of the medical image to be registered is referenced to a lung delineation of the medical image.

Step S102, carrying out resolution grading on a reference medical image and a medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and carrying out resolution grading on first organ delineation information and second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2;

specifically, the resolution grading can be performed on the reference medical image and the medical image to be registered through a multi-resolution pyramid type iterative algorithm to obtain N layers of reference medical images and N layers of medical images to be registered. Because the resolution grading is carried out on the reference medical image and the floating image and the reference medical image and the floating image are divided into N layers of resolution images, the resolution grading is carried out on the delineation information of the reference medical image and the delineation information of the medical image to be registered, and the layer number and the resolution are consistent with the grading of the reference medical image and the floating image, so that the resolution grading is carried out on the first organ delineation information and the second organ delineation information, and N layers of first target organ delineation information and N layers of second target organ delineation information are obtained.

In an alternative embodiment, N may be set directly to 2. And carrying out multi-layer resolution grading on the reference medical image and the medical image to be registered by using a multi-resolution pyramid iterative method to obtain a high-to-low 2-layer resolution reference medical image and a 2-layer resolution medical image to be registered.

The resolution grading is carried out on the reference medical image and the medical image to be registered, so that the calculation amount is favorably reduced, and the calculation efficiency is improved.

Step S103, determining a target deformation field according to N layers of reference medical images, N layers of medical images to be registered, N layers of first target organ delineation information and N layers of second target organ delineation information;

specifically, after obtaining N layers of reference medical images and N layers of medical images to be registered, the target deformation field is determined by the N layers of reference medical images, the N layers of medical images to be registered, the first target organ delineation information and the second target organ delineation information.

And S104, registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

Specifically, the target deformation field obtained in step S103 is used to perform deformation registration on the medical image to be registered and the second organ delineation information, and finally the registered medical image and the registered second organ delineation information are output.

In summary, before the medical image to be registered is registered, the resolution of the image to be registered and the resolution of the reference medical image are graded, so as to obtain N layers of reference medical images and N layers of medical images to be registered. Therefore, the calculated amount of the deformation field during iteration can be effectively reduced, and then the deformation field is calculated by utilizing the organ delineation information, so that the accuracy and reasonability of organ deformation are further improved, and the effect of improving the accuracy of image registration is further achieved.

In order to improve the accuracy of the registration of the medical image, in the registration method of the medical image provided by the embodiment of the present application, after acquiring the medical image information, the registration method further includes: rigidly registering the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and rigidly registering second organ delineation information of the medical image to be registered according to the first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; performing resolution classification on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution classification on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein the steps of: performing resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

The rigid registration of the medical image to be registered according to the reference medical image to obtain a first medical image to be registered comprises the following steps: acquiring a conversion matrix between a reference medical image and a medical image to be registered; and carrying out rigid registration on the medical image to be registered according to the transformation matrix to obtain a first medical image to be registered.

Specifically, rigid registration is performed on the medical image to be registered according to the reference medical image, and a transformation matrix required by the rigid registration is obtained. Secondly, rigid transformation is carried out on the medical image to be registered by using the transformation matrix, and a first medical image to be registered after rigid registration is obtained. And finally, carrying out rigid transformation on the first organ delineation information by using the transformation matrix to obtain rigid registered third organ delineation information.

The accuracy of the registration of the medical images can be effectively improved by rigidly registering the medical images to be registered through the reference medical images.

In addition, if rigid registration is performed on the medical image to be registered by referring to the medical image, when resolution grading is performed correspondingly, resolution grading is performed on the first medical image to be registered after rigid registration to obtain N layers of first medical images to be registered.

In the registration method of medical images provided in the embodiment of the present application, the performing resolution classification on a reference medical image and a medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered includes: respectively carrying out Gaussian filtering on the reference medical image and the medical image to be registered to obtain a filtered reference medical image and a filtered medical image to be registered; and respectively carrying out down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered.

Specifically, the resolution grading of the reference medical image and the medical image to be registered includes performing gaussian filtering smoothing processing on the reference medical image and the medical image to be registered respectively to obtain a filtered reference medical image and a filtered medical image to be registered, and then performing down-sampling processing or down-sampling processing on the filtered reference medical image and the filtered medical image to be registered respectively to obtain N layers of reference medical images and N layers of medical images to be registered, where the resolutions are from high to low.

It should be noted that, the step of performing resolution classification on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information is consistent with the step of performing resolution classification on the reference medical image and the medical image to be registered, and therefore, the steps are not repeated.

Therefore, in the registration method of medical images provided in the embodiment of the present application, determining a target deformation field according to N layers of reference medical images, N layers of medical images to be registered, N layers of first target organ delineation information, and N layers of second target organ delineation information includes: setting an initial deformation field of a first layer for a first layer of reference medical image and a first layer of medical image to be registered, and obtaining a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered and second target organ delineation information corresponding to the first layer of medical image to be registered; iteratively updating the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of a first layer; for the M layer of parameter medical image and the M layer of medical image to be registered, determining the deformation field of the M-1 layer as the initial deformation field of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N; and performing up-sampling processing on the deformation field of the Nth layer, and determining the processed deformation field of the Nth layer as a target deformation field.

Specifically, for the first layer reference medical image and the first layer medical image to be registered, an initial deformation field corresponding to the first layer is set, for example, the initial deformation field is set to a zero matrix with a size of [ x, y, z,3], where 3 represents an offset of the pixel in three directions (x, y, z). The offset will change each time the field is updated. It should be noted that the first layer reference medical image and the first layer medical image to be registered are the lowest resolution images of the N layers of reference medical images and the N layers of medical images to be registered.

And deforming the first layer medical image to be registered through the initial deformation field corresponding to the first layer, iteratively updating the initial deformation field according to an objective constraint function of a gradient descent method until the updating iteration times reach a specified time (for example, 50 times), and returning to the deformation field corresponding to the first layer.

It should be noted that the target constraint function is obtained through the first layer of reference medical image, the first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered, and the second target organ delineation information corresponding to the first layer of medical image to be registered.

And after the deformation field corresponding to the first layer is obtained, taking the deformation field corresponding to the first layer as the initial deformation field of the next layer. Because the resolution ratios of the medical image of the first layer and the medical image of the second layer are different, the deformation field corresponding to the first layer can be subjected to upsampling processing, the processed deformation field corresponding to the first layer is used as the initial deformation field of the next layer, then the iteration process of the initial deformation field is repeated to obtain the deformation field corresponding to the next layer until the corresponding deformation field of the last layer is obtained, and finally the deformation field corresponding to the last layer is subjected to upsampling processing to obtain the final target deformation field.

Therefore, in the registration method of medical images provided in the embodiment of the present application, obtaining the target constraint function according to the first layer of reference medical image, the first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered, and the second target organ delineation information corresponding to the first layer of medical image to be registered includes: deforming the first layer of medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer of medical image to be registered; calculating according to pixel points in the deformed first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function; calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; and determining the target constraint function according to the first constraint function, the second constraint function, the third constraint function and the fourth constraint function.

Specifically, the target constraint function is composed of four constraint functions, specifically as follows:

(1) In order to measure the degree of alignment of the medical image to be registered with the reference medical image after each deformation, a mode Independent neighbor probability Descriptor (a mode Independent similarity Descriptor) and a block-based self-similarity context (self-similarity context SSC) are used as similarity constraints (i.e. the first constraint function mentioned above) between the two images. And calculating according to the deformed pixels in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function.

(2) And calculating a similarity value (also called a Dice value) between the deformed delineation data of the first layer of medical image to be registered and the delineation data corresponding to the first layer of reference medical image to serve as a constraint condition for organ delineation guidance, namely calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the first layer of medical image to be registered to obtain a second constraint function.

(3) In order to improve the smoothness of the deformation field, a diffusion regularizer is used for smooth constraint and averaging is carried out to serve as a deformation field smoothness constraint condition, namely a third constraint function is obtained according to the gradient value of the initial deformation field of the first layer in the target direction.

In an alternative embodiment, the third constraint function is L _smooth The corresponding functional expression is shown as follows:

the mean is a mean function, phi is the current deformation field, and V U (phi) is the gradient value of the current deformation field in the x, y and z directions.

(4) To optimize the folding problem, jacobian matrix coefficients are introduced as a fourth constraint function.

In an alternative embodiment, the fourth constraint function is L _Jac The corresponding functional expression is shown in the following formula:

wherein, JD ₀ ，JD ₁ ，JD ₂ Three direction matrixes of the current deformation field Jacobian matrix are respectively as follows:

wherein 0 corresponds to the x-direction, 1 corresponds to the y-direction, and 2 corresponds to the z-direction.

After the first constraint function, the second constraint function, the third constraint function and the fourth constraint function are obtained, the target constraint function is determined by using the constraint functions.

In conclusion, by introducing the first constraint function, the second constraint function, the third constraint function and the fourth constraint function, the generalization and accuracy of the deformation field are improved, the deformation of the crisis organ or the target region is further guided, and the method is suitable for multi-modal images.

In the registration method for medical images provided in the embodiment of the present application, the calculating according to the deformed pixel points in the first layer of medical image to be registered and the first layer of reference medical image to obtain the first constraint function includes: calculating the sum of first means of variance of pixel points in the deformed first layer medical image to be registered, and calculating according to the sum of the first means of variance and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image, and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and calculating according to the first self-similarity function and the second self-similarity function to obtain a first constraint function.

Specifically, the calculating of the first constraint function includes: and calculating pixel points in the deformed first layer medical image to be registered to obtain a first mean square difference sum, and calculating according to the first mean square difference sum and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function. And calculating pixel points in the first layer of reference medical image to obtain a second square difference sum, and calculating according to the second square difference sum and the noise estimation in the first layer of reference medical image to obtain a second self-similarity function. And finally, calculating according to the first self-similarity function and the second self-similarity function to obtain a first constraint function.

In an alternative embodiment, the first self-similarity function and the second self-similarity function are calculated using the following formulas:

wherein the reference medical image is F, MS _F For the second self-similarity function, the deformed first layer medical image to be registered is Mo phi, MS _MoΦ Is the first self-similarity function described above. SSD (x, y) is the sum of squared differences, sigma, of pixel points in the reference medical image F or the first layer medical image Mo phi to be registered ² For noise estimation in medical images, N is a specific domain for computing self-similar context, y is the center point of N, and x is all points of N. The specific field refers to the field that the pixel point y is used as a central point, and a plurality of pixel points are selected from eight adjacent pixel points of the pixel point y.

After the first self-similarity function and the second self-similarity function are obtained through calculation, a first constraint function is obtained through calculation by adopting the following formula:

in summary, the alignment degree between the medical image to be aligned and the reference image after deformation at each time is used as a constraint function, so that the method is applicable to multi-modal medical images, and the accuracy of deformation of the target deformation field is effectively improved.

In the registration method of the medical image provided in the embodiment of the present application, the following steps are adopted to calculate according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered, so as to obtain a second constraint function: acquiring first target organ delineation information of each organ in the first layer of reference medical image, and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; similarity calculation is carried out on first target organ delineation information of each organ in the first layer of reference medical image and second target organ delineation information of each organ in the deformed first layer of medical image to be registered, and a similarity constraint value corresponding to each organ is obtained; and acquiring the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value and the total volume value of each organ and the similarity constraint value corresponding to each organ to obtain a second constraint function.

Specifically, first target organ delineation information of each organ in a first layer of reference medical image and second target organ delineation information of each organ in a first layer of medical image to be registered are respectively obtained, then the similarity between the first target organ delineation information of each organ and the second target organ delineation information of each organ is calculated, a similarity constraint value corresponding to each organ is obtained, then the weight of each organ is distributed according to a plurality of organ volume proportions, and finally a second constraint function is obtained.

In an alternative embodiment, the similarity constraint value corresponding to each organ is calculated by the following formula:

wherein the content of the first and second substances,

for the above-mentioned similarity constraint value for the k-th organ, < >>

Drawing information for a first target organ of the kth organ, based on the comparison result>

And drawing information for a second target organ of the kth organ of the deformed first layer medical image to be registered.

In an alternative embodiment, the second constraint function is obtained by the following equation:

wherein the content of the first and second substances,

in the second constraint function mentioned above, T is the total number of organs,. T is>

Represents the volume delineated by the kth organ>

Is the total volume>

W is a constant weight for the similarity constraint value of the kth organ. In specific implementation, W =4 may be set.

Through quoting a plurality of organ sketches, further carry out deformation to the organ region, can effectively guide the deformation in crisis organ or target area region, improve the accuracy of deformation.

In order to improve the reasonableness of the setting of the target constraint function, in the registration method of the medical image provided by the embodiment of the application, the determining the target constraint function according to the first constraint function, the second constraint function, the third constraint function and the fourth constraint function includes: respectively setting weight values for a first constraint function, a second constraint function, a third constraint function and a fourth constraint function to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function; and calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

Specifically, all the above constraint functions L _sim ，L _dice ，L _smooth And L _Jac Respectively setting the weight value of each constraint function through weight integration, wherein the first weight value corresponding to the first constraint function is W ₀ The second weight value corresponding to the second constraint function is W ₁ The third weight value corresponding to the third constraint function is W ₂ And a fourth weight value corresponding to the fourth constraint function is W ₃ . And integrating through the weight values to obtain the target constraint function.

In an alternative embodiment, the objective constraint function is as follows:

in concrete implementation, W may be set ₀ =4，W ₁ =1，W ₂ =1.2，W ₃ =1。

In an alternative embodiment, the flowchart shown in fig. 2 may be used to realize the registration of the medical image, and the reference image, the floating image, and the organ delineation information corresponding to the reference image and the organ delineation information corresponding to the floating image are read. And carrying out rigid registration on the data, and grading the image data based on a multi-resolution pyramid iterative method. Initializing a deformation field for the image data of the ith layer, carrying out deformation registration on the data based on the deformation field, calculating a constraint function based on the deformation result, iterating the initialized deformation field by using the constraint function to obtain the deformation field of the ith layer, taking the deformation field of the ith layer as the initial deformation field of the next layer, repeatedly executing the steps until the deformation field of the last layer is obtained, then standardizing the size of the deformation field of the last layer to obtain a target deformation field, and finally carrying out registration on the floating image by using the target deformation field to obtain the deformation registration result.

The registration method of the medical image provided by the embodiment of the application acquires medical image information, wherein the medical image information at least comprises: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information, so that the problem of low accuracy of image registration caused by the fact that deformation registration depends on global deformation of the medical image during image registration in the related technology is solved. According to the scheme, before the medical image to be registered is registered, the resolution of the image to be registered and the resolution of the reference medical image are graded, so that N layers of reference medical images and N layers of medical images to be registered are obtained. Therefore, the calculated amount of the deformation field during iteration can be effectively reduced, and then the deformation field is calculated by utilizing the organ delineation information, so that the accuracy and the reasonability of the organ deformation are further improved, and the effect of improving the accuracy of image registration is further achieved.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer-executable instructions and that, although a logical order is illustrated in the flowcharts, in some cases, the steps illustrated or described may be performed in an order different than here.

The embodiment of the present application further provides a registration apparatus for medical images, and it should be noted that the registration apparatus for medical images according to the embodiment of the present application may be used to execute the registration method for medical images according to the embodiment of the present application. The following describes a registration apparatus for medical images provided by an embodiment of the present application.

Fig. 3 is a schematic diagram of a registration apparatus of medical images according to an embodiment of the application. As shown in fig. 3, the apparatus includes: an acquisition unit 301, a ranking unit 302, a determination unit 303 and a first registration unit 304.

An obtaining unit 301, configured to obtain medical image information, where the medical image information at least includes: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered;

the grading unit 302 is configured to perform resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and perform resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, where N is a positive integer greater than or equal to 2;

the determining unit 303 is configured to determine a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information, and the N layers of second target organ delineation information;

a first registration unit 304, configured to register the medical image to be registered and the second organ delineation information according to the target deformation field, so as to obtain a registered medical image and registered second organ delineation information.

The registration apparatus for medical images provided by the embodiment of the present application acquires medical image information through an acquisition unit 301, where the medical image information at least includes: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; the grading unit 302 performs resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performs resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, where N is a positive integer greater than or equal to 2; the determining unit 303 determines a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; the first registration unit 304 registers the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information, so that the problem that the accuracy of image registration is low due to the fact that deformation registration depends on global deformation of the medical image when image registration is performed in the related art is solved. In the scheme, before the medical image to be registered is registered, the resolution of the image to be registered and the resolution of the reference medical image are graded to obtain N layers of reference medical images and N layers of medical images to be registered. Therefore, the calculated amount of the deformation field during iteration can be effectively reduced, and then the deformation field is calculated by utilizing the organ delineation information, so that the accuracy and the reasonability of the organ deformation are further improved, and the effect of improving the accuracy of image registration is further achieved.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, after acquiring the medical image information, the apparatus further includes: the second registration unit is used for performing rigid registration on the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and performing rigid registration on second organ delineation information of the medical image to be registered according to the first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; the classification unit includes: the first grading subunit is used for carrying out resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and the second grading subunit is used for carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the second registration unit includes: the acquisition subunit is used for acquiring a conversion matrix between the reference medical image and the medical image to be registered; and the first registration subunit is used for performing rigid registration on the medical image to be registered according to the transformation matrix to obtain a first medical image to be registered.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the classifying unit includes: the filtering subunit is used for respectively carrying out Gaussian filtering on the reference medical image and the medical image to be registered to obtain a filtered reference medical image and a filtered medical image to be registered; and the down-sampling sub-unit is used for respectively performing down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered.

Optionally, in a registration apparatus for medical images provided in an embodiment of the present application, the determining unit includes: the device comprises a setting subunit, a target constraint function acquiring unit and a registration processing unit, wherein the setting subunit is used for setting an initial deformation field of a first layer for a first layer of reference medical images and a first layer of medical images to be registered, and acquiring a target constraint function according to the first layer of reference medical images, first target organ delineation information corresponding to the first layer of reference medical images, the first layer of medical images to be registered and second target organ delineation information corresponding to the first layer of medical images to be registered; the iteration subunit is used for iteratively updating the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of the first layer; the determining subunit is used for determining the deformation field of the M-1 layer as the initial deformation field of the M layer for the parameter medical image of the M layer and the medical image to be registered of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N; and the up-sampling subunit is used for performing up-sampling treatment on the deformation field of the Nth layer and determining the treated deformation field of the Nth layer as a target deformation field.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the setting subunit includes: the deformation module is used for deforming the first layer of medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer of medical image to be registered; the first calculation module is used for calculating according to the deformed pixels in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function; the second calculation module is used for calculating according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; the first processing module is used for obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; the second processing module is used for obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; and the determining module is used for determining the target constraint function according to the first constraint function, the second constraint function, the third constraint function and the fourth constraint function.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the first calculation module includes: the first calculation submodule is used for calculating the sum of the first mean square differences of pixel points in the deformed first layer medical image to be registered and calculating according to the sum of the first mean square differences and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; the second calculation submodule is used for calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and the third calculation submodule is used for calculating according to the first self-similarity function and the second self-similarity function to obtain a first constraint function.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the second calculating module includes: the acquisition submodule is used for acquiring first target organ delineation information of each organ in the first layer of reference medical image and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; the fourth calculation submodule is used for carrying out similarity calculation on the first target organ delineation information of each organ in the first layer of reference medical image and the second target organ delineation information of each organ in the deformed first layer of medical image to be registered to obtain a similarity constraint value corresponding to each organ; and the fifth calculation submodule is used for acquiring the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value and the total volume value of each organ and the similarity constraint value corresponding to each organ to obtain a second constraint function.

Optionally, in the registration apparatus for medical images provided in the embodiment of the present application, the determining module includes: the setting submodule is used for setting weight values for the first constraint function, the second constraint function, the third constraint function and the fourth constraint function respectively to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function; and the sixth calculating submodule is used for calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

The registration apparatus for medical images comprises a processor and a memory, wherein the above-mentioned acquisition unit 301, the ranking unit 302, the determination unit 303, the first registration unit 304, etc. are stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. One or more than one kernel can be set, and the registration of the medical image is realized by adjusting the kernel parameters.

The memory may include volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), including at least one memory chip.

In an alternative embodiment, the registration of medical images may be achieved using a system as shown in fig. 4. The system comprises: and the data reading module is used for the reference image, the floating image, the organ delineation information corresponding to the reference image and the organ delineation information corresponding to the floating image. And the rigid registration module is used for carrying out rigid registration on the data. The data grading module is used for grading the medical image and the drawing information; the deformation field calculation module is used for calculating to obtain a target deformation field; and the data deformation registration module is used for registering the floating image by using the target deformation field. And the output module is used for outputting the deformation registration result.

An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which when executed by a processor, implements a registration method for medical images.

An embodiment of the invention provides a processor for running a program, wherein the program executes a registration method of a medical image during running.

As shown in fig. 5, an embodiment of the present invention provides an electronic device, where the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and the processor implements the following steps when executing the program: acquiring medical image information, wherein the medical image information at least comprises: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

Optionally, after acquiring the medical image information, the method further comprises: rigidly registering the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and rigidly registering second organ delineation information of the medical image to be registered according to the first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein the steps of: performing resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

Optionally, the rigidly registering the medical image to be registered according to the reference medical image to obtain the first medical image to be registered includes: acquiring a conversion matrix between a reference medical image and a medical image to be registered; and carrying out rigid registration on the medical image to be registered according to the transformation matrix to obtain a first medical image to be registered.

Optionally, the step of performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered includes: respectively carrying out Gaussian filtering on the reference medical image and the medical image to be registered to obtain a filtered reference medical image and a filtered medical image to be registered; and respectively carrying out down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered.

Optionally, determining the target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information, and the N layers of second target organ delineation information includes: setting an initial deformation field of a first layer for a first layer of reference medical image and a first layer of medical image to be registered, and obtaining a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered and second target organ delineation information corresponding to the first layer of medical image to be registered; iteratively updating the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of a first layer; for the M layer of parameter medical image and the M layer of medical image to be registered, determining the deformation field of the M-1 layer as the initial deformation field of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N; and performing up-sampling processing on the deformation field of the Nth layer, and determining the processed deformation field of the Nth layer as a target deformation field.

Optionally, the obtaining a target constraint function according to the first layer of reference medical image, the first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered, and the second target organ delineation information corresponding to the first layer of medical image to be registered includes: deforming the first layer of medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer of medical image to be registered; calculating according to pixel points in the deformed first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function; calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; and determining an objective constraint function according to the first constraint function, the second constraint function, the third constraint function and the fourth constraint function.

Optionally, the calculating according to the deformed pixel points in the first layer of medical image to be registered and the first layer of reference medical image to obtain the first constraint function includes: calculating the sum of first square differences of pixel points in the deformed first layer medical image to be registered, and calculating according to the sum of the first square differences and noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image, and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and calculating according to the first self-similarity function and the second self-similarity function to obtain a first constraint function.

Optionally, the obtaining a second constraint function by calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered includes: acquiring first target organ delineation information of each organ in the first layer of reference medical image, and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; similarity calculation is carried out on first target organ delineation information of each organ in the first layer of reference medical image and second target organ delineation information of each organ in the deformed first layer of medical image to be registered, and a similarity constraint value corresponding to each organ is obtained; and obtaining the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value and the total volume value of each organ and the similarity constraint value corresponding to each organ to obtain a second constraint function.

Optionally, determining the target constraint function according to the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function includes: respectively setting weight values for a first constraint function, a second constraint function, a third constraint function and a fourth constraint function to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function; and calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

The device herein may be a server, a PC, a PAD, a mobile phone, etc.

The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device: acquiring medical image information, wherein the medical image information at least comprises: the method comprises the steps of referring to a medical image, referring to first organ delineation information of the medical image, a medical image to be registered and second organ delineation information of the medical image to be registered; performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2; determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information; and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

Optionally, after acquiring the medical image information, the method further comprises: rigidly registering the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and rigidly registering second organ delineation information of the medical image to be registered according to the first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered; performing resolution classification on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution classification on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein the steps of: performing resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered; and carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

Optionally, the obtaining a target constraint function according to the first layer of reference medical image, the first target organ delineation information corresponding to the first layer of reference medical image, the first layer of medical image to be registered, and the second target organ delineation information corresponding to the first layer of medical image to be registered includes: deforming the first layer of medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer of medical image to be registered; calculating according to the deformed pixels in the first layer of medical image to be registered and the first layer of reference medical image to obtain a first constraint function; calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function; obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction; obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer; and determining an objective constraint function according to the first constraint function, the second constraint function, the third constraint function and the fourth constraint function.

Optionally, the calculating according to the deformed pixel points in the first layer of medical image to be registered and the first layer of reference medical image to obtain the first constraint function includes: calculating the sum of first means of variance of pixel points in the deformed first layer medical image to be registered, and calculating according to the sum of the first means of variance and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function; calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image, and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function; and calculating according to the first self-similarity function and the second self-similarity function to obtain a first constraint function.

Optionally, the obtaining a second constraint function by calculating according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered includes: acquiring first target organ delineation information of each organ in the first layer of reference medical image, and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered; similarity calculation is carried out on first target organ delineation information of each organ in the first layer of reference medical image and second target organ delineation information of each organ in the deformed first layer of medical image to be registered, and a similarity constraint value corresponding to each organ is obtained; and obtaining the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value and the total volume value of each organ and the similarity constraint value corresponding to each organ to obtain a second constraint function.

Optionally, determining the target constraint function from the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function comprises: respectively setting weight values for the first constraint function, the second constraint function, the third constraint function and the fourth constraint function to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function; and calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method of registration of medical images, comprising:

acquiring medical image information, wherein the medical image information at least comprises: the method comprises the steps of obtaining a reference medical image, first organ delineation information of the reference medical image, a medical image to be registered and second organ delineation information of the medical image to be registered;

performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2;

determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information;

and registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

2. The method of claim 1,

after acquiring the medical image information, the method further comprises:

carrying out rigid registration on the medical image to be registered according to the reference medical image to obtain a first medical image to be registered, and carrying out rigid registration on second organ delineation information of the medical image to be registered according to first organ delineation information of the reference medical image to obtain third organ delineation information of the medical image to be registered;

performing resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and performing resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, including:

performing resolution grading on the reference medical image and the first medical image to be registered to obtain N layers of reference medical images and N layers of first medical images to be registered;

and carrying out resolution grading on the first organ delineation information of the reference medical image and the third organ delineation information of the medical image to be registered to obtain N layers of first target organ delineation information and N layers of second target organ delineation information.

3. The method according to claim 2, wherein rigidly registering the medical image to be registered according to the reference medical image, resulting in a first medical image to be registered comprises:

acquiring a conversion matrix between the reference medical image and the medical image to be registered;

and carrying out rigid registration on the medical image to be registered according to the transformation matrix to obtain the first medical image to be registered.

4. The method according to claim 1, wherein performing resolution scaling on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered comprises:

respectively carrying out Gaussian filtering on the reference medical image and the medical image to be registered to obtain a filtered reference medical image and a filtered medical image to be registered;

and respectively performing down-sampling processing on the filtered reference medical image and the filtered medical image to be registered to obtain the N layers of reference medical images and the N layers of medical images to be registered.

5. The method of claim 1, wherein determining a target deformation field from the N layers of reference medical images, N layers of medical images to be registered, the N layers of first target organ delineation information, and the N layers of second target organ delineation information comprises:

setting an initial deformation field of a first layer for a first layer of reference medical image and a first layer of medical image to be registered, and obtaining a target constraint function according to the first layer of reference medical image, first target organ delineation information corresponding to the first layer of reference medical image, and second target organ delineation information corresponding to the first layer of medical image to be registered and the first layer of medical image to be registered;

iteratively updating the initial deformation field to a preset number of times according to the target constraint function to obtain a deformation field of a first layer;

for the M layer of parameter medical image and the M layer of medical image to be registered, determining the deformation field of the M-1 layer as the initial deformation field of the M layer, and performing iterative processing on the initial deformation field of the M layer to obtain the deformation field of the M layer, wherein M is less than or equal to N;

and performing up-sampling treatment on the deformation field of the Nth layer, and determining the treated deformation field of the Nth layer as the target deformation field.

6. The method of claim 5, wherein deriving an object constraint function from the first layer of reference medical images, first target organ delineation information corresponding to the first layer of reference medical images, the first layer of medical images to be registered, and second target organ delineation information corresponding to the first layer of medical images to be registered comprises:

deforming the first layer medical image to be registered according to the initial deformation field of the first layer to obtain a deformed first layer medical image to be registered;

calculating according to the deformed pixel points in the first layer medical image to be registered and the first layer reference medical image to obtain a first constraint function;

calculating according to first target organ delineation information corresponding to the first layer of reference medical image and second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain a second constraint function;

obtaining a third constraint function according to the gradient value of the initial deformation field of the first layer in the target direction;

obtaining a fourth constraint function according to the Jacobian matrix of the initial deformation field of the first layer;

determining the target constraint function in dependence on the first, second, third and fourth constraint functions.

7. The method according to claim 6, wherein calculating according to the deformed pixel points in the first layer medical image to be registered and the first layer reference medical image to obtain the first constraint function comprises:

calculating the sum of first means of variance of pixel points in the deformed first layer medical image to be registered, and calculating according to the sum of the first means of variance and the noise estimation in the deformed first layer medical image to be registered to obtain a first self-similarity function;

calculating the sum of second mean square deviations of pixel points in the first layer of reference medical image, and calculating according to the sum of the second mean square deviations and noise estimation in the first layer of reference medical image to obtain a second self-similarity function;

and calculating according to the first self-similarity function and the second self-similarity function to obtain the first constraint function.

8. The method according to claim 6, wherein calculating according to the first target organ delineation information corresponding to the first layer of reference medical image and the second target organ delineation information corresponding to the deformed first layer of medical image to be registered to obtain the second constraint function comprises:

acquiring first target organ delineation information of each organ in the first layer of reference medical image, and acquiring second target organ delineation information of each organ in the deformed first layer of medical image to be registered;

similarity calculation is carried out on the first target organ delineation information of each organ in the first layer of reference medical image and the second target organ delineation information of each organ in the deformed first layer of medical image to be registered, and a similarity constraint value corresponding to each organ is obtained;

and acquiring the volume value of each organ and the total volume values of all the organs, and calculating according to the volume value of each organ, the total volume value and the similarity constraint value corresponding to each organ to obtain the second constraint function.

9. The method of claim 6, wherein determining the target constraint function from the first constraint function, the second constraint function, the third constraint function, and the fourth constraint function comprises:

respectively setting weight values for the first constraint function, the second constraint function, the third constraint function and the fourth constraint function to obtain a first weight value corresponding to the first constraint function, a second weight value corresponding to the second constraint function, a third weight value corresponding to the third constraint function and a fourth weight value corresponding to the fourth constraint function;

and calculating according to the first constraint function, the second constraint function, the third constraint function, the fourth constraint function, the first weight value, the second weight value, the third weight value and the fourth weight value to obtain the target constraint function.

10. A registration apparatus for medical images, comprising:

an obtaining unit, configured to obtain medical image information, wherein the medical image information at least includes: the method comprises the steps of obtaining a reference medical image, first organ delineation information of the reference medical image, a medical image to be registered and second organ delineation information of the medical image to be registered;

the grading unit is used for carrying out resolution grading on the reference medical image and the medical image to be registered to obtain N layers of reference medical images and N layers of medical images to be registered, and carrying out resolution grading on the first organ delineation information and the second organ delineation information to obtain N layers of first target organ delineation information and N layers of second target organ delineation information, wherein N is a positive integer greater than or equal to 2;

the determining unit is used for determining a target deformation field according to the N layers of reference medical images, the N layers of medical images to be registered, the N layers of first target organ delineation information and the N layers of second target organ delineation information;

and the first registration unit is used for registering the medical image to be registered and the second organ delineation information according to the target deformation field to obtain the registered medical image and the registered second organ delineation information.

11. A computer-readable storage medium, characterized in that the storage medium stores a program, wherein the program performs the registration method of a medical image according to any one of claims 1 to 9.

12. An electronic device comprising one or more processors and memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of registration of medical images of any of claims 1-9.