CN114677415A - Image registration method, apparatus, computer equipment and readable storage medium - Google Patents

Abstract

The present application relates to an image registration method, apparatus, computer equipment and readable storage medium. The method includes: acquiring respiration data and initial dynamic PET data within a first preset time period after the tracer is administered to the object to be tested, and determining corresponding multi-frame dynamic MR images within the first preset time period according to the respiration data; Obtain the dynamic-like PET images corresponding to the dynamic MR images of each frame, and obtain the registered dynamic-like PET images from the dynamic-like PET images of each frame. Registration is performed to obtain a registered dynamic PET image. The method can obtain the registered dynamic PET images, and then register multiple frames of initial dynamic PET images based on the registered dynamic PET images, so that the registered dynamic PET images can reflect the structural information and obvious Therefore, the accuracy of dynamic PET image registration results can be improved.

Description

Image registration method, apparatus, computer equipment and readable storage medium

technical field

The present application relates to the technical field of medical image processing, and in particular, to an image registration method, apparatus, computer equipment and readable storage medium.

Background technique

Kinetic parameter analysis based on dynamic positron emission computed tomography (PET) can reveal the biochemical process of the tracer and provide a basis for clinically revealing the pathological mechanism.

In order to analyze the biochemical process of the tracer, it is necessary to inject the tracer into the detection imaging site of the object to be tested, and then collect the initial PET images for a continuous period of time, and then analyze the kinetic parameters of the initial PET images. Since it is difficult for the object to be tested to keep the body posture completely fixed for a long time, the initial PET images collected at different time points will be displaced. Therefore, before the dynamic parameter analysis, it is generally necessary to perform registration processing on the initial PET images to improve the dynamic Accuracy of quantification of scientific parameters.

However, in the related art, when the initial PET image is registered, there is a problem that the registration result is not accurate enough.

SUMMARY OF THE INVENTION

Based on this, it is necessary to provide an image registration method, apparatus, computer device and readable storage medium for the above technical problems.

In a first aspect, the present application provides an image registration method, the method comprising:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET data are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

In one embodiment, based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images, including:

Based on the dynamic PET images of each frame type, time mapping is performed on the initial dynamic PET data within the first preset time period to obtain the mapped initial dynamic PET data of the dynamic PET images of each frame type;

Reconstructing the mapped initial dynamic PET data of each frame-like dynamic PET image to obtain the corresponding mapped initial dynamic PET image;

Through the registered quasi-dynamic PET images, the corresponding mapped initial dynamic PET images are registered to obtain the registered dynamic PET images.

In one embodiment, the registered dynamic-like PET images are determined by each frame of dynamic-like PET images, including:

Obtain the deformation field corresponding to each frame type of dynamic MR image;

According to the deformation field corresponding to each frame-like dynamic MR image, the corresponding dynamic-like PET image is registered to obtain the registered dynamic-like PET image.

In one embodiment, obtaining the deformation field corresponding to each frame-like dynamic MR image includes:

determining a reference image based on multiple frames of sample MR images;

Image registration is performed on the dynamic MR images of each frame type through the reference image, and the deformation fields corresponding to the dynamic MR images of each frame type are obtained.

In one embodiment, determining the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data includes:

The respiratory data is input into the prediction model to obtain corresponding multi-frame dynamic MR images within the first preset time period.

In one embodiment, the above method further includes:

acquiring multiple frames of sample MR images and sample respiration data within a second preset time period before the subject to be tested is administered the tracer;

The initial prediction model is trained by using multiple frames of sample MR images and sample respiration data to obtain the prediction model.

In one embodiment, acquiring the dynamic-like PET images corresponding to the dynamic-like MR images of each frame includes:

obtaining initial PET data within a first preset time period;

performing time mapping on the dynamic MR images of each frame type and the initial PET data in the first preset time period to obtain the mapped PET data of the dynamic MR images of each frame type;

The mapped PET data is reconstructed to obtain the dynamic-like PET images corresponding to the dynamic-like MR images of each frame.

In a second aspect, the present application provides an image registration device, the device comprising:

a data acquisition module, used to acquire respiratory data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

a first image determination module, configured to determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

The second image determination module is used to obtain the dynamic-like PET images corresponding to the dynamic MR images of each frame, and determine the registered dynamic-like PET images through the dynamic-like PET images of each frame;

The registration module is used for registering multiple frames of initial dynamic PET images according to the registered quasi-dynamic PET images to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are obtained by reconstructing the initial dynamic PET data Image.

In a third aspect, the present application provides a computer device, including a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

In a fourth aspect, the present application provides a readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

The above image registration method, device, computer equipment and readable storage medium, the computer equipment obtains the respiration data and the initial dynamic PET data within the first preset time period after the tracer is administered to the object to be tested, and determines the first image according to the respiration data. The corresponding multi-frame dynamic MR images within a preset time period are obtained, and the dynamic PET images corresponding to each frame of dynamic MR images are obtained, and the registered dynamic PET images are determined from the dynamic PET images of each frame. The dynamic-like PET image is obtained by registering multiple frames of the initial dynamic PET image to obtain the registered dynamic PET image; the above method can directly obtain the dynamic-like MR image through the respiratory data, and then based on the dynamic-like MR image and the initial dynamic PET data Obtain a dynamic-like PET image with the structural features of the tissue/organ of the imaging site, and based on the registered dynamic-like PET image, perform registration on multiple frames of the initial dynamic PET image, so that the registered dynamic PET image can reflect the structural information. And obvious image features, which can improve the accuracy of dynamic PET image registration results.

Description of drawings

1 is an application environment diagram of an image registration method in one embodiment;

2 is a schematic flowchart of an image registration method in one embodiment;

3 is a schematic flowchart of a method for registering multiple frames of initial dynamic PET images in one embodiment;

4 is a schematic flowchart of a method for obtaining a registered dynamic-like PET image through each frame of dynamic-like PET images in another embodiment;

5 is a schematic flowchart of a method for obtaining a deformation field corresponding to each frame type dynamic MR image in another embodiment;

6 is a schematic diagram of multiple frames of different types of images in an image registration process in another embodiment;

7 is a schematic flowchart of a method for obtaining a dynamic-like PET image corresponding to each frame of dynamic-like MR images in another embodiment;

Fig. 8 is a corresponding relationship diagram between time points and sample MR data and respiration data in another embodiment;

9 is a structural block diagram of an image registration apparatus in one embodiment;

Figure 10 is a diagram of the internal structure of a computer device in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

The image registration method provided by the present application can be applied to the image registration system shown in FIG. 1 , and can be applied to medical image registration or medical image calibration scenarios. The above-mentioned image registration system includes a scanning system and computer equipment. The scanning system includes a medical scanning equipment and a tool capable of carrying the object to be measured. The tool can be a scanning bed, a scanning frame, a scanning board, and the like. The computer device and the medical scanning device in the scanning system may be a communication connection, and the communication mode may be Wi-Fi, a mobile network, or a Bluetooth connection, and so on. The above-mentioned medical scanning device can be an electronic computed tomography system, a computer X-ray photography system or a direct digital X-ray photography system, a magnetic resonance (Magnetic Resonance, MR) scanning device, etc. Scanning system; the above-mentioned computer equipment can be various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices, but not limited to these. In practical applications, the object to be tested can lie on the carrier in different postures, the medical scanning device can scan the target imaging part of the object to be tested to obtain scan data, the medical scanning device sends the scanned data to the computer device, and the computer device The scan data is reconstructed and analyzed to achieve image registration. Optionally, after the medical scanning device scans, the magnetic resonance coil in the medical scanning device can be removed from the body of the object to be measured, and the carrying tool can be returned to its position. Optionally, the target imaging site may be imaging sites such as the brain, lung, abdomen, heart, blood vessels, and joints of the object to be tested; the object to be tested may be a human body, an animal body, or the like. In medicine, in order to provide a basis for clinically revealing the pathological mechanism, it is necessary to analyze the dynamic parameters of the scanned medical images.

Among them, kinetic parameter analysis can reveal the biochemical process of the tracer. Therefore, it is necessary to administer the tracer to the test object, and then scan the test object to which the tracer is applied to obtain a medical image, and then obtain the medical image at this time. Dynamic parameter analysis of medical images. Optionally, the object to be tested is limited by the accumulation time characteristics of the tracer in the body. Therefore, in this embodiment, it is necessary to continuously collect data for a certain period of time, generate a dynamic medical image after reconstruction, and then perform dynamic parameter analysis on the dynamic medical image. In order to reveal the biochemical process of the tracer, further medical staff can determine the pathological result of the object to be tested according to the biochemical process of the tracer. In the actual scene, it is difficult for the object to be tested to maintain a complete body posture for a long time, and the obtained dynamic medical image may deviate from the actual one. Registration processing to improve the accuracy of kinetic parameter quantification. The above dynamic medical images may be dynamic positron emission tomography (Positron Emission Computed Tomography, PET) images, Computed Tomography (Computed Tomography, CT) images, MR images, and the like. However, in this embodiment, the dynamic medical image is a dynamic PET image.

In one embodiment, as shown in FIG. 2, an image registration method is provided, and the method is applied to a computer device as an example for description, including the following steps:

S100. Acquire respiration data and initial dynamic PET data within a first preset time period after the subject to be tested is administered the tracer.

Specifically, before the medical scanning device scans any one or more imaging parts of the object to be measured, the medical staff can firstly apply a certain amount of tracer to the vein of the one or more imaging parts of the object to be measured. Accumulation of the tracer, within a specific period of time, will produce a certain biochemical reaction with the tissue/organ at the imaging site to which the tracer is administered, and during this biochemical reaction, the tissue/organ in the focal area is compared with the non-focal area. The tissues/organs will produce obvious differences, which can be reflected from dynamic PET images.

Optionally, during the biochemical reaction of the tissue/organ, the medical scanning device may scan the imaging site where the tracer is administered to obtain a dynamic PET image, wherein, with the accumulation of the tracer in the tissue/organ at the imaging site , the contrast of the dynamic PET image will change significantly, but it is not that the longer the time, the greater the contrast of the dynamic PET image, and the contrast of the dynamic PET image will reach a peak within a period of time. Therefore, the above specific time period The time points at which the contrast of dynamic PET images peaked were included. Optionally, the starting time point of the specific time period may be the time point at which the tracer administration ends, or may be a certain time point after the time point at which the tracer administration ends. Optionally, the above-mentioned tracer may be a different nuclide drug, which is harmless to the object to be tested.

In practical applications, the medical scanning device may scan one or more imaging sites where the tracer is administered to the object to be tested, acquire initial PET data scanned within a first preset time period, and at the time of acquiring the initial PET data At the same time, the breathing data of the object to be tested can also be collected synchronously. The medical scanning device can send both the initial PET data and the respiration data in the first preset time period collected by the scan to the computer device, and the computer device can perform the initial PET data in the multiple sub-time periods in the first preset time period. Reconstruction to obtain multiple frames of dynamic PET images corresponding to the first preset time period. Each sub-period corresponds to one frame of dynamic PET image.

It should be noted that each frame of dynamic PET images may correspond to initial PET data in a sub-time period within the first preset time period. Optionally, the first preset time period may include a plurality of sub-time periods, and each sub-time period has a corresponding initial dynamic PET image; the time lengths corresponding to each sub-time period may be equal or unequal; The combined duration of the time periods may be less than or equal to the duration corresponding to the first preset time period.

In addition, the computer equipment can perform static reconstruction of the initial PET data scanned by the medical scanning equipment to obtain a static PET image, but in this embodiment, the computer equipment reconstructs the initial PET data scanned by the medical scanning equipment into a dynamic PET image, and reconstructs Dynamic reconstruction algorithms can be used for dynamic PET images. The contrast of the dynamic PET image is greater than that of the static PET image. The above dynamic reconstruction algorithm may be a back-projection method, an iterative reconstruction algorithm, a filtered back-projection method, a Fourier transform method, etc., which is not limited in this embodiment.

S200. Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the breathing data.

Specifically, the computer device may perform arithmetic operations, data conversion, analysis, data comparison and/or reconstruction, etc. on the respiratory data within the first preset time period, to obtain corresponding multi-frame classes within the first preset time period Dynamic MR images. Alternatively, the computer device may also perform preprocessing such as arithmetic operation, data conversion, analysis, data comparison and/or reconstruction on the respiratory data within the first preset time period, and then use a specific algorithm to perform specific preprocessing on the result. processing to obtain corresponding multi-frame dynamic MR images within the first preset time period. Optionally, the above arithmetic operations may be addition operations, subtraction operations, division operations, multiplication operations, exponential operations, and/or logarithmic operations, and the like.

It should be noted that the lengths of the sub-time periods corresponding to the dynamic MR images of each frame type may be equal or unequal. Optionally, the duration of the sub-period of the dynamic MR image of each frame type may be greater than or equal to the duration of the sub-period of the initial dynamic PET image of the corresponding frame. Optionally, multiple frames of quasi-dynamic MR images may be acquired within the first preset time period; each frame of quasi-dynamic MR images may be dynamic MR images generated within a certain sub-period within the first preset period of time. The sub-period corresponding to each frame of the dynamic MR image is different from the sub-period corresponding to each frame of the dynamic PET image.

S300 , acquiring a dynamic-like PET image corresponding to each frame of the dynamic-like MR image, and determining the registered dynamic-like PET image by using each frame of the dynamic-like PET image.

In this embodiment, the initial PET data within the first preset time period has corresponding multiple frames of dynamic-like PET images, and each frame of the initial dynamic PET image has a corresponding dynamic-like PET image. In this embodiment, the initial PET data may also be referred to as dynamic PET data. Optionally, the duration of the sub-period of the initial dynamic PET image of each frame may be less than or equal to the duration of the sub-period of the dynamic-like PET image of the corresponding frame.

It can be understood that the computer equipment can perform mapping processing, conversion processing, and/or analysis processing, etc., through the dynamic MR image of each frame and the dynamic PET image of the corresponding frame, so as to obtain the dynamic PET corresponding to each frame of the dynamic MR image of each frame. image. Optionally, the mapping process can be understood as the mapping between the dynamic MR image of the same size and the pixel value of the pixel at the corresponding position in the corresponding dynamic PET image, and the pixel values of the two pixels at the mapped position are superimposed. Or the process of subtraction and other operations. Optionally, the conversion process can be understood as a process of performing an arithmetic operation on a preset value or multiple preset values and pixel values of pixel points at different positions in the quasi-dynamic MR image. The analysis process can be understood as a process of analyzing the pixel resolution of each pixel point in the dynamic MR image.

Further, the computer equipment may use any frame of the dynamic MR image as a reference image, and perform image registration on each frame of the dynamic PET image to obtain the registered dynamic PET image. The above-mentioned dynamic-like PET image can reflect information such as the boundary and shape of the tissue/organ in the imaging part of the object to be tested, that is, the dynamic-like PET image carries the structural information of the tissue/organ, and the dynamic-like PET image and the Compared with other medical images, it can reflect the obvious feature points in the image.

S400. Based on the registered quasi-dynamic PET images, register multiple frames of initial dynamic PET images to obtain registered dynamic PET images. The multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

Further, the computer equipment can register multiple frames of initial dynamic PET images through the registered dynamic PET images carrying tissue/organ structure information, so as to improve the accuracy of dynamic PET image registration results. The computer equipment can use the registered dynamic PET image as a reference image, and use a registration algorithm to register the initial dynamic PET image of each frame to obtain a registered dynamic PET image. Optionally, the computer device may also perform arithmetic operations on the registered quasi-dynamic PET images and the initial dynamic PET images of each frame, so as to realize the registration of the initial dynamic PET images of each frame, and obtain the registered dynamic PET images. .

In this embodiment, the object to be measured has almost the same posture within the first preset time period. In the actual processing process, the size of the dynamic PET image, the dynamic PET image and the dynamic MR image can be the same.

In the above image registration method, the computer equipment can obtain the respiration data and the initial dynamic PET data in the first preset time period after the tracer is administered to the object to be tested, and determine the corresponding multiple data in the first preset time period according to the respiration data. Frame-like dynamic MR images, obtain the dynamic-like PET images corresponding to each frame-like dynamic MR image, and determine the registered dynamic-like PET images through each frame-like dynamic PET image. Frame initial dynamic PET images are registered to obtain registered dynamic PET images; the above method can directly obtain dynamic MR images through respiratory data, and then obtain tissues/organs with imaging sites based on dynamic MR images and initial dynamic PET data. Similar dynamic PET images with structural features, and based on the registered dynamic PET images, multiple frames of initial dynamic PET images are registered, so that the registered dynamic PET images can reflect the structural information and obvious image features, so as to be able to Improve the accuracy of dynamic PET image registration results; at the same time, the above method can reduce the complexity of obtaining dynamic MR images, and also allow medical staff to accurately obtain the pathological mechanism of the object to be tested from the registered dynamic PET images, The accuracy of diagnosis and treatment methods can be further improved, and effective treatment methods can be taken in time to treat the object to be tested.

As one of the embodiments, as shown in FIG. 3 , in the above step S400 , based on the registered dynamic PET images, multiple frames of initial dynamic PET images are registered, and the registered dynamic PET images can be obtained through the following steps Steps to achieve:

S410. Based on the dynamic PET images of each frame type, perform time mapping on the initial dynamic PET data in the first preset time period to obtain the mapped initial dynamic PET data of the dynamic PET images of each frame type.

Specifically, each frame-like dynamic PET image in the first preset time period corresponds to part of the initial dynamic PET data. Therefore, the computer device can correspond to each frame of initial dynamic PET data according to the sub-time period corresponding to each frame-like dynamic PET image and each frame. In the sub-time period, time mapping is performed on each frame-like dynamic PET image and each frame's initial dynamic PET data, and the mapped initial dynamic PET data of each frame-like dynamic PET image is obtained.

S420: Reconstruct the mapped initial dynamic PET data of each frame-like dynamic PET image to obtain a corresponding mapped initial dynamic PET image.

Specifically, the computer equipment can use the direct back-projection method, the iterative method or the two-dimensional Fourier transform reconstruction method to reconstruct the mapped initial dynamic PET data of each frame type of dynamic PET image, and obtain the mapped initial dynamic PET image corresponding to each frame type of dynamic PET image. PET image.

Exemplarily, if there are two frames of quasi-dynamic PET images and four frames of initial dynamic PET images within the first preset time period, the two frames of quasi-dynamic PET images are respectively quasi-dynamic PET image 1 and quasi-dynamic PET image 2, and the four frames of initial PET images are respectively quasi-dynamic PET image 1 and quasi-dynamic PET image 2. The dynamic PET images are respectively the initial dynamic PET image 1, the initial dynamic PET image 2, the initial dynamic PET image 3, and the initial dynamic PET image 4, wherein the sub-time period corresponding to the dynamic PET image 1 is [0, 2], and the dynamic PET image 1 corresponds to The sub-period corresponding to PET image 2 is [2, 4], the sub-period corresponding to the initial dynamic PET image 1 is [0, 1], and the sub-period corresponding to the initial dynamic PET image 2 is [1, 2]. The sub-time period corresponding to the dynamic PET image 3 is [2, 3], and the sub-time period corresponding to the initial dynamic PET image 4 is [3, 4] (the data in the interval represent time points), then the computer equipment can The sub-period [0, 1] corresponding to the PET image 1 and the sub-period [1, 2] corresponding to the initial dynamic PET image 2 are mapped to the sub-period [0, 2] corresponding to the quasi-dynamic PET image 1, and the The sub-period [2, 3] corresponding to the initial dynamic PET image 3 and the sub-period [3, 4] corresponding to the initial dynamic PET image 4 are mapped to the sub-period [2, 4] corresponding to the quasi-dynamic PET image 2, The dynamic PET image-like 1 is mapped to the initial dynamic PET image 1 and the mapped initial dynamic PET image 2, and the dynamic PET image-like 2 is mapped to the initial dynamic PET image 3 and the mapped initial dynamic PET image 4, respectively.

Or, if the sub-time period corresponding to the dynamic-like PET image 1 in the first preset time period is [0, 4], the sub-time period corresponding to the dynamic-like PET image 2 is [5, 9], and the initial dynamic PET image 1 The corresponding sub-period is [0, 1.5], the sub-period corresponding to the initial dynamic PET image 2 is [2.5, 4], the sub-period corresponding to the initial dynamic PET image 3 is [5, 6.5], and the initial dynamic PET image corresponds to the sub-period [5, 6.5]. 4 The corresponding sub-time period is [7.5, 9] (the data in the interval all represent time points), then the computer equipment can correspond the sub-time period [0, 1.5] corresponding to the initial dynamic PET image 1 to the initial dynamic PET image 2. The sub-time period [2.5, 4] of is mapped to the sub-time period [0, 4] corresponding to the dynamic PET image 1, and the sub-time period [5, 6.5] corresponding to the initial dynamic PET image 3 and the initial dynamic PET image 4 The corresponding sub-time period [7.5, 9] is mapped to the sub-time period [5, 9] corresponding to the dynamic PET image 2, and the mapped initial dynamic PET image 1 and the mapped initial dynamic PET image 2 of the dynamic PET image 1 are respectively mapped. , the like dynamic PET image 2 is mapped to the initial dynamic PET image 3 and the mapped initial dynamic PET image 4.

In this embodiment, the durations of the sub-periods corresponding to the initial dynamic PET images of each frame may be equal; the duration of the sub-periods corresponding to the dynamic PET images of each frame type is greater than the duration of the sub-periods corresponding to the corresponding initial dynamic PET images of each frame duration.

S430 , performing registration on the corresponding mapped initial dynamic PET image through the registered quasi-dynamic PET image, to obtain a registered dynamic PET image.

Specifically, each frame-like dynamic PET image has a corresponding registered dynamic-like PET image. The computer equipment can use the registered dynamic-like PET images of each frame as a reference image, and use an image registration algorithm to register the mapped initial dynamic PET images corresponding to the registered dynamic-like PET images of each frame, and obtain the registered dynamic PET image. Post-calibration dynamic PET images.

The above-mentioned image registration method can register the corresponding initial dynamic PET image of the mapping based on the registered quasi-dynamic PET image, so that the registered dynamic PET image can reflect the structural information and obvious image features, so that the dynamic PET image can be improved. The accuracy of the PET image registration results; at the same time, the above method can also allow medical staff to accurately obtain the pathological mechanism of the object to be tested from the registered dynamic PET images, further improve the accuracy of the diagnosis and treatment methods, and can take effective measures in time. The treatment method treats the subject to be tested.

As one of the embodiments, as shown in FIG. 4 , the step of determining the registered quasi-dynamic PET image by using each frame of quasi-dynamic PET images in the above S300 can be realized by the following steps:

S310: Acquire a deformation field corresponding to each frame type dynamic MR image.

Specifically, the computer equipment may perform arithmetic operation, conversion, analysis, and/or comparison processing on each frame type dynamic MR image, and obtain the deformation field corresponding to each frame type dynamic MR image. In this embodiment, the image may be represented in the form of a matrix, and the size of the matrix may be the same as the size of the image. If the size of the image is 3×3, the size of the matrix is also 3×3, and the data in the first row and the first column of the matrix can be the pixel value of the pixel in the first row and the first column in the image and the first The position of the pixel in the first row and column (1, 1), the data in the first row and the second column in the matrix can be the pixel value of the pixel in the first row and the second column in the image and the pixel value in the first row and the second column in the image. The position of the pixel point (1, 2) also has a corresponding relationship between the pixel values of other pixel points in the image and the data at the corresponding position in the matrix.

It should be noted that the deformation field may be represented by a deformation field matrix, and the size of the deformation field matrix may be equal to the size of the pixel matrix of the dynamic MR image. Optionally, the numerical values at different positions in the deformation field matrix may represent the deformation values of the corresponding pixel points in the dynamic MR image.

Wherein, as shown in FIG. 5 , the step of obtaining the deformation field corresponding to each frame type dynamic MR image in the above S310 may include:

S311. Determine a reference image based on the multi-frame sample MR images.

Specifically, within a period of time before the medical staff administers a certain amount of tracer to one or more imaging parts of the object to be measured, the medical scanning device can scan the imaging parts of the object to be measured, obtain sample MR data, and then The sample MR data is sent to a computer device. The computer equipment can reconstruct the sample MR data in the period of time to obtain multiple frames of sample MR images. Each frame of sample MR image has a corresponding sub-time period, and the combined time period of the sub-time periods corresponding to each frame of sample MR image is equal to the time period of the sample MR data collected by the medical scanning device.

Wherein, the computer device may select any frame of MR images from the multiple frames of sample MR images as a reference image. However, in this embodiment, the computer device may select a frame of sample MR images corresponding to the sub-time period closest to the start time point of the first preset time period as the reference image. The corresponding sample MR data may carry the relaxation properties of the tissue/organ, that is, the sample MR data may include T1WI sequences and/or T2WI sequences, and the like. The T1WI sequence represents the T1 sequence of NMR, and the T2WI sequence represents the T2 sequence of NMR.

S312 , performing image registration on the dynamic MR images of each frame type respectively by using the reference image, to obtain a deformation field corresponding to the dynamic MR image of each frame type.

It can be understood that the computer equipment can use an image registration algorithm to perform image registration on each frame type of dynamic MR image through the selected reference image, and obtain the deformation field corresponding to each frame type of dynamic MR image. The deformation field matrix corresponding to the deformation field may be equal to the difference between the pixel value of each pixel point in the reference image and each frame-type dynamic MR image.

In this embodiment, the deformation fields corresponding to the dynamic MR images of each frame type can be obtained, and then the deformation fields corresponding to the dynamic MR images of each frame type can be used to facilitate the registration of the dynamic PET images of each frame type in the corresponding time period. Image registration can take into account the structure density of dynamic MR images, so that the accuracy of image registration results can be improved.

S320. According to the deformation fields corresponding to the dynamic MR images of each frame, register the corresponding dynamic PET images to obtain the registered dynamic PET images.

At the same time, the computer equipment can use an image registration algorithm to register the corresponding dynamic PET images through the deformation fields corresponding to the dynamic MR images of each frame, so as to obtain the registered dynamic PET images.

In this embodiment, the above-mentioned image registration algorithm may be an image grayscale-based matching algorithm or an image feature-based matching algorithm, or may be other image matching algorithms, which are not limited in this embodiment. Wherein, the above-mentioned matching algorithms based on image grayscale may be mean absolute difference algorithm, absolute error sum algorithm, error sum of squares algorithm, mean error sum of squares algorithm, normalized product correlation algorithm, sequential similarity algorithm, etc.; The matching algorithm of image features can be feature extraction, feature matching, model parameter estimation, image transformation and grayscale interpolation algorithms, and so on. Fig. 6 is a multi-frame dynamic MR image corresponding to a certain imaging part of an object to be tested, a registered dynamic MR image, a dynamic PET image, a registered dynamic PET image, a registered dynamic PET image, Schematic diagram of the dynamic PET image and the registered dynamic PET image.

In the above-mentioned image registration method, the computer equipment can acquire the respiration data and multiple frames of initial dynamic PET images within the first preset time period after the tracer is administered to the object to be tested, and determine the corresponding parameters within the first preset time period according to the respiration data. The multi-frame quasi-dynamic MR images are obtained, the quasi-dynamic PET images corresponding to each frame of quasi-dynamic MR images are obtained, and the registered quasi-dynamic PET images are obtained through each frame of quasi-dynamic PET images. Based on the registered quasi-dynamic PET images, Multi-frame initial dynamic PET images are registered to obtain registered dynamic PET images; the above method can directly obtain dynamic-like MR images through respiratory data, and then obtain tissues with imaging sites based on dynamic-like MR images and initial dynamic PET data. Organ-like dynamic PET images, and based on the registered dynamic PET images, multiple frames of initial dynamic PET images are registered, so that the registered dynamic PET images can reflect structural information and obvious image features, Therefore, the accuracy of the dynamic PET image registration result can be improved; at the same time, the above method can also allow medical staff to accurately obtain the pathological mechanism of the object to be measured from the registered dynamic PET image, further improve the accuracy of the diagnosis and treatment method, and can Take effective treatment methods in a timely manner to treat the subjects to be tested.

As one of the embodiments, the step of determining the corresponding multi-frame dynamic MR images in the first preset time period according to the respiratory data in the above S200 may include: inputting the respiratory data into the prediction model to obtain the first preset time period corresponding multi-frame dynamic MR images.

Specifically, the computer device may input the respiratory data of the first preset time period into the prediction model, and the prediction model outputs the corresponding multi-frame dynamic MR images in the first preset time period. The prediction model may be a pre-trained network model, and the prediction model has been trained before the steps in S200 are performed. The above prediction model may be composed of at least one of a convolutional neural network model, a recurrent neural network model, and an adversarial neural network model.

Wherein, before performing the above-mentioned step of inputting the respiratory data into the prediction model to obtain the corresponding multi-frame dynamic MR images in the first preset time period, the above-mentioned image registration method may further include: acquiring the object to be measured and applying the tracer multi-frame sample MR images and sample respiration data in the second preset time period before the dose; the initial prediction model is trained by using the multi-frame sample MR images and sample respiration data to obtain a prediction model.

It should be noted that the computer equipment can perform network model training on the initial prediction model by using multiple frames of sample MR images and sample respiration data within the second preset time period before the tracer is administered to obtain a pre-trained prediction model. That is, the multi-frame sample MR images and sample respiration data in the second preset time period are input into the initial prediction model, and cyclic iterative training is performed to obtain the optimal prediction model. Optionally, the duration of the second preset time period may be greater than, less than or equal to the duration of the first preset time period.

It can be understood that the computer equipment can input the multi-frame sample MR images and sample respiration data in the second preset time period into the initial prediction model, obtain the dynamic-like MR prediction result, and calculate the dynamic-like MR prediction result through the loss function. The prediction error value between the standard dynamic MR image and the standard dynamic MR image, and update the initial network parameters in the initial prediction model according to the prediction error value, and iterate the above training steps continuously until the prediction error value meets the preset error threshold or the number of iterations reaches the preset iteration. Up to the threshold of times, a pre-trained prediction model is obtained. During the second time period before the tracer is applied to the imaging site of the object to be tested, the medical scanning device may scan the imaging site of the object to be tested to obtain sample MR data. The computer device may perform segmentation processing on the second time period to obtain multiple sub-time periods, and then reconstruct the sample MR data in each time period to obtain multi-frame sample MR images corresponding to the multiple sub-time periods. The durations of the sub-periods corresponding to the sample MR images of each frame may or may not be equal. The time period corresponding to the combination of the sub-time periods corresponding to the sample MR images of each frame may be equal to the second preset time period. Optionally, the above-mentioned standard dynamic-like MR image may be an idealized dynamic-like MR image.

The above-mentioned image registration method can obtain a dynamic MR image with structural information by using the respiratory data collected synchronously with the dynamic PET data, thereby shortening the data acquisition time. It is not necessary to collect dynamic MR data first, then reconstruct the dynamic MR data into dynamic MR images, and then indirectly process the dynamic MR images to obtain the process of quasi-dynamic MR images, thereby reducing the data processing process of the entire image registration and shortening the image registration time.

As one of the embodiments, as shown in FIG. 7 , the step of obtaining the dynamic-like PET images corresponding to the dynamic-like MR images of each frame in the above S300 may include:

S330: Acquire initial PET data within a first preset time period.

Specifically, after the tracer is applied to the imaging part of the object to be tested, the medical scanning device can scan the imaging part to which the tracer is applied in real time for a period of time to collect initial PET data within the first preset time period, and The acquired initial PET data is sent to a computer device. Alternatively, the computer device may also acquire the initial PET data collected in the historical time period, that is, the initial PET data in the first preset time period, from the cloud or local data. In this embodiment, the manner of acquiring the initial PET data within the first preset time period may not be limited. FIG. 8 shows the first preset time period, the second preset time period, the time point of administering the tracer, and the corresponding relationship diagram of initial PET data, sample MR data and respiratory data on the same time axis. The training sequence in FIG. 8 may be the sample MR data of multiple sub-time periods collected when training the prediction model; the sequence 1, the sequence 2 and the sequence 3 may be the sample MR data of the three sub-time periods corresponding to the administration of the tracer, However, the sample MR data of the three sub-periods corresponding to the administration of the tracer may not be used in the image registration process.

S340. Perform time mapping on the dynamic MR images of each frame type and the initial PET data in the first preset time period to obtain the mapped PET data of the dynamic MR images of each frame type.

Specifically, the time mapping of the dynamic MR images of each frame type and the initial PET data within the first preset time period can be understood as: mapping the sub-time periods of the dynamic MR images of each frame type to the corresponding sub-periods within the first preset time period Time mapping is performed in the time period, and the mapped PET data of each frame type dynamic MR image is obtained.

S350. Reconstruct the mapped PET data to obtain a dynamic-like PET image corresponding to each frame of the dynamic-like MR image.

Further, the computer equipment can reconstruct the mapped PET data in each sub-time period to obtain the dynamic-like PET images corresponding to the dynamic-like MR images of each frame.

The above image registration method can obtain the dynamic-like PET images, further register the dynamic-like PET images, and then register multiple initial dynamic PET images based on the registered dynamic-like PET images, so that the registered dynamic PET images are registered. PET images can reflect structural information and obvious image features, which can improve the accuracy of dynamic PET image registration results.

In order to facilitate the understanding of those skilled in the art, the image registration method provided by the present application is introduced by taking the execution subject as a computer device as an example. Specifically, the method includes:

(1) Acquiring respiration data and initial dynamic PET data within a first preset time period after the subject to be tested is administered the tracer.

(2) Acquiring multiple frames of sample MR images and sample respiration data within a second preset time period before the tracer is administered to the object to be tested.

(3) The initial prediction model is trained by using multiple frames of sample MR images and sample respiration data to obtain the prediction model.

(4) Input the respiratory data into the prediction model to obtain the corresponding multi-frame dynamic MR images within the first preset time period.

(5) Time mapping is performed between the dynamic MR images of each frame type and the initial PET data in the first preset time period to obtain the mapped PET data of the dynamic MR images of each frame type.

(6) Reconstructing the mapped PET data to obtain the dynamic-like PET images corresponding to the dynamic-like MR images of each frame.

(7) Determine the reference image based on the multi-frame sample MR images.

(8) Perform image registration on each frame type dynamic MR image through the reference image, and obtain the deformation field corresponding to each frame type dynamic MR image.

(9) According to the deformation field corresponding to each frame-like dynamic MR image, the corresponding dynamic-like PET image is registered to obtain the registered dynamic-like PET image.

(10) Based on the dynamic PET images of each frame type, time mapping is performed on the initial dynamic PET data of each frame within the first preset time period to obtain the mapped initial dynamic PET data of the dynamic PET image of each frame type.

(11) Reconstructing the mapped initial dynamic PET data of each frame-like dynamic PET image to obtain the corresponding mapped initial dynamic PET image.

(12) Registering the corresponding mapped initial dynamic PET image through the registered dynamic PET image to obtain the registered dynamic PET image.

For details of the execution processes of the above (1) to (12), reference may be made to the descriptions of the foregoing embodiments, and the implementation principles and technical effects thereof are similar, and are not repeated here.

It should be understood that although the various steps in the flowcharts of FIGS. 2-5 and 7 are shown in sequence as indicated by arrows, these steps are not necessarily performed sequentially in the sequence indicated by arrows. Unless explicitly stated herein, the execution of these steps is not strictly limited to the order, and these steps may be performed in other orders. Moreover, at least a part of the steps in FIGS. 2-5 and 7 may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but may be executed at different times. These steps or stages The order of execution of the steps is not necessarily sequential, but may be performed alternately or alternately with other steps or at least a portion of the steps or stages in the other steps.

In one embodiment, as shown in FIG. 9 , an image registration device is provided, including: a data acquisition module 11, a first image determination module 12, a second image determination module 13, and a registration module 14, wherein:

The data acquisition module 11 is used to acquire respiratory data and initial dynamic PET data within the first preset time period after the tracer is administered to the subject to be tested;

The first image determination module 12 is configured to determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

The second image determination module 13 is configured to obtain the dynamic-like PET images corresponding to the dynamic MR images of each frame, and determine the registered dynamic-like PET images through the dynamic-like PET images of each frame;

The registration module 14 is configured to register multiple frames of initial dynamic PET images according to the registered quasi-dynamic PET images to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are used for reconstructing the initial dynamic PET data. obtained image.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one of the embodiments, the registration module 14 includes: a first time mapping unit, a reconstruction unit and a first registration unit, wherein:

a first time mapping unit, configured to perform time mapping on the initial dynamic PET data in the first preset time period according to the dynamic PET images of each frame type, to obtain the mapped initial dynamic PET data of the dynamic PET images of each frame type;

The reconstruction unit is used for reconstructing the mapped initial dynamic PET data of each frame-like dynamic PET image to obtain the corresponding mapped initial dynamic PET image;

The first registration unit is used for registering the corresponding mapped initial dynamic PET image through the registered quasi-dynamic PET image to obtain the registered dynamic PET image.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one of the embodiments, the second image determination module 13 includes: a deformation field acquisition unit and a second registration unit, wherein:

A deformation field acquisition unit, used to acquire the deformation field corresponding to each frame type of dynamic MR image;

The second registration unit is configured to register the corresponding dynamic-like PET images according to the deformation fields corresponding to the dynamic-like MR images of each frame, so as to obtain the registered dynamic-like PET images.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one of the embodiments, the deformation field acquisition unit includes: a reference image determination subunit and an image registration subunit, wherein:

a reference image determination subunit, configured to determine a reference image according to the multi-frame sample MR images;

The image registration subunit is used to perform image registration on the dynamic MR images of each frame type respectively by using the reference image, so as to obtain the deformation field corresponding to the dynamic MR image of each frame type.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one embodiment, the first image determination module 12 is specifically configured to input respiratory data into the prediction model to obtain corresponding multi-frame dynamic MR images within the first preset time period.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one embodiment, the above-mentioned image registration apparatus further includes: a data acquisition module and a training module, wherein:

a data acquisition module, configured to acquire multiple frames of sample MR images and sample respiration data within the second preset time period before the tracer is administered to the subject to be tested;

The training module is used to train the initial prediction model by using the multi-frame sample MR images and sample breathing data to obtain the prediction model.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

In one of the embodiments, the second image determination module 13 includes: a data acquisition unit, a second time mapping unit and a data reconstruction unit, wherein:

a data acquisition unit for acquiring initial PET data within a first preset time period;

a second time mapping unit, configured to perform time mapping on the dynamic MR images of each frame type and the initial PET data in the first preset time period to obtain the mapped PET data of the dynamic MR images of each frame type;

The data reconstruction unit is used for reconstructing the mapped PET data to obtain the dynamic-like PET images corresponding to the dynamic-like MR images of each frame.

The image registration apparatus provided in this embodiment can execute the above method embodiments, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

For the specific limitation of the image registration apparatus, reference may be made to the limitation of the image registration method above, which will not be repeated here. Each module in the above-mentioned image registration apparatus may be implemented in whole or in part by software, hardware, and combinations thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided, and the computer device may be a server, and its internal structure diagram may be as shown in FIG. 10 . The computer device includes a processor, memory, and a network interface connected by a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The computer facility's database is used to store respiratory data and dynamic PET images. The network interface of the computer device is used to communicate with external endpoints over a network connection. The computer program, when executed by a processor, implements an intensification degree analysis method.

Those skilled in the art can understand that the structure shown in FIG. 10 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, including a memory and a processor, a computer program is stored in the memory, and the processor implements the following steps when executing the computer program:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

In one embodiment, a readable storage medium is provided on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

In one embodiment, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the following steps:

Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer;

Determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiratory data;

Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image;

Based on the registered quasi-dynamic PET images, multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images reconstructed from the initial dynamic PET data.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media used in the various embodiments provided in this application may include at least one of non-volatile and volatile memory. The non-volatile memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, and the like. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, the RAM may be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM).

The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, all It is considered to be the range described in this specification.

The above-mentioned embodiments only represent several embodiments of the present application, and the descriptions thereof are specific and detailed, but should not be construed as a limitation on the scope of the invention patent. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the scope of protection of the patent of the present application shall be subject to the appended claims.

Claims (10)

1. an image registration method, is characterized in that, described method comprises: Acquiring respiration data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer; determining the corresponding multi-frame dynamic MR images within the first preset time period according to the breathing data; Obtaining the quasi-dynamic PET image corresponding to each frame of quasi-dynamic MR image, and determining the registered quasi-dynamic PET image through each frame of quasi-dynamic PET image; Based on the registered quasi-dynamic PET images, the multiple frames of initial dynamic PET images are registered to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are images of the initial dynamic PET data reconstructed image. 2 . The image registration method according to claim 1 , wherein the registration is performed on the multiple frames of initial dynamic PET images based on the registered quasi-dynamic PET images to obtain the registration. 3 . Post dynamic PET images, including: Based on the dynamic PET images of each frame type, time mapping is performed on the initial dynamic PET data in the first preset time period to obtain the mapped initial dynamic PET data of the dynamic PET images of each frame type; reconstructing the mapped initial dynamic PET data of the dynamic PET images of each frame type to obtain a corresponding mapped initial dynamic PET image; The registered dynamic PET image is obtained by registering the corresponding mapped initial dynamic PET image through the registered dynamic PET image. 3. The image registration method according to claim 1 or 2, wherein, determining the registered dynamic PET image by each frame of dynamic PET image, comprising: obtaining the deformation fields corresponding to the dynamic MR images of each frame type; According to the deformation fields corresponding to the dynamic MR images of each frame, the corresponding dynamic PET images are registered to obtain the registered dynamic PET images. 4. The image registration method according to claim 3, wherein the obtaining the deformation field corresponding to the dynamic MR images of each frame type comprises: Determining a reference image based on multiple frames of sample MR images; Image registration is performed on the dynamic MR images of each frame type respectively by using the reference image, so as to obtain the deformation field corresponding to the dynamic MR image of each frame type. 5. The image registration method according to claim 1 or 2, wherein the determining according to the respiration data the corresponding multi-frame dynamic MR images in the first preset time period comprises: The respiration data is input into the prediction model to obtain corresponding multi-frame dynamic MR images within the first preset time period. 6. The image registration method according to claim 5, wherein the method further comprises: acquiring multiple frames of sample MR images and sample respiration data within a second preset time period before the tracer is administered to the subject to be tested; The initial prediction model is obtained by training the initial prediction model by using the multi-frame sample MR images and the sample respiration data. 7. The image registration method according to claim 1 or 2, wherein the obtaining of the dynamic PET images corresponding to the dynamic MR images of each frame comprises: obtaining initial PET data within the first preset time period; performing time mapping on the dynamic MR images of each frame type and the initial PET data in the first preset time period to obtain the mapped PET data of the dynamic MR images of each frame type; Reconstructing the mapped PET data to obtain dynamic-like PET images corresponding to the dynamic-like MR images of each frame. 8. An image registration device, wherein the device comprises: a data acquisition module, used to acquire respiratory data and initial dynamic PET data within the first preset time period after the subject to be tested is administered the tracer; a first image determination module, configured to determine the corresponding multi-frame dynamic MR images within the first preset time period according to the respiration data; The second image determination module is used to obtain the dynamic-like PET images corresponding to the dynamic MR images of each frame, and determine the registered dynamic-like PET images through the dynamic-like PET images of each frame; The registration module is used for registering multiple frames of initial dynamic PET images according to the registered quasi-dynamic PET images to obtain registered dynamic PET images; the multiple frames of initial dynamic PET images are for the Image reconstructed from initial dynamic PET data. 9. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the method according to any one of claims 1-7 when the processor executes the computer program . 10. A readable storage medium on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the method according to any one of claims 1-7 are implemented.

Images