Multi-modal image registration method and system based on deformation field, storage medium and equipment

Abstract

The invention provides a multi-modal image registration method and system based on a deformation field, a storage medium and equipment, and the method comprises the steps: obtaining multi-modal images, including a 2D ultrasonic image and a 3D reconstruction image, and carrying out the coarse registration; obtaining a 2D ultrasonic image sequence of continuous frames, and obtaining rigid change parameters of the organ in a respiratory cycle; acquiring organ contour feature data points of each frame of 2D ultrasonic image, and performing elastic registration to obtain a 2D elastic deformation field of the organ; constructing a 2D-3D deformation function of the organ, and obtaining a 3D deformation field in a three-dimensional space; and updating vertex coordinates of the target organ in the 3D reconstruction image according to the 3D deformation field to obtain a mapped three-dimensional image to represent rigid and elastic changes of the organ. According to the method, by fusing the multi-modal image information and combining rigid and elastic deformation analysis, accurate registration of the ultrasonic image and the CT / MRI image is achieved, and the precision of multi-modal medical image registration is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and specifically designs a multimodal image fusion method, system, storage medium and device based on deformation field. Background Art

[0002] Medical image registration and fusion is a commonly used technique in medical image analysis. It converts the coordinates of one set of image sequences into another set of image sequences so that the corresponding positions of the two sets of image sequences match, resulting in a registered and fused image sequence.

[0003] The core technical challenge facing multimodal medical image registration is how to accurately align organ images acquired by different imaging methods. In particular, there are significant differences between 2D ultrasound images and 3D reconstructed CT / MRI images. How to achieve accurate correspondence between these two modal images during dynamic breathing becomes a complex problem.

[0004] Traditional registration methods find it difficult to take into account both rigid changes and elastic deformations at the same time, resulting in insufficient registration accuracy, which in turn affects the accuracy of medical diagnosis and surgical planning. Summary of the Invention

[0005] The purpose of the present invention is to address the registration problem of 2D ultrasound images and 3D reconstructed images under respiratory motion, and to propose a multimodal image registration method, system, storage medium and device based on deformation field.

[0006] The technical solution of the present invention is: The present invention provides a multimodal image registration method based on deformation field, which comprises the following steps: S1. A coarse registration step: acquiring a multimodal image, wherein the multimodal image includes a 2D ultrasound image and a 3D reconstructed image, determining a spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on a coordinate system transformation, and obtaining a coarse registration result; S2, rigid registration step: acquiring a 2D ultrasound image sequence of consecutive frames, extracting organ contours and anatomical features for rigid registration, and obtaining rigidity change parameters of the organ during the respiratory cycle; S3, elastic registration step: obtaining organ contour feature data points of each frame of 2D ultrasound image, and performing elastic registration to obtain a 2D elastic deformation field of the organ; S4, deformation mapping step: constructing a 2D-3D deformation function of the organ to obtain a 3D deformation field in three-dimensional space; updating the vertex coordinates of the target organ in the 3D reconstructed image according to the 3D deformation field to obtain a mapped three-dimensional image to characterize the rigidity and elasticity changes of the organ.

[0007] Furthermore, the 3D reconstructed image is CT or MRI image data; determining the spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on coordinate system transformation to obtain a rough registration result specifically includes: Acquiring coordinate data of an electromagnetic sensor attached to an ultrasound probe; Establishing a conversion relationship between an ultrasound coordinate system and a 3D reconstructed image coordinate system according to the coordinate data; The coordinates of the 2D ultrasound image are mapped to the coordinate system of the 3D reconstructed image through the conversion relationship, and the position of the 2D ultrasound image in the 3D reconstructed image is obtained as a coarse registration result for position tracking.

[0008] Furthermore, the rigid registration step specifically includes: Acquire a 2D ultrasound image sequence of consecutive frames; Processing the 2D ultrasound image sequence using a segmentation algorithm to extract organ contours; extracting features from the organ contours based on a feature extraction network to obtain organ anatomical features; Based on the spatial relationship between the electromagnetic sensor bound to the ultrasound probe and the ultrasound probe, the three-dimensional coordinates of the organ contour and anatomical features are obtained; since the organ position will return to its original state during the respiratory cycle, the translation range and rotation range of the organ during the respiratory cycle are obtained as the rigidity change parameters of the organ during the respiratory cycle.

[0009] Furthermore, the elastic registration step specifically includes: A window of preset width is used to slide on each frame of 2D ultrasound image to obtain organ contour feature data points of each frame of 2D ultrasound image, wherein the organ contour feature data covers the complete organ contour; Elastic registration is performed on the organ contour feature data points at corresponding positions in the 2D ultrasound images of consecutive frames to obtain the 2D elastic deformation field of the organ, which is used to describe the elastic changes of the organ in the 2D ultrasound image.

[0010] Furthermore, the elastic registration uses an interpolation method to construct a 2D elastic deformation field of continuous organ images. ; ; in, is the interpolation function; Indicates the position of any point on the 2D ultrasound image; Indicates the 2D ultrasound image frame number, Indicates the total number of frames; and Respectively represent the original position and deformation of the corresponding feature data points of the previous and next continuous image frames; Represents any point between 2D ultrasound image sequences and The Euclidean distance between them.

[0011] Furthermore, the 2D-3D deformation function of the organ constructed in S4 is specifically: Construct an optimization equation to obtain the optimal 3D deformation field:

[0012] in: x, y, z Represents the position of any point on the 3D reconstructed image, Δ x、 Δ y、 Δ z They represent the continuous changes of points on the organ surface in the 3D reconstructed three-dimensional space; Shift and Rotate Representative 2D ultrasound images The range of translation and rotation of organs during the respiratory cycle, are the coefficients of the 2D elastic deformation field, is the organ tissue density, is the force parameter, which serves as the condition constraint of the optimization algorithm function; 2D-3D deformation mapping for MRI or CT volume data change, V It is 3D reconstructed volume data of MRI or CT; min L () represents the projection section image of the corresponding position of the ultrasound probe on the 3D reconstructed image in real time after obtaining the 2D-3D deformation mapping Upper organ contours and 2D ultrasound images The differences in organ contours are optimized iteratively until the differences are minimized.

[0013] Furthermore, the loss function of the 2D-3D deformation function is:

[0014] in: Representing an image and The outline of the organ The overlap, Penalize the gradient of the deformation field to maintain smoothness; defined as: ; deformation field The optimization is done by gradient descent algorithm.

[0015] A system used in the deformation field-based multimodal image registration method according to any one of claims 1 to 7, comprising: A coarse registration module is used to obtain a multimodal image, wherein the multimodal image includes a 2D ultrasound image and a 3D reconstructed image, and determine the spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on a coordinate system transformation to obtain a coarse registration result; Rigid registration module: used to acquire a 2D ultrasound image sequence of continuous frames, extract organ contours and anatomical features for rigid registration, and obtain the rigidity change parameters of the organ during the respiratory cycle; Elastic registration module: used to obtain the organ contour feature data points of each frame of 2D ultrasound image, and perform elastic registration to obtain the 2D elastic deformation field of the organ; Deformation mapping module: used to construct the 2D-3D deformation function of the organ and obtain the 3D deformation field in three-dimensional space; based on the 3D deformation field, the vertex coordinates of the target organ in the 3D reconstructed image are updated to obtain a mapped three-dimensional image, which characterizes the rigidity and elasticity changes of the organ.

[0016] A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, the deformation field-based multimodal image registration method according to any one of claims 1 to 7 is implemented.

[0017] An ultrasound device comprising: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, the deformation field-based multimodal image registration method according to any one of claims 1 to 7 is implemented.

[0018] Beneficial effects of the present invention: The present invention discloses a multimodal medical image registration method. The method achieves coarse registration by acquiring ultrasound and CT / MRI images, determining the spatial coordinates of the ultrasound probe, and constructing a spatial transformation relationship between images. The 2D ultrasound image is then segmented to extract organ contours and features, and rigidity and elasticity changes are calculated. Combined with force analysis and tissue density data, a 2D to 3D deformation mapping function is generated. This function is used to update the vertex positions and voxel deformation coefficients of the organs in the 3D model. Finally, a visualization result of the dynamic deformation of the organs is generated through rendering, thereby achieving accurate characterization of the rigidity and elasticity changes of the organs.

[0019] The present invention achieves precise registration of ultrasound and CT / MRI images by fusing multimodal imaging information and combining rigid and elastic deformation analysis, effectively improving the accuracy of multimodal medical image registration, providing more accurate organ deformation information for medical diagnosis and surgical planning, and providing reliable technical support for real-time surgical navigation and radiotherapy, thereby improving diagnosis and treatment effects.

[0020] Other features and advantages of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other objects, features and advantages of the present invention will become more apparent through a more detailed description of exemplary embodiments of the present invention with reference to the accompanying drawings, wherein like reference numerals generally represent like components throughout the exemplary embodiments of the present invention.

[0022] Figure 1 A registration flow chart according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0023] The preferred embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the preferred embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0024] Example 1: Figure 1 A registration flow chart according to an embodiment of the present invention is shown.

[0025] like Figure 1 As shown, the present invention provides a multimodal image registration method based on deformation field, which includes the following steps: S1. Coarse registration step: obtaining coordinate data of an electromagnetic sensor attached to an ultrasound probe; establishing a conversion relationship between an ultrasound coordinate system and a 3D reconstructed image coordinate system based on the coordinate data; mapping the coordinates of the 2D ultrasound image to the coordinate system of the 3D reconstructed image using the conversion relationship, and obtaining the position of the 2D ultrasound image in the 3D reconstructed image as a coarse registration result for position tracking; S2, rigid registration step: obtaining a 2D ultrasound image sequence of consecutive frames; processing the 2D ultrasound image sequence using a segmentation algorithm, such as a level set or convolutional neural network (CNN), to extract organ contours; performing feature extraction on the organ contours based on a feature extraction network, such as a SuperPoint network, to obtain organ anatomical features; Based on the spatial relationship between the electromagnetic sensor bound to the ultrasound probe and the ultrasound probe, the three-dimensional coordinates of the organ contour and anatomical features are obtained; since the organ position will return to its original state during the respiratory cycle, the translation range and rotation range of the organ during the respiratory cycle are obtained as the rigidity change parameters of the organ during the respiratory cycle.

[0026] Specifically, in this step, 1. A 2D ultrasound image sequence of consecutive frames is acquired, ensuring coverage of the complete respiratory cycle. The image sequence is timestamped when transferred to the buffer to facilitate subsequent alignment with the electromagnetic sensor data. For example, in a liver examination, images covering two complete respiratory cycles are collected, and the key frames at the end of inspiration and expiration are filtered using the timestamps. 2. When processing the image sequence using the level set segmentation algorithm, the contour lines are initialized based on the liver's grayscale distribution. Alternatively, a CNN (such as U-Net) is used, which requires a pre-trained model. 3. SuperPoint network training: The SuperPoint network is trained based on basic data of standard shapes such as triangles and rectangles. To improve DICE for ultrasound image processing, local features in the ultrasound image can be selected as input, such as the portal vein of the liver, the renal pelvis of the kidney, vascular bifurcations, and organ edges. 4. The probe pose data output by the electromagnetic sensor is aligned with the image timestamps, and the 3D coordinates of the contour points are obtained through coordinate system transformation. 5. When calculating the rigid parameters within the respiratory cycle, several periodic data are selected to calculate the translation trajectory of the organ's center of mass, extract the translation range and rotation range, and model the periodic changes as rigid motion.

[0027] S3, elastic registration step: obtaining organ contour feature data points of each frame of 2D ultrasound image, and performing elastic registration to obtain a 2D elastic deformation field of the organ; S4, deformation mapping step: constructing a 2D-3D deformation function of the organ to obtain a 3D deformation field in three-dimensional space; updating the vertex coordinates of the target organ in the 3D reconstructed image according to the 3D deformation field to obtain a mapped three-dimensional image to characterize the rigidity and elasticity changes of the organ.

[0028] Specifically, the elastic registration step in S3 includes: A window of preset width is used to slide on each frame of 2D ultrasound image to obtain organ contour feature data points of each frame of 2D ultrasound image, wherein the organ contour feature data covers the complete organ contour; Elastic registration is performed on the organ contour feature data points at corresponding positions in the 2D ultrasound images of the previous and next consecutive frames to obtain the 2D elastic deformation field of the organ, which is used to describe the elastic changes of the organ in the 2D ultrasound image; the elastic registration uses the interpolation method to construct the 2D elastic deformation field of the continuous image of the organ ; ; in, is the interpolation function; Indicates the position of any point on the 2D ultrasound image; Indicates the 2D ultrasound image frame number, Indicates the total number of frames; and Respectively represent the original position and deformation of the corresponding feature data points of the previous and next continuous image frames; Represents any point between 2D ultrasound image sequences and The Euclidean distance between them.

[0029] In this example, when a window of preset width is slid across each frame of a 2D ultrasound image, the window size is set to cover the entire organ. For example, for a liver ultrasound image, the window sliding can capture the feature point set at the bifurcation of the hepatic vein, and its coordinate sequence forms a discrete representation of the contour.

[0030] Elastic registration is performed on feature points in consecutive frames to establish point correspondences between frames. For example, the hepatic vein bifurcation point p (x = 120, y = 80) in frame i is associated with the corresponding point p' (x = 122, y = 83) in frame i+1 using optical flow or feature descriptor matching. The deformation variable Δp = (+2, +3) reflects the displacement of the point due to respiratory motion. Radial basis functions, such as thin plate splines, are used as interpolation functions. Euclidean distance is used to constrain the locality of the deformation field and adjust the influence of each corner point on the overall deformation field.

[0031] The overall rigid deformation of the liver is determined by the organ's translation and rotational range during the respiratory cycle, while local distortions, or elastic deformations, are refined using interpolation functions. If feature points in five consecutive frames consistently show an upward displacement of 3 pixels in the right lobe of the liver, while only 1 pixel in the left lobe, the deformation field reveals the organ's elastic stretching properties and can be used to quantify differences in tissue stiffness. The resulting 2D elastic deformation field provides pixel-by-pixel deformation vectors for subsequent 3D mapping, supporting the real-time requirements of multimodal fusion.

[0032] Specifically, the 2D-3D deformation function of the organ constructed in S4 is: Construct an optimization equation to obtain the optimal 3D deformation field:

[0033] in: x, y, z Represents the position of any point on the 3D reconstructed image, Δ x、 Δ y、 Δ z They represent the continuous changes of points on the organ surface in the 3D reconstructed three-dimensional space; Shift and Rotate Representative 2D ultrasound images The range of translation and rotation of organs during the respiratory cycle, are the coefficients of the 2D elastic deformation field, is the organ tissue density, is the force parameter, which serves as the condition constraint of the optimization algorithm function; 2D-3D deformation mapping for MRI or CT volume data change, V It is 3D reconstructed volume data of MRI or CT; min L () represents the projection section image of the corresponding position of the ultrasound probe on the 3D reconstructed image in real time after obtaining the 2D-3D deformation mapping Upper organ contours and 2D ultrasound images The differences in organ contours are optimized iteratively until the differences are minimized.

[0034] The loss function of the 2D-3D deformation function is:

[0035] in: Representing an image and The outline of the organ The overlap, Penalize the gradient of the deformation field to maintain smoothness; defined as: ; deformation field The optimization is done by gradient descent algorithm.

[0036] In this example, an optimization equation is constructed to determine the optimal deformation field. The translation and rotation ranges (Shift and Rotate) in the 2D ultrasound image are mapped to the 3D reconstructed image. Using the elastic deformation field coefficient (DDF) and other constraints, such as organ tissue density and force parameters, an optimal model is constructed that reflects organ deformation during the respiratory cycle. The goal of optimizing the deformation field is to minimize the difference between the organ contours in the 2D ultrasound image and the 3D reconstructed image. This difference is gradually reduced through iterative optimization until the optimal deformation field is achieved.

[0037] When constructing the loss function, not only must the degree of overlap between the organ contours in the 2D ultrasound image and the 3D reconstructed image be considered, but the smoothness of the deformation field must also be maintained by penalizing the gradient of the deformation field. Using a gradient descent algorithm, the deformation field can be gradually optimized to minimize the loss function, thereby obtaining the optimal deformation field and achieving high-precision registration of the 2D ultrasound image and the 3D reconstructed image.

[0038] Example 2:

[0039] The embodiment of the present application further provides a multimodal image registration system based on a deformation field. It should be noted that the multimodal image registration system based on a deformation field in the embodiment of the present application can be used to execute the multimodal image registration method based on a deformation field provided in the embodiment of the present application, including: A coarse registration module is used to obtain a multimodal image, wherein the multimodal image includes a 2D ultrasound image and a 3D reconstructed image, and determine the spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on a coordinate system transformation to obtain a coarse registration result; Rigid registration module: used to acquire a 2D ultrasound image sequence of continuous frames, extract organ contours and anatomical features for rigid registration, and obtain the rigidity change parameters of the organ during the respiratory cycle; Elastic registration module: used to obtain the organ contour feature data points of each frame of 2D ultrasound image, and perform elastic registration to obtain the 2D elastic deformation field of the organ; Deformation mapping module: used to construct the 2D-3D deformation function of the organ and obtain the 3D deformation field in three-dimensional space; based on the 3D deformation field, the vertex coordinates of the target organ in the 3D reconstructed image are updated to obtain a mapped three-dimensional image, which characterizes the rigidity and elasticity changes of the organ.

[0040] Example 3:

[0041] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed, the multimodal image registration method based on deformation field is implemented.

[0042] Example 4: An embodiment of the present invention provides an ultrasound device, comprising: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, the deformation field-based multimodal image registration method is implemented.

[0043] While various embodiments of the present invention have been described above, the above description is intended to be illustrative, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A multimodal image registration method based on deformation field, characterized in that: The method comprises the following steps: S1. A coarse registration step: acquiring a multimodal image, wherein the multimodal image includes a 2D ultrasound image and a 3D reconstructed image, determining a spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on a coordinate system transformation, and obtaining a coarse registration result; S2, rigid registration step: acquiring a 2D ultrasound image sequence of consecutive frames, extracting organ contours and anatomical features for rigid registration, and obtaining rigidity change parameters of the organ during the respiratory cycle; S3, elastic registration step: obtaining organ contour feature data points of each frame of 2D ultrasound image, and performing elastic registration to obtain a 2D elastic deformation field of the organ; S4, deformation mapping step: constructing a 2D-3D deformation function of the organ to obtain a 3D deformation field in three-dimensional space; updating the vertex coordinates of the target organ in the 3D reconstructed image according to the 3D deformation field to obtain a mapped three-dimensional image to characterize the rigidity and elasticity changes of the organ.

2. The multimodal image registration method based on deformation field according to claim 1 is characterized in that In the coarse registration step, the 3D reconstructed image is CT or MRI image data; determining the spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on coordinate system transformation to obtain a coarse registration result specifically includes: Acquiring coordinate data of an electromagnetic sensor attached to an ultrasound probe; Establishing a conversion relationship between an ultrasound coordinate system and a 3D reconstructed image coordinate system according to the coordinate data; The coordinates of the 2D ultrasound image are mapped to the coordinate system of the 3D reconstructed image through the conversion relationship, and the position of the 2D ultrasound image in the 3D reconstructed image is obtained as a coarse registration result for position tracking.

3. The multimodal image registration method based on deformation field according to claim 1, characterized in that The rigid registration steps specifically include: Acquiring a 2D ultrasound image sequence of consecutive frames; Processing the 2D ultrasound image sequence using a segmentation algorithm to extract organ contours; extracting features from the organ contours based on a feature extraction network to obtain organ anatomical features; Based on the spatial relationship between the electromagnetic sensor bound to the ultrasound probe and the ultrasound probe, the three-dimensional coordinates of the organ contour and anatomical features are obtained; since the organ position will return to its original state during the respiratory cycle, the translation range and rotation range of the organ during the respiratory cycle are obtained as the rigidity change parameters of the organ during the respiratory cycle.

4. The multimodal image registration method based on deformation field according to claim 1, characterized in that The elastic registration steps specifically include: A window of preset width is used to slide on each frame of 2D ultrasound image to obtain organ contour feature data points of each frame of 2D ultrasound image, wherein the organ contour feature data covers the complete organ contour; Elastic registration is performed on the organ contour feature data points at corresponding positions in the 2D ultrasound images of consecutive frames to obtain the 2D elastic deformation field of the organ, which is used to describe the elastic changes of the organ in the 2D ultrasound image.

5. The multimodal image registration method based on deformation field according to claim 4 is characterized in that The elastic registration uses interpolation to construct the 2D elastic deformation field of the continuous images of the organ ; ； in, is the interpolation function; Indicates the position of any point on the 2D ultrasound image; Indicates the 2D ultrasound image frame number, Indicates the total number of frames; and Respectively represent the original position and deformation of the corresponding feature data points of the previous and next continuous image frames; Represents any point between 2D ultrasound image sequences and The Euclidean distance between .

6. The multimodal image registration method based on deformation field according to claim 1, characterized in that The 2D-3D deformation function of the organ constructed in S4 is specifically: Construct an optimization equation to obtain the optimal 3D deformation field: ； in: x, y, z Represents the position of any point on the 3D reconstructed image, Δ x、 Δ y、 Δ z They represent the continuous changes of points on the organ surface in the 3D reconstructed three-dimensional space; Shift and Rotate Representative 2D ultrasound images The range of translation and rotation of organs during the respiratory cycle, DDF are the coefficients of the 2D elastic deformation field, is the organ tissue density, is the force parameter, which serves as the condition constraint of the optimization algorithm function; 2D-3D deformation mapping for MRI or CT volume data change, V It is 3D reconstructed volume data of MRI or CT; min L () represents the projection section image of the corresponding position of the ultrasound probe on the 3D reconstructed image in real time after obtaining the 2D-3D deformation mapping Upper organ contours and 2D ultrasound images The differences in organ contours are optimized iteratively until the differences are minimized.

7. The deformation field-based multimodal image registration method according to claim 6, characterized in that The loss function of the 2D-3D deformation function is: ； in: Representing an image and The outline of the organ The overlap, Penalize the gradient of the deformation field to maintain smoothness; defined as: ; deformation field The optimization is done by gradient descent algorithm.

8. A system used in the deformation field-based multimodal image registration method according to any one of claims 1 to 7, characterized in that: include: A coarse registration module is used to obtain a multimodal image, wherein the multimodal image includes a 2D ultrasound image and a 3D reconstructed image, and determine the spatial relationship between the 2D ultrasound image and the 3D reconstructed image based on a coordinate system transformation to obtain a coarse registration result; Rigid registration module: used to acquire a 2D ultrasound image sequence of continuous frames, extract organ contours and anatomical features for rigid registration, and obtain the rigidity change parameters of the organ during the respiratory cycle; Elastic registration module: used to obtain the organ contour feature data points of each frame of 2D ultrasound image, and perform elastic registration to obtain the 2D elastic deformation field of the organ; Deformation mapping module: used to construct the 2D-3D deformation function of the organ and obtain the 3D deformation field in three-dimensional space; based on the 3D deformation field, the vertex coordinates of the target organ in the 3D reconstructed image are updated to obtain a mapped three-dimensional image, which characterizes the rigidity and elasticity changes of the organ.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed, the multimodal image registration method based on deformation field as claimed in any one of claims 1 to 7 is implemented.

10. An ultrasonic device, characterized in that: include: one or more processors; The processor is configured to store one or more programs; When the one or more programs are executed by the one or more processors, the deformation field-based multimodal image registration method according to any one of claims 1 to 7 is implemented.

Images