High-precision real-time medical image splicing method and system

Abstract

The invention provides a high-precision real-time medical image splicing method and system, and the method comprises the following steps: in a scanning process, obtaining a to-be-spliced pathological section image obtained through the scanning of a scanner at a certain position, and carrying out the optimization processing of the pathological section image, thereby obtaining the optimal position of the pathological section image, and obtaining the optimal position of the pathological section image; splicing the pathological section images according to the optimal position; the method comprises the following steps: acquiring a to-be-spliced pathological section image obtained by scanning when a scanner moves to a new position, and repeating an optimization processing process on the pathological section image, so that the pathological section image can be spliced while being scanned in the scanning process until a complete pathological section image is obtained; the method has the beneficial effects that the optimal position of each pathological section image to be spliced is calculated by using the factor graph, Bayesian reasoning and nonlinear optimization, so that the image integrity can be kept, the image can be spliced while being scanned, and the image splicing result is more accurate by correcting the splicing position of the image.

Description

Technical Field

[0001] The present invention relates to the technical field of image stitching, and in particular to a high-precision real-time medical image stitching method and system. Background Art

[0002] In digital pathology, in order to obtain a complete pathology slice image, it is usually necessary to stitch multiple high-resolution images. These images often cover different areas of the entire slice, so they must be combined into a complete image through stitching technology. However, since the resolution of pathology images is very high, usually reaching billions of pixels, the amount of calculation in the stitching process is very large and the processing speed is very slow, which seriously affects the efficiency of real-time diagnosis and analysis.

[0003] Defects and shortcomings of existing technologies: 1. Highly dependent on image matching and unable to cope with poor matching: First, the matching algorithm is sensitive to deformation. Since pathological sections may deform during production and scanning, existing spatial domain jigsaw puzzle methods are less robust to these deformations and prone to calculation errors; Second, for samples lacking feature points, it is impossible to calculate a relatively accurate displacement distance and even produce a large offset error. 2. Unable to scan and stitch at the same time: The digital imaging process of the entire slide requires the calculation of the global position of each image. This process requires waiting for all images to be scanned and the displacement deviation to be calculated before global optimization can be performed. Then the machine scanning motion process cannot be imaged, resulting in too slow imaging time. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a high-precision real-time medical image stitching method and system for solving the problem that the prior art is highly dependent on image matching and thus cannot cope with poor matching and cannot stitch while scanning.

[0005] In order to achieve the above-mentioned purpose and other related purposes, the present invention provides the following technical solutions: A high-precision real-time medical image stitching method comprises the following steps: scanning a pathological slice using a scanner, and obtaining a pathological slice image to be stitched according to the scanning result; in the scanning process, obtaining the pathological slice image to be stitched obtained by scanning the pathological slice at a certain position by the scanner, optimizing the pathological slice image, obtaining the optimal position of the pathological slice image according to the optimization result, and stitching the pathological slice image according to the optimal position of the pathological slice image; The pathological slice image to be spliced ​​is obtained by moving the scanner to a new position to scan the pathological slice, and the optimization process is repeated for the pathological slice image, so that each pathological slice image scanned by the scanner is iteratively optimized in sequence according to the scanning order, so that the pathological slice image can be spliced ​​while scanning until a complete pathological slice image is obtained, wherein the complete pathological slice image includes several spliced ​​pathological slice images, and the content of each pathological slice image contains different areas of the entire pathological slice.

[0006] A high-precision real-time medical image stitching system, comprising: an image optimization processing module, used to scan a pathological slice using a scanner, and obtain a pathological slice image to be stitched according to the scanning result; in the scanning process, the pathological slice image to be stitched obtained by scanning the pathological slice at a certain position by the scanner, and the pathological slice image is optimized, the optimal position of the pathological slice image is obtained according to the optimization result, and the pathological slice image is stitched according to the optimal position of the pathological slice image; The iterative optimization processing module is used to obtain the pathological slice image to be spliced ​​obtained by moving the scanner to a new position to scan the pathological slice, and repeat the optimization processing process on the pathological slice image, so that each pathological slice image scanned by the scanner is iteratively optimized in sequence according to the scanning order, so that it can be scanned and spliced ​​at the same time during the scanning process until a complete pathological slice image is obtained, wherein the complete pathological slice image includes several spliced ​​pathological slice images, and the content of each pathological slice image contains different areas of the entire pathological slice.

[0007] An electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the high-precision real-time medical image stitching method as described above.

[0008] In one embodiment of the present invention, after the pathological slice image to be spliced ​​is obtained by scanning the pathological slice by the scanner, and before the pathological slice image is optimized, it also includes: confirming whether the pathological slice image to be spliced ​​is a starting image; if so, the pathological slice image is not optimized and spliced ​​directly; if not, confirming whether the pathological slice image has adjacent pathological slice images; if so, the pathological slice image is optimized; if not, the image weight of the non-pathological slice area in the corresponding direction is set to 0.

[0009] In one embodiment of the present invention, the pathological slice image is optimized and the optimal position of the pathological slice image is obtained according to the optimization result, including: determining a factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice image, wherein the factor graph includes relative pose factors and prior factors, the relative pose factors include the displacement, matching strength and rotation angle between the pathological slice images, and the prior factors include the initial position of each pathological slice image to be spliced; establishing a Bayesian inference model according to the factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice image, and obtaining the optimal position of the pathological slice image through the Bayesian inference model and nonlinear optimization in turn, wherein the Bayesian inference model includes a minimum error model.

[0010] In one embodiment of the present invention, the relative pose factor also includes a shooting position of the pathological slice image, the shooting position of the pathological slice image is the actual grid position of the pathological slice image to be spliced ​​calculated by the SURF_NCC matching method, and the error between the initial position and the shooting position of the pathological slice image can be obtained through the factor graph, wherein the initial position of the pathological slice image to be spliced ​​is the preset ideal grid position of the pathological slice image; There are also observation positions between pathological slice images. and The observations between can be expressed as factors , the factor represents the pathological slice image The position of , and obtain the relative pose factor, and compare it with the pathological slice image The position is calculated, and Pathological slice image Observation location.

[0011] In one embodiment of the present invention, determining the factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice images includes: using the SURF_NCC matching method to determine the displacement, matching strength and rotation angle between adjacent pathological slice images, wherein the SURF_NCC matching method combines the SURF feature extraction algorithm and the NCC matching algorithm.

[0012] In one embodiment of the present invention, the use of the NCC matching algorithm to determine the matching strength between adjacent pathological slice images includes: determining the matching strength between adjacent pathological slice images according to the following formula: in, and Pathological slice image and The mean of and is the overlapping area of ​​the two registered pathological slice images, , is the matching strength between pathological slice images.

[0013] In one embodiment of the present invention, the step of sequentially obtaining the optimal position of the pathological slice image through the Bayesian inference model and nonlinear optimization includes: obtaining the optimal position of the pathological slice image according to the following formula: ; ; in, is the optimal position of the pathological slice image; yes The initial solution; C is the pathological slice image to be spliced, A is the pathological slice image adjacent to the left, and B is the pathological slice image adjacent to the top. and is the matching strength; , represents the rotation angle between image A and image C; Represents the rotation angle between image B and image C; represents the observation position of image C obtained through image A; represents the observation position of image C obtained through image B; is the minimum error model; is the likelihood function, which imposes constraints on the pathological slice image pose through the observation factor. is the initial position of the pathological slice image; represents the position of image A, Indicates the position of image B.

[0014] As described above, the high-precision real-time medical image stitching method and system of the present invention have the following beneficial effects: The present invention designs a factor graph between a pathological slice image to be spliced ​​and its adjacent pathological slice images through a SURF_NCC matching method, and then establishes a Bayesian reasoning model according to the factor graph, and reduces the error between the initial position and the shooting position of the pathological slice image to be spliced ​​through a minimum error model in the Bayesian reasoning model, thereby obtaining the optimal position of the pathological slice image to be spliced, and then splicing the pathological slice images according to the optimal position, and then repeating the above steps for each acquired pathological slice image to be spliced ​​until a complete pathological slice image is obtained by splicing. Therefore, the present invention uses factor graphs, Bayesian reasoning and nonlinear optimization to calculate the optimal position of each pathological slice image to be spliced, which can not only maintain image integrity, but also perform scanning and splicing, and make the image splicing result more accurate by correcting the splicing position of the image, thereby solving the problem of being highly dependent on matching between images and being unable to cope with poor matching and being unable to scan and splice. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of a high-precision real-time medical image stitching method in a first embodiment of the present invention; Figure 2 is a schematic diagram of a high-precision real-time medical image stitching system in a second embodiment of the present invention; Figure 3 is a schematic diagram of an electronic device in a third embodiment of the present invention; Figure 4 is a schematic diagram of a factor graph in the present invention; Figure 5 It is a schematic diagram of the factor graph in the present invention (including only the connection strength of the observation position during the calculation process, and the sample is a concentrated type); Figure 6 It is a schematic diagram of the factor graph in the present invention (including only the connection strength of the observation position during the calculation process, and the samples are dispersed); Figure 7 It is a schematic diagram of the factor relationship between nodes A, B and C in the present invention. DETAILED DESCRIPTION

[0016] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0017] The first embodiment of the present invention relates to a high-precision real-time medical image stitching method. It should be noted that the present invention is only for the area containing the sample, that is, the area containing the pathological section. The process is as follows: Figure 1 As shown, the details are as follows: Step 101, use a scanner to scan the pathological slices, and obtain the pathological slice images to be spliced ​​according to the scanning results.

[0018] Step 102, during the scanning process, a pathological slice image to be spliced ​​is obtained by scanning the pathological slice at a certain position by a scanner, and a factor graph between the pathological slice image and its adjacent pathological slice images is determined.

[0019] Specifically, after obtaining the pathological slice image to be spliced ​​obtained by scanning the pathological slice with a scanner, before determining the factor graph between the pathological slice image and its adjacent pathological slice images, it also includes: confirming whether the pathological slice image to be spliced ​​is a starting image; if so, the pathological slice image is not subsequently optimized and spliced ​​directly; if not, confirming whether the pathological slice image has adjacent pathological slice images; if so, the pathological slice image is subsequently optimized; if not, the image weight of the non-pathological slice area in the corresponding direction is set to 0.

[0020] More specifically, in this embodiment, the pathological slice image can be regarded as a node, and the process of determining the factor graph between the pathological slice image and its adjacent pathological slice images in practical application is as follows: Design the factor graph: the factors include: 1. Relative pose factor, the relative pose factor includes the displacement offset, matching strength and rotation angle between nodes, and the error between the shooting position (the actual grid position calculated by the SURF_NCC matching method) and the initial position (the preset ideal grid position); 2. Priori factor, the priori factor includes the initial position of the node; It should be noted that there is also a SURF_NCC matching method, the SURF_NCC matching method combines the SURF matching algorithm and the method of taking the negative logarithm of the NCC, and the displacement, matching strength and rotation angle between the pathological slice images are calculated by the SURF_NCC matching method at the same time, which are used as the relative pose factor. The NCC (Normalized Cross-Correlation) matching algorithm is an important indicator for measuring the local similarity of two images, and the calculation method is as follows: , and Pathological slice image and The mean of and is the overlapping area of ​​the two registered pathological slice images, , is the matching strength between pathological slice images; There are also nodes and The observations between can be expressed as factors , which represents The position of the observation node , get the relative pose factor, and compare it with the node The position of is calculated (translation and rotation), yes Observation position; and each newly added node will be matched with the node on the left and the node above. The node on the left is A, the node above is B, and the newly added node is C. For details on the factor relationship, please refer to Figure 7 ,in, and It is the negative logarithm of the correlation NCC after registration calculation, recorded as ENCC, that is, the matching strength, and They are the displacements calculated by registration, and the factor graph is detailed in Figures 4 to 6 .

[0021] Step 103: establishing a Bayesian inference model based on the factor graph between the pathological slice image and its adjacent pathological slice images, and obtaining the optimal position of the pathological slice image through the Bayesian inference model and nonlinear optimization in sequence.

[0022] Specifically, a Bayesian inference model is established based on the factor graph between the pathological slice image and its adjacent pathological slice images, and the optimal position of the pathological slice image is obtained through the Bayesian inference model and nonlinear optimization in turn. The process in practical application is as follows: a Bayesian optimization framework is established, and the factor graph optimization is based on the Bayesian inference framework, and the posterior probability of the node is calculated incrementally: First, the input needs to be simplified. The conventional Bayesian input is the input image , each pathological slice image is used as a node , since the scanner scans in a translational manner and the scanned data is planar data, that is, a planar image, then for the input node To simplify, , Corresponding to their preset coordinate positions (preset ideal grid positions), is the initial rotation angle (the specific value is 0), which is used to correct some system errors such as camera rotation; Bayesian formula: , in order to establish a Bayesian inference model, is the initial position of the node (that is, the ideal grid position of the node); is the optimal location of the node, through (All observation positions) weighted sum is obtained; is the target posterior distribution, which indicates the probability of the best node position after observation; is the likelihood function, which imposes constraints on the node pose through the observation factor; is the prior distribution (error probability of the node); usually a priori pose constraints are imposed on the starting node to ensure the stability of the graph; set up The error probability model of the node conforms to the Gaussian distribution. Then for the nodes added later, the joint probability distribution of all positions can be obtained by multiplying the Gaussian error of each position: , convert the maximum probability problem into the likelihood function to solve the minimum error; assuming that the new node C, the node on its left is A, the node above is B, and the matching strengths are and , then the minimum error model established is: ; ,in, yes The initial solution of It is calculated through factor graph, see Figure 4 In this embodiment, that is Figure 4 The optimized position of the new node C in the factor graph is obtained, that is, the optimized position of the new node can be obtained through the factor graph, and then the optimized position of the new node is obtained by iterating the minimum error model in the Bayesian inference model to obtain the optimal position of the new node. ; Represents the rotation angle between node A and node C; Represents the rotation angle between node B and node C; represents the observed position of node C obtained through node A; represents the observed position of node C obtained through node B; Indicates the location of node A; Indicates the position of node B; this can control the influence from the left and top respectively. The optimization process will automatically balance the influence of the two factors on the final estimate. Finally, nonlinear optimization is used to calculate the new node position in real time, that is, the optimal position Y of the pathological slice image. Among them, nonlinear optimization will be performed using the C++ gtsam library, which is a dedicated optimization tool. Mathematically speaking, it is solved through sparse matrices, and the nonlinear optimization is a mature mathematical calculation method.

[0023] Step 104: splicing the pathological slice images according to the optimal positions of the pathological slice images.

[0024] Step 105, obtain the pathological slice image to be spliced ​​obtained by moving the scanner to a new position to scan the pathological slice, and repeat steps 102 to 104 for the pathological slice image, that is, scan to obtain a pathological slice image to be spliced, then calculate the optimal position of the pathological slice image and splice it until a complete pathological slice image is obtained.

[0025] Specifically, a complete pathological slice image includes several spliced ​​pathological slice images, the content of each pathological slice image includes different areas of the entire pathological slice, and the complete pathological slice image can be considered as a grid map composed of several images, each small grid of the grid map is a pathological slice image. The more important technical point in the present invention is to calculate the specific splicing positions of each pathological slice image to be spliced ​​in the final grid map.

[0026] Further, the above process can also be specifically as follows: Step 1: Use a scanner to scan the pathological slice, and the following steps are all performed during the scanning process; Step 2: First determine whether node C is the starting node, skip the optimization of the starting node, and then determine whether there are adjacent nodes A and B; if not, set the node weights in the corresponding direction to 0; if so, design a factor graph through the matching relationship, that is, use the SURF_NCC matching method to calculate the factors of the new node and the adjacent node, which include: the preset grid position, the actual position of the scanning camera (obtained by the displacement calculation of the SURF registration method), the displacement, the matching strength and the rotation angle; Step 3: Establish a Bayesian reasoning model through the factor graph in the second step to simplify The spatial model is three-dimensional , construct the minimum error model: ; ; Get the best position by minimizing the error: that is ; Step 4: Optimize the inference of the position of the new image through the Bayesian inference model in the third step and paste the pathological slice image on the canvas; Step 5: Move the scanner's camera to the new position to shoot, repeat steps 2 to 3, that is, scan a position, deduce the exact position of the new image and paste the pathological slice image at the new position.

[0027] The second embodiment of the present invention relates to a high-precision real-time medical image stitching system, see Figure 2 ,include: The image optimization processing module is used to scan the pathological slices with a scanner and obtain the pathological slice images to be spliced ​​according to the scanning results; in the scanning process, the pathological slice images to be spliced ​​are obtained by scanning the pathological slices at a certain position by the scanner, and the pathological slice images are optimized, the optimal position of the pathological slice images is obtained according to the optimization processing results, and the pathological slice images are spliced ​​according to the optimal positions of the pathological slice images; The iterative optimization processing module is used to obtain the pathological slice image to be spliced ​​obtained by moving the scanner to a new position to scan the pathological slice, and repeat the optimization processing process on the pathological slice image, so that each pathological slice image scanned by the scanner is iteratively optimized in sequence according to the scanning order, so that the scanning process can be carried out while splicing until a complete pathological slice image is obtained, wherein the complete pathological slice image includes several spliced ​​pathological slice images, and the content of each pathological slice image contains different areas of the entire pathological slice.

[0028] It is not difficult to find that this embodiment is a system embodiment corresponding to the first embodiment, and this embodiment can be implemented in conjunction with the first embodiment. The relevant technical details mentioned in the first embodiment are still valid in this embodiment, and in order to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied in the first embodiment.

[0029] It is worth mentioning that all modules involved in this embodiment are logic modules. In practical applications, a logic unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by the present invention, but this does not mean that there are no other units in this embodiment.

[0030] The third embodiment of the present invention relates to an electronic device, see Figure 3 ,include: At least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the high-precision real-time medical image stitching method as described above.

[0031] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.

[0032] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0033] A fourth embodiment of the present invention relates to a computer-readable storage medium storing a computer program, and the computer program implements the above method embodiment when executed by a processor.

[0034] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0035] In summary, the present invention designs a reasonable factor graph to describe the relationship between adjacent pathological slice images in the sample area, designs a Bayesian reasoning model, and uses two constraints of a weighted method to infer the positions of the pathological slice images to be spliced ​​and the surrounding pathological slice images. Therefore, the present invention uses a factor graph and a Bayesian reasoning model to correct the position of each pathological slice image to be spliced, and finally splices all images into a complete digital image to achieve the purpose of real-time splicing of large-scale images to generate digital pathological slices, so that scanning and splicing can be performed in real time, and even for poor matches, the position can be optimized, thereby solving the problem of being highly dependent on matching between images and being unable to cope with poor matches and being unable to scan and splice.

[0036] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. All equivalent modifications or changes made by those skilled in the art without departing from the spirit and technical ideas disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A high-precision real-time medical image stitching method, characterized in that: The following steps are involved: Scanning the pathological sections with a scanner, and obtaining the pathological section images to be spliced ​​according to the scanning results; During the scanning process, a pathological slice image to be spliced ​​is obtained by scanning the pathological slice at a certain position by a scanner, and the pathological slice image is optimized, and the optimal position of the pathological slice image is obtained according to the optimization result, and the pathological slice image is spliced ​​according to the optimal position of the pathological slice image; The pathological slice image to be spliced ​​is obtained by moving the scanner to a new position to scan the pathological slice, and the optimization process is repeated for the pathological slice image, so that each pathological slice image scanned by the scanner is iteratively optimized in sequence according to the scanning order, so that the pathological slice image can be spliced ​​while scanning until a complete pathological slice image is obtained, wherein the complete pathological slice image includes several spliced ​​pathological slice images, and the content of each pathological slice image contains different areas of the entire pathological slice.

2. A high-precision real-time medical image stitching method according to claim 1, characterized in that: After the pathological slice images to be spliced ​​are obtained by scanning the pathological slices with a scanner, and before the pathological slice images are optimized, the method further includes: Confirming whether the pathological section image to be spliced ​​is a starting image; If so, the pathological slice image is not optimized and is directly spliced. If not, it is confirmed whether the pathological slice image has adjacent pathological slice images. If so, the pathological slice image is optimized. If not, the image weight of the non-pathological slice area in the corresponding direction is set to 0.

3. The high-precision real-time medical image stitching method according to claim 1, characterized in that: The step of optimizing the pathological slice image and obtaining the optimal position of the pathological slice image according to the optimization result includes: Determine a factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice images, wherein the factor graph includes a relative pose factor and a priori factors, the relative pose factor includes a displacement, a matching strength and a rotation angle between the pathological slice images, and the priori factor includes an initial position of each pathological slice image to be spliced; A Bayesian inference model is established according to a factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice image, and the optimal position of the pathological slice image is obtained through the Bayesian inference model and nonlinear optimization in sequence, wherein the Bayesian inference model includes a minimum error model.

4. A high-precision real-time medical image stitching method according to claim 3, characterized in that: The relative pose factor also includes the shooting position of the pathological slice image, the shooting position of the pathological slice image is the actual grid position of the pathological slice image to be spliced ​​calculated by the SURF_NCC matching method, and the error between the initial position and the shooting position of the pathological slice image can be obtained through the factor graph, wherein the initial position of the pathological slice image to be spliced ​​is the preset ideal grid position of the pathological slice image; There are also observation positions between pathological slice images. and The observations between can be expressed as bit factors , the factor represents the pathological slice image The position of the pathological slice image is observed , and obtain the relative pose factor, and compare it with the pathological slice image The position is calculated, and Pathological slice image Observation location.

5. The high-precision real-time medical image stitching method according to claim 3, characterized in that: The step of determining a factor graph between the pathological slice image to be spliced ​​and its adjacent pathological slice images includes: The displacement, matching strength and rotation angle between adjacent pathological slice images are determined using the SURF_NCC matching method, wherein the SURF_NCC matching method combines the SURF feature extraction algorithm and the NCC matching algorithm.

6. A high-precision real-time medical image stitching method according to claim 5, characterized in that: Determining the matching strength between adjacent pathological slice images using the NCC matching algorithm includes: The matching strength between adjacent pathological slice images is determined according to the following formula: ； in, and Pathological slice image and The mean of and is the overlapping area of ​​the two registered pathological slice images, , is the matching strength between pathological slice images.

7. The high-precision real-time medical image stitching method according to claim 4, characterized in that: The step of obtaining the optimal position of the pathological slice image by sequentially using a Bayesian inference model and nonlinear optimization includes: The optimal position of the pathological slice image is obtained according to the following formula: ； ； in, is the optimal position of the pathological slice image; yes The initial solution; C is the pathological slice image to be spliced, A is the pathological slice image adjacent to the left, and B is the pathological slice image adjacent to the top. and is the matching strength; Represents the rotation angle between image A and image C; Represents the rotation angle between image B and image C; represents the observation position of image C obtained through image A; represents the observation position of image C obtained through image B; is the minimum error model; is the likelihood function, which imposes constraints on the pathological slice image pose through the observation factor. is the initial position of the pathological slice image; represents the position of image A, Indicates the position of image B.

8. A high-precision real-time medical image stitching system, characterized by: include: An image optimization processing module is used to scan the pathological slices using a scanner, and obtain the pathological slice images to be spliced ​​according to the scanning results; During the scanning process, a pathological slice image to be spliced ​​is obtained by scanning the pathological slice at a certain position by a scanner, and the pathological slice image is optimized, and the optimal position of the pathological slice image is obtained according to the optimization result, and the pathological slice image is spliced ​​according to the optimal position of the pathological slice image; The iterative optimization processing module is used to obtain the pathological slice image to be spliced ​​obtained by moving the scanner to a new position to scan the pathological slice, and repeat the optimization processing process on the pathological slice image, so that each pathological slice image scanned by the scanner is iteratively optimized in sequence according to the scanning order, so that it can be scanned and spliced ​​at the same time during the scanning process until a complete pathological slice image is obtained, wherein the complete pathological slice image includes several spliced ​​pathological slice images, and the content of each pathological slice image contains different areas of the entire pathological slice.

9. An electronic device, characterized in that: include: at least one processor; as well as, A memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor so that the at least one processor can execute the high-precision real-time medical image stitching method as described in any one of claims 1 to 7.

Images