Method and system for selecting optimal fit of allogeneic intervertebral disc

Abstract

The present invention provides a method and system for selecting the best fit of an allogeneic intervertebral disc, and relates to the technical field of image data processing. The method comprises: obtaining intervertebral disc image data of a target patient; interpolating the intervertebral disc image data of the target patient through an optical flow algorithm; constructing a three-dimensional model of the intervertebral disc of the target patient according to the interpolated intervertebral disc image data of the target patient; obtaining the intervertebral disc volume of the target patient according to the three-dimensional model of the intervertebral disc of the target patient; calculating the volume difference between each intervertebral disc sample in an allogeneic intervertebral disc sample library and the target patient; screening out intervertebral disc samples to be matched whose volume difference is less than a preset difference from the allogeneic intervertebral disc sample library; calculating the matching degree between each intervertebral disc sample to be matched and the target patient through a deep learning network; and determining the intervertebral disc sample to be matched with the highest matching degree as the best fit intervertebral disc sample.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a method and system for optimally adapting and selecting an allogeneic intervertebral disc. Background Art

[0002] With the aging of the population and the change of lifestyle, the incidence of intervertebral disc degeneration has increased year by year, becoming one of the major problems affecting human health. Traditional treatments such as physical therapy and drug therapy are often difficult to fundamentally solve the problem, while intervertebral disc replacement surgery is an effective treatment method.

[0003] Allograft disc replacement is a surgical procedure to treat cervical spondylosis that involves removing the patient's damaged disc and replacing it with a disc donated from someone else (an allograft disc). This procedure is an alternative to spinal fusion surgery and artificial disc replacement surgery for cervical spondylosis.

[0004] The purpose of allogeneic disc replacement surgery is to relieve the symptoms of patients with cervical spondylosis, such as neck pain, arm pain, numbness, etc., while maintaining the stability and mobility of the cervical spine. By replacing the damaged intervertebral disc with an allogeneic disc, the aim is to restore the normal structure and function of the cervical spine.

[0005] The key to allogeneic disc transplantation is to establish an allogeneic disc library and collect as many samples as possible. When treating patients with disc lesions, it is necessary to screen the samples in the disc library based on imaging data to select the allogeneic disc that best matches the patient's disc volume for surgery to ensure the best surgical results.

[0006] Currently, the selection of allogeneic intervertebral discs mainly relies on the doctor's experience, with low matching efficiency and lack of precise quantitative standards, which can easily lead to mismatch between the samples in the intervertebral disc library and the patient's intervertebral disc, affecting the surgical effect and patient recovery. Summary of the invention

[0007] In order to solve the technical problems that the existing technology mainly relies on the doctor's experience, has low matching efficiency, lacks precise quantitative standards, and easily causes the samples in the intervertebral disc library to not match the patient's intervertebral disc, thus affecting the surgical effect and patient recovery, the present invention provides a method and system for optimal adaptation selection of allogeneic intervertebral discs.

[0008] The technical solution provided by the embodiment of the present invention is as follows:

[0009] First aspect

[0010] An embodiment of the present invention provides a method for selecting an optimally adapted allogeneic intervertebral disc, comprising:

[0011] S1: Acquire intervertebral disc image data of the target patient;

[0012] S2: performing interpolation processing on the intervertebral disc image data of the target patient by using an optical flow algorithm;

[0013] S3: constructing a three-dimensional model of the intervertebral disc of the target patient according to the interpolated intervertebral disc image data of the target patient;

[0014] S4: acquiring the intervertebral disc volume of the target patient according to the three-dimensional model of the intervertebral disc of the target patient;

[0015] S5: calculating the volume difference between each intervertebral disc sample in the allogeneic intervertebral disc sample library and the target patient;

[0016] S6: Screening out intervertebral disc samples to be matched whose volume difference is less than a preset difference from the allogeneic intervertebral disc sample library;

[0017] S7: calculating the matching degree between each of the intervertebral disc samples to be matched and the target patient through a deep learning network;

[0018] S8: Determine the intervertebral disc sample to be matched with the highest matching degree as the best matching intervertebral disc sample.

[0019] Second aspect

[0020] An embodiment of the present invention provides an optimally adapted allogeneic intervertebral disc selection system, comprising:

[0021] processor;

[0022] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for selecting the best fit of the allogeneic intervertebral disc as described in the first aspect is implemented.

[0023] The third aspect

[0024] An embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method for selecting the best-fitting allogeneic intervertebral disc as described in the first aspect is implemented.

[0025] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0026] In the present invention, a three-dimensional model of the intervertebral disc is reconstructed based on the patient's intervertebral disc medical image data, and then artificial intelligence is used to automatically calculate the matching degree between each intervertebral disc sample to be matched in the allogeneic intervertebral disc sample library and the target patient, and the best matching intervertebral disc sample is determined, avoiding reliance on the doctor's subjective experience and eliminating the need for human intervention, thereby improving the efficiency and accuracy of matching, providing strong support for allogeneic intervertebral disc replacement surgery, and helping patients recover. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 A schematic diagram of a flow chart of a method for selecting the best-fitting allogeneic intervertebral disc provided by an embodiment of the present invention;

[0029] Figure 2 A schematic diagram of a flow chart of a method for constructing a three-dimensional model of an intervertebral disc provided in an embodiment of the present invention;

[0030] Figure 3 A schematic diagram of a matching degree calculation method provided by an embodiment of the present invention;

[0031] Figure 4 A flowchart of a matching degree calculation method provided by an embodiment of the present invention;

[0032] Figure 5 A schematic structural diagram of an optimally adapted allogeneic intervertebral disc selection system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0034] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0035] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0036] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0037] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0038] Reference Manual Attached Figure 1 , showing a flow chart of a method for selecting the best fit of an allogeneic intervertebral disc provided by an embodiment of the present invention.

[0039] The embodiment of the present invention provides a method for selecting an optimal allogeneic intervertebral disc, which can be implemented by an optimal allogeneic intervertebral disc selection device, which can be a terminal or a server. The processing flow of the method for selecting an optimal allogeneic intervertebral disc can include the following steps:

[0040] S1: Acquire intervertebral disc image data of a target patient.

[0041] The intervertebral disc image data may be a computed tomography (CT scan) image.

[0042] S2: Interpolate the intervertebral disc image data of the target patient through the optical flow algorithm.

[0043] The optical flow algorithm is a technique used to estimate the motion between consecutive frames. The basic idea is to infer the motion trajectory of an object or scene by tracking the position changes of specific pixels in the image between different frames. The displacement of these pixels in the image is called optical flow.

[0044] It should be noted that CT scans are image data obtained through tomographic scanning, that is, there are discontinuities between each CT slice. Interpolation processing can fill the gaps between adjacent CT slices, making the image sequence more continuous and smooth in space, which can better reflect the structure and morphology of the intervertebral disc and facilitate subsequent analysis. At the same time, the discontinuity in CT scan images may cause artifacts or artifacts, affecting image quality and diagnostic accuracy. Interpolation processing can reduce the appearance of these artifacts and artifacts and improve image reliability and visualization.

[0045] In a possible implementation, S2 specifically includes sub-steps S201 and S202:

[0046] S201: Calculate the displacement vector field estimation between adjacent intervertebral disc image data.

[0047] Optionally, S201 specifically includes:

[0048] Extract image features of intervertebral disc image data.

[0049] The image features may specifically be edge features, texture features, shape features, gradient features, and the like.

[0050] According to the image features, the feature points between adjacent intervertebral disc image data are matched using the K nearest neighbor matching algorithm.

[0051] Among them, the K-nearest neighbor matching algorithm is a simple and effective machine learning algorithm, which is often used for classification and regression tasks, and can also be used for applications such as feature matching. In image processing, the KNN algorithm is often used for matching feature points.

[0052] After the traditional K nearest neighbor matching algorithm determines K nearest neighbors, it is still difficult to determine the best matching point pair. Therefore, the present invention provides an improved K nearest neighbor matching algorithm.

[0053] According to the image features, the similarity between each pixel in the previous intervertebral disc image data and the next intervertebral disc image data is calculated.

[0054] Optionally, the Euclidean distance or cosine similarity between each pixel point in the previous intervertebral disc image data and the next intervertebral disc image data may be calculated.

[0055] Arrange the pixels in the next intervertebral disc image data in descending order of similarity.

[0056] The top K pixels are selected as the K nearest neighbors of the current pixel in the previous intervertebral disc image data.

[0057] The pixel points in the previous intervertebral disc image data are randomly combined, and a first reference pixel point, a second reference pixel point and a third reference pixel point are selected to form a reference triangle.

[0058] A first pixel to be registered is randomly selected from the K neighbors of the first reference pixel, a second pixel to be registered is randomly selected from the K neighbors of the second reference pixel, and a third pixel to be registered is randomly selected from the K neighbors of the third reference pixel, and a triangle to be registered is formed by the first pixel to be registered, the second pixel to be registered and the third pixel to be registered.

[0059] Calculate the similarity between each triangle to be registered and the reference triangle.

[0060] Specifically, the similarity between the triangle to be registered and the reference triangle can be calculated by calculating the Euclidean distance or the cosine similarity based on the angle and the length of the side of the triangle.

[0061] The three pixels to be registered corresponding to the triangle to be registered with the highest similarity are determined as the best matching points of the three reference pixels, and the best matching point pair is formed.

[0062] According to the best matching point pair, the image registration is performed on the next intervertebral disc image data after perspective transformation.

[0063] In the present invention, by introducing a triangle structure and selecting the triangle to be registered with the highest similarity, the best matching point pair can be determined more accurately, which can avoid possible mismatching in the traditional KNN algorithm and improve the accuracy and reliability of matching.

[0064] Based on the matched feature points, the displacement vector field estimation between adjacent intervertebral disc image data is calculated.

[0065] Specifically, the displacement vector is:

[0066]

[0067]

[0068] Among them, (v x , v y ) represents the displacement vector, (x t (j), y t (j)) represents the coordinates of the jth matching feature point in the tth intervertebral disc image data, (x t+1 (j), y t+1 (j)) represents the coordinates of the j-th matching feature point in the t+1-th intervertebral disc image data, and J represents the total number of matching feature points.

[0069] S202: performing interpolation processing on adjacent intervertebral disc image data by using an optical flow algorithm based on the displacement vector field estimation.

[0070] Optionally, S202 specifically includes:

[0071] Based on the optical flow algorithm, the forward prediction image and the backward prediction image are determined.

[0072] The forward prediction image satisfies:

[0073]

[0074] in, represents the forward prediction image, (x, y) represents the pixel coordinates, (v x , v y ) represents the displacement vector, α represents the interpolation coefficient, Represents the t-th intervertebral disc image data.

[0075] The backward prediction image satisfies:

[0076]

[0077] in, represents the forward prediction image, Represents the t+1th intervertebral disc image data.

[0078] According to the forward prediction image and the backward prediction image, the adjacent intervertebral disc image data are interpolated:

[0079]

[0080] in, Represents the interpolated image.

[0081] In the present invention, the forward and backward images predicted by the optical flow algorithm can be combined to interpolate adjacent image data to improve the accuracy of the interpolation result. The interpolation process can reduce artifacts and artifacts caused by discontinuities between images. By combining the forward and backward predicted images, the interpolated image can better reflect the morphology and structure of the intervertebral disc and avoid the unrealistic effect caused by missing image data.

[0082] S3: constructing a three-dimensional model of the intervertebral disc of the target patient according to the interpolated intervertebral disc image data of the target patient.

[0083] Reference Manual Attached Figure 2 , showing a schematic flow chart of a method for constructing a three-dimensional model of an intervertebral disc provided in an embodiment of the present invention.

[0084] In a possible implementation, S3 specifically includes sub-steps S301 to S310:

[0085] S301: performing image segmentation on each intervertebral disc image data after interpolation processing of the target patient, and determining the intervertebral disc region image in each intervertebral disc image data.

[0086] Specifically, K-means clustering algorithm, region growing algorithm, neural network, etc. can be used to perform image segmentation on each intervertebral disc image data of the target patient after interpolation processing.

[0087] S302: stacking the intervertebral disc region images at preset intervals.

[0088] The preset interval is the interval between the intervertebral disc area images during tomographic scanning and interpolation processing.

[0089] S303: Connect the edges of the intervertebral disc region image to form a volume data space.

[0090] S304: Initialize the virtual viewpoint.

[0091] S305: Obtain a set of rays starting from the virtual viewpoint and having intersections with the volume data space.

[0092] S306: Perform multi-threaded sampling on rays passing through the volume data space through the CPU and GPU with the goal of minimizing load imbalance.

[0093] Optionally, the load imbalance degree is specifically:

[0094]

[0095] Among them, R represents the load imbalance, R max represents the highest load on each thread, R min Indicates the lowest load on each thread.

[0096] In the present invention, by utilizing the parallel computing capabilities of the CPU and GPU, hardware resources can be fully utilized to improve the overall performance and efficiency of the system. The CPU and GPU are each good at processing different types of tasks. By reasonably allocating tasks, their advantages can be fully utilized to improve sampling speed and efficiency. At the same time, by minimizing the load imbalance, a reasonable allocation of computing resources can be achieved, so that the load between each thread is balanced. Doing so can avoid performance bottlenecks caused by overloading certain threads and improve the overall stability and reliability of the system.

[0097] S307: Calculate the color and opacity of the sampling points intersecting the intervertebral disc region image.

[0098] Specifically, the color and opacity in the three-dimensional model may be converted according to the grayscale values ​​of the sampling points where the two-dimensional intervertebral disc region images intersect.

[0099] S308: Perform interpolation processing on the data points between two sampling points.

[0100] It should be noted that in the process of constructing a 3D model, the sampling points are usually discretely distributed points in the volume data, and do not necessarily correspond to the actual data points in the volume data. Therefore, when calculating the color and opacity of the sampling points that intersect with the intervertebral disc area image, interpolation processing is required to obtain the color and opacity of the data points between the two sampling points.

[0101] Specifically, commonly used interpolation methods include linear interpolation, bilinear interpolation, trilinear interpolation, etc. These interpolation methods can infer the data values ​​of other positions between the known sampling points based on the positional relationship and data values ​​between these sampling points, so as to achieve the continuity of color and opacity processing of the entire 3D model.

[0102] S309: Accumulate the color and opacity of each data point in the volume data space through the optical model.

[0103] The optical model may be a Fresnel equation, a Lambertian model, a Ray Tracing model, or the like.

[0104] It should be noted that in 3D graphics rendering, optical models are usually used to simulate the interaction between light and objects to determine the final color and opacity. Through the optical model, the color and opacity of each data point in the volume data space are accumulated, which means that the impact of light passing through each data point is considered to determine the final color and opacity.

[0105] S310: Construct a three-dimensional model of the intervertebral disc of the target patient based on the accumulated results.

[0106] In the present invention, by converting the intervertebral disc image data after interpolation processing into a three-dimensional model, the morphology and structure of the intervertebral disc can be displayed in a more intuitive and three-dimensional way. This helps doctors and patients to have a more comprehensive understanding of the intervertebral disc, including information such as morphology, size, and position.

[0107] S4: Obtain the intervertebral disc volume of the target patient according to the intervertebral disc three-dimensional model of the target patient.

[0108] Specifically, the three-dimensional model can be represented as a set of discrete voxels (volume pixels), and then the volume of the entire intervertebral disc can be obtained by accumulating or integrating the volumes of these voxels.

[0109] S5: Calculate the volume difference between each intervertebral disc sample in the allogeneic intervertebral disc sample library and the target patient.

[0110] Optionally, the volume difference is specifically:

[0111]

[0112] Among them, σ i Represents the volume difference with the i-th intervertebral disc sample, V a represents the intervertebral disc volume of the target patient, V i represents the volume of the i-th intervertebral disc sample.

[0113] Furthermore, in addition to the difference in volume, which can be used for preliminary screening, height, width, depth, etc. can also be considered for preliminary screening.

[0114] S6: From the allogeneic intervertebral disc sample library, select intervertebral disc samples to be matched whose volume difference is less than a preset difference.

[0115] In the present invention, the calculation of volume difference can be used to preliminarily screen intervertebral disc samples, helping doctors to quickly identify samples with similar volumes to those of target patients, which helps to narrow the scope of the intervertebral disc sample library, reduce the amount of calculation, and improve the efficiency of matching.

[0116] S7: Through the deep learning network, the matching degree between each intervertebral disc sample to be matched and the target patient is calculated.

[0117] Specifically, a deep learning network can be a Siamese network. A Siamese network is a deep learning architecture that is often used to process paired data, such as images, text, or other types of data. It consists of two sub-networks of the same structure, each with its own parameters. The two sub-networks are called "twins" because they have the same structure and parameters. The goal of a twin network is to learn how to map two input data into a shared space for comparison, matching, or classification.

[0118] Reference Manual Attached Figure 3 , shows a structural schematic diagram of a matching degree calculation method provided by an embodiment of the present invention.

[0119] Reference Manual Attached Figure 4 , which shows a flow chart of a matching degree calculation method provided in an embodiment of the present invention.

[0120] In a possible implementation, S7 specifically includes sub-steps S701 to S704:

[0121] S701: Construct a two-branch twin network based on convolutional neural network.

[0122] Among them, Convolutional Neural Network (CNN) is a deep learning model specifically designed for processing neural networks with grid-structured data.

[0123] S702: Extracting a first depth feature of the three-dimensional model of the intervertebral disc in the intervertebral disc sample to be matched by using a first convolutional neural network.

[0124] S703: Extracting a second depth feature of the target patient's intervertebral disc three-dimensional model through a second convolutional neural network.

[0125] S704: Calculate the matching degree between the intervertebral disc sample to be matched and the target patient according to the first depth feature and the second depth feature:

[0126]

[0127] Among them, δ i represents the matching degree between the i-th intervertebral disc sample to be matched and the target patient, E1(i) represents the first depth feature of the intervertebral disc three-dimensional model in the i-th intervertebral disc sample to be matched, and E2 represents the second depth feature of the intervertebral disc three-dimensional model of the target patient.

[0128] In the present invention, a deep learning network is used to calculate the matching degree, which can realize an automated matching process and reduce the need for human intervention. This can improve the matching efficiency and reduce the impact of human factors on the matching results, making the matching process more stable and reliable.

[0129] In a possible implementation, the training method of the two-branch twin network specifically includes:

[0130] Construct the inter-sample loss function of the two-branch twin network:

[0131]

[0132] Among them, L( ) represents the loss function, θ represents the parameters of the two-branch twin network, and y i represents the matching result between the i-th intervertebral disc sample to be matched and the target patient. When , it means that the i-th intervertebral disc sample to be matched is successfully matched with the target patient. When , it means that the i-th intervertebral disc sample to be matched fails to match the target patient, max( ) means taking a larger value, and δ0 means the preset matching degree.

[0133] The two-branch Siamese network is trained with the goal of minimizing the inter-sample loss function.

[0134] Specifically, the two-branch twin network can be trained by gradient descent method, genetic algorithm, particle swarm optimization algorithm, etc.

[0135] In the present invention, the use of the inter-sample loss function to train the two-branch twin network can fully consider the matching results and preset matching degree between samples, thereby improving the matching accuracy and optimizing the network parameters, making the network more suitable for the matching task of intervertebral disc samples.

[0136] S8: Determine the intervertebral disc sample to be matched with the highest matching degree as the best matching intervertebral disc sample.

[0137] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0138] In the present invention, a three-dimensional model of the intervertebral disc is reconstructed based on the patient's intervertebral disc medical image data, and then artificial intelligence is used to automatically calculate the matching degree between each intervertebral disc sample to be matched in the allogeneic intervertebral disc sample library and the target patient, and the best-matching intervertebral disc sample is determined, avoiding reliance on the doctor's subjective experience and eliminating the need for human intervention, thereby improving the efficiency and accuracy of matching, providing strong support for allogeneic intervertebral disc replacement surgery, and helping patients recover.

[0139] Reference Manual Attached Figure 5 , showing a schematic structural diagram of an optimally adapted allogeneic intervertebral disc selection system provided by the present invention.

[0140] The present invention further provides an optimally adapted allogeneic intervertebral disc selection system 20, which is applied to the above-mentioned optimally adapted allogeneic intervertebral disc selection method, comprising:

[0141] Processor 201.

[0142] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201 , the method for selecting the best-fitting allogeneic intervertebral disc in the method embodiment is implemented.

[0143] The best-fit allogeneic intervertebral disc selection system 20 provided by the present invention can execute the above-mentioned best-fit allogeneic intervertebral disc selection method and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate on them.

[0144] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0145] In the present invention, a three-dimensional model of the intervertebral disc is reconstructed based on the patient's intervertebral disc medical image data, and then artificial intelligence is used to automatically calculate the matching degree between each intervertebral disc sample to be matched in the allogeneic intervertebral disc sample library and the target patient, and the best matching intervertebral disc sample is determined, avoiding reliance on the doctor's subjective experience and eliminating the need for human intervention, thereby improving the efficiency and accuracy of matching, providing strong support for allogeneic intervertebral disc replacement surgery, and helping patients recover.

[0146] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0147] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0148] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0149] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0150] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0151] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0152] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0153] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0154] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0155] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0156] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0157] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0158] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the method for selecting the best fit of an allogeneic intervertebral disc as described in the method embodiment is implemented.

[0159] A computer-readable storage medium provided by the present invention can implement the steps and effects of the best-fitting selection method for allogeneic intervertebral discs in the above method embodiment, and to avoid repetition, the present invention will not elaborate on them.

[0160] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0161] In the present invention, a three-dimensional model of the intervertebral disc is reconstructed based on the patient's intervertebral disc medical image data, and then artificial intelligence is used to automatically calculate the matching degree between each intervertebral disc sample to be matched in the allogeneic intervertebral disc sample library and the target patient, and the best matching intervertebral disc sample is determined, avoiding reliance on the doctor's subjective experience and eliminating the need for human intervention, thereby improving the efficiency and accuracy of matching, providing strong support for allogeneic intervertebral disc replacement surgery, and helping patients recover.

[0162] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0163] There are a few points to note:

[0164] (1) The drawings of the embodiments of the present invention only involve structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0165] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0166] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0167] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for selecting the best-fitting allogeneic intervertebral disc, characterized in that: include: S1: Acquire intervertebral disc image data of the target patient; S2: performing interpolation processing on the intervertebral disc image data of the target patient by using an optical flow algorithm; S3: constructing a three-dimensional model of the intervertebral disc of the target patient according to the interpolated intervertebral disc image data of the target patient; S4: acquiring the intervertebral disc volume of the target patient according to the three-dimensional model of the intervertebral disc of the target patient; S5: calculating the volume difference between each intervertebral disc sample in the allogeneic intervertebral disc sample library and the target patient; S6: Screening out intervertebral disc samples to be matched whose volume difference is less than a preset difference from the allogeneic intervertebral disc sample library; S7: calculating the matching degree between each of the intervertebral disc samples to be matched and the target patient through a deep learning network; S8: determining the intervertebral disc sample to be matched with the highest matching degree as the best matching intervertebral disc sample; Wherein, the S3 specifically includes: S301: performing image segmentation on each intervertebral disc image data after interpolation processing of the target patient to determine an intervertebral disc region image in each intervertebral disc image data; S302: stacking the intervertebral disc region images at preset intervals; S303: Connecting the edges of the intervertebral disc region image to form a volume data space; S304: Initialize virtual viewpoint; S305: Acquire a set of rays starting from the virtual viewpoint and having intersections with the volume data space; S306: performing multi-threaded sampling on rays passing through the volume data space by means of the CPU and the GPU with the goal of minimizing load imbalance; S307: Calculating the color and opacity of the sampling points intersecting with the intervertebral disc region image; S308: performing interpolation processing on the data points between two sampling points; S309: Accumulating the color and opacity of each data point in the volume data space through an optical model; S310: Constructing a three-dimensional model of the intervertebral disc of the target patient according to the accumulated results.

2. The best-fitting selection method for allogeneic intervertebral disc according to claim 1, characterized in that: The S2 specifically includes: S201: Calculate the displacement vector field estimation between adjacent intervertebral disc image data; S202: performing interpolation processing on adjacent intervertebral disc image data by using an optical flow algorithm based on the displacement vector field estimation.

3. The best-fit selection method for allogeneic intervertebral disc according to claim 2, characterized in that: The S201 specifically includes: extracting image features of intervertebral disc image data; According to the image features, K The nearest neighbor matching algorithm matches the feature points between adjacent intervertebral disc image data; Based on the matched feature points, the displacement vector field estimation between adjacent intervertebral disc image data is calculated.

4. The best-fitting selection method for allogeneic intervertebral disc according to claim 2, characterized in that: The S202 specifically includes: Based on the optical flow algorithm, determine the forward prediction image and the backward prediction image; The forward prediction image satisfies: ； in, represents the forward prediction image, ( x , y ) represents the pixel coordinates, ( v x , v y ) represents the displacement vector, α represents the interpolation coefficient, Indicates t Zhang intervertebral disc image data; The backward prediction image satisfies: ； in, represents the backward prediction image, Indicates t +1 intervertebral disc image data; According to the forward prediction image and the backward prediction image, interpolation processing is performed on adjacent intervertebral disc image data: ； in, Represents the interpolated image.

5. The best-fit selection method for allogeneic intervertebral disc according to claim 1, characterized in that: The load imbalance degree is specifically: ； in, R Indicates the load imbalance. R max Indicates the highest load on each thread, R min Indicates the lowest load on each thread.

6. The best-fitting selection method for allogeneic intervertebral disc according to claim 1, characterized in that: The volume difference is specifically: ； in, σ i Indicates i The volume difference of the intervertebral disc samples is V a represents the intervertebral disc volume of the target patient, V i Indicates i The volume of the intervertebral disc sample.

7. The best-fitting selection method for allogeneic intervertebral disc according to claim 1, characterized in that: The S7 specifically includes: S701: Construct a dual-branch twin network based on convolutional neural network; S702: extracting a first depth feature of the three-dimensional intervertebral disc model in the intervertebral disc sample to be matched by a first convolutional neural network; S703: extracting a second depth feature of the three-dimensional model of the intervertebral disc of the target patient through a second convolutional neural network; S704: Calculate the matching degree between the intervertebral disc sample to be matched and the target patient according to the first depth feature and the second depth feature: ； in, δ i Indicates i The matching degree between the intervertebral disc samples to be matched and the target patient, E 1( i ) indicates the i The first depth feature of the three-dimensional model of the intervertebral disc in the intervertebral disc sample to be matched, E 2 represents the second depth feature of the three-dimensional model of the intervertebral disc of the target patient.

8. The best-fitting selection method for allogeneic intervertebral disc according to claim 7, characterized in that: The training method of the dual-branch twin network specifically includes: Construct the inter-sample loss function of the two-branch twin network: ； in, L ( ) represents the loss function, θ represents the parameters of the two-branch twin network, y i Indicates i The matching results between the intervertebral disc samples to be matched and the target patients are When i The intervertebral disc samples to be matched are successfully matched with the target patient. When i The intervertebral disc samples to be matched failed to match the target patient, and max( ) means taking a larger value. δ 0 indicates the preset matching degree; The two-branch twin network is trained with the goal of minimizing the inter-sample loss function.

9. An optimally adapted allogeneic intervertebral disc selection system, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for selecting the best-fitting allogeneic intervertebral disc according to any one of claims 1 to 8 is implemented.

Images