Three-dimensional model generation method for personalized tibial osteotomy based on CT scan data

Abstract

The present invention relates to the technical field of constructing three-dimensional models of tibial osteotomy, and specifically relates to a method for generating a personalized three-dimensional model of tibial osteotomy based on CT scan data. This method obtains the tibial CT slices of a patient; according to the appearance and distribution of corner points in the tibial CT slices in images with different scale transformations, obtains the structural feature indexes of the corner points; according to the matching situation of corner points in the tibial CT slices of the current patient and historical patients, obtains the stable distribution degree of the corner points; according to the structural feature indexes and the stable distribution degree, obtains the type of each corner point in the tibial CT slices of the current patient, and constructs a three-dimensional model of tibial osteotomy for the current patient. The present invention constructs a three-dimensional model of tibial osteotomy on the basis of obtaining the type of each corner point in the tibial CT slices of the current patient, which is beneficial to accurately identify each part of the tibia of the current patient, avoid damaging the blood vessels in the tissue during tibial treatment, and effectively improve the effect of tibial treatment.

Description

Technical Field

[0001] The present invention relates to the technical field of constructing three-dimensional models of tibial osteotomy, and particularly relates to a method for generating a personalized three-dimensional model of tibial osteotomy based on CT scan data. Background Art

[0002] Medial Open-Wedge High Tibial Osteotomy (MOWHTO) is a surgical method for treating osteoarthritis of the medial compartment of the knee joint. During the operation, it is necessary to accurately determine the osteotomy angle and distraction distance to accurately correct tibial deformities.

[0003] In existing methods, by analyzing CT slices of the tibia and using 3D printing technology to construct a three-dimensional model of tibial osteotomy, the osteotomy angle and distraction distance are obtained, thereby improving the surgical accuracy and effectively shortening the operation time. However, during the treatment of the tibia, it is also necessary to ensure avoiding important blood vessels in the surrounding tissues of the tibia to prevent a large amount of bleeding during the operation, which may affect the progress of the operation. However, in existing methods, only a three-dimensional model of tibial osteotomy is constructed, without considering the blood vessels in the surrounding tissues of the tibia, which easily leads to damage to the blood vessels in the tissues by the set osteotomy angle and distraction distance of the tibia, affecting the treatment of the tibia. Summary of the Invention

[0004] In order to solve the technical problem that, due to the failure to consider the blood vessels in the surrounding tissues of the tibia, the set osteotomy angle and distraction distance of the tibia damage the blood vessels in the tissues, thereby affecting the treatment of the tibia, the purpose of the present invention is to provide a method for generating a personalized three-dimensional model of tibial osteotomy based on CT scan data, and the specific technical solution adopted is as follows:

[0005] An embodiment of the present invention provides a method for generating a personalized three-dimensional model of tibial osteotomy based on CT scan data, and the method includes the following steps:

[0006] Obtain tibial CT slices of a preset number of historical patients and the current patient;

[0007] Perform scale transformation on each tibial CT slice through a Gaussian pyramid, and obtain the structural feature index of each corner point in each tibial CT slice according to the appearance and distribution of each corner point in the tibial CT slice in different scale-transformed images;

[0008] According to the difference in the structural feature indexes of the matching corner points in the tibial CT slices of the current patient and each historical patient in different scale-transformed images and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient, obtain the stable distribution degree of each corner point in each tibial CT slice of the current patient;

[0009] Based on the structural feature index and stable distribution degree of each corner point in each tibial CT slice of the current patient, obtain the type of each corner point in each tibial CT slice of the current patient, and construct a three-dimensional tibial osteotomy model of the current patient;

[0010] The method for obtaining the structural feature index is as follows:

[0011] For any tibial CT slice and any corner point in this tibial CT slice, obtain the prominence of this corner point according to the appearance of this corner point in different scale-transformed images corresponding to this tibial CT slice;

[0012] Obtain the distribution characteristic value of this corner point according to the distribution of this corner point in different scale-transformed images corresponding to this tibial CT slice;

[0013] Take the product of the prominence of this corner point and the distribution characteristic value as the structural feature index of this corner point;

[0014] The method for obtaining the stable distribution degree is as follows:

[0015] Obtain the matching degree of each tibial CT slice of the current patient and each historical patient according to the difference in the structural feature index of the matching corner points in the tibial CT slices of the current patient and each historical patient in different scale-transformed images;

[0016] Number each tibial CT slice of each patient in the same specified order; among them, each tibial CT slice of a patient corresponds to a unique number;

[0017] For any corner point in any numbered tibial CT slice of the current patient, obtain the corner points that match this corner point in the same numbered tibial CT slices of all historical patients as reference corner points;

[0018] Take the cumulative result of the matching degree between the current patient and the numbered tibial CT slices of the historical patients corresponding to each reference corner point as the stable analysis value of this corner point;

[0019] Take the result of normalizing the ratio of the stable analysis value to the total number of all historical patients as the stable distribution degree of this corner point.

[0020] Furthermore, the method for obtaining the prominence is as follows:

[0021] Obtain the number of images in which this corner point exists in different scale-transformed images corresponding to this tibial CT slice as the first number;

[0022] Take the ratio of the first number to the total number of all different scale-transformed images corresponding to this tibial CT slice as the prominence of this corner point.

[0023] Further, the method for obtaining the distribution eigenvalue is as follows:

[0024] For any scale image in the differently scaled transformed images corresponding to the tibia CT slice, use the corner point corresponding to this corner point in this scale image as the target corner point, and with the target corner point as the center and a preset length as the radius, construct the local area of the target corner point;

[0025] Take the ratio of the number of corner points existing in the local area to the area of the local area as the density distribution degree of this corner point in this scale image;

[0026] For any edge line passing through the target corner point, take the ratio of the number of corner points on this edge line to the number of all pixel points on this edge line as the first eigenvalue of this edge line;

[0027] Take the summation result of the first eigenvalues of all edge lines passing through the target corner point as the position analysis value of this corner point in this scale image;

[0028] Take the product of the density distribution degree of this corner point in this scale image and the position analysis value as the overall reference value of this corner point in this scale image;

[0029] Take the result of negative correlation and normalization after adding the overall reference values of this corner point in all differently scaled transformed images corresponding to the tibia CT slice as the distribution eigenvalue of this corner point.

[0030] Further, the method for obtaining the matching degree is as follows:

[0031] For any historical patient and any label, use the tibia CT slice with the said any label of the current patient as the first image, and use the tibia CT slice with the said any label of this historical patient as the second image;

[0032] For any scale, use all the corner points of the corner points in the first image that exist in the image corresponding to the first image after this scale transformation as the first corner points, and use all the corner points of the corner points in the second image that exist in the image corresponding to the second image after this scale transformation as the second corner points;

[0033] Take all the matching pairs obtained by matching the first corner points and the second corner points as the reference matching pairs;

[0034] For any reference matching pair, take the result of negative correlation and normalization of the difference in the structural feature indexes of the two corner points in this reference matching pair as the matching analysis value of this reference matching pair;

[0035] Take the summation result of the matching analysis values of all reference matching pairs as the reference matching value of the tibia CT slices with the said any label of the current patient and this historical patient under this scale transformation;

[0036] Obtain the reference matching value of any labeled tibial CT slice of the current patient and the historical patient under each different scale transformation, and take the maximum reference matching value as the matching degree of any labeled tibial CT slice of the current patient and the historical patient.

[0037] Further, the method for obtaining the type of each corner point in each tibial CT slice of the current patient is as follows:

[0038] For any corner point in any tibial CT slice of the current patient, when the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the normal part of the tibial contour;

[0039] When the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is the abnormal diseased part of the tibial contour;

[0040] When the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the blood vessel in the tissue;

[0041] When the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is the interference point.

[0042] Further, the method for constructing the three-dimensional tibial osteotomy model of the current patient is as follows:

[0043] Match the corner points in different tibial CT slices of the current patient according to the type of the corner points, and use the matched corner points as the center points of the quadtree to divide the voxel blocks to construct the three-dimensional tibial osteotomy model of the current patient.

[0044] Further, the method for obtaining the corner points is: obtain the corner points in the tibial CT slice through the SIFT corner detection algorithm.

[0045] The present invention has the following beneficial effects:

[0046] The present invention first performs scale transformation on each tibia CT slice through a Gaussian pyramid, which is beneficial for accurately analyzing the characteristics of each corner point in the tibia CT slice. Furthermore, according to the appearance and distribution of each corner point in each tibia CT slice in images with different scale transformations, the structural feature index of each corner point in each tibia CT slice is obtained, accurately reflecting the performance characteristics of each corner point, which is beneficial for subsequently accurately obtaining the type of each corner point in the tibia CT slice of the current patient. In order to more accurately obtain the type of each corner point in the tibia CT slice of the current patient and make each part of the three-dimensional tibial osteotomy model constructed for the current patient more complete and accurate, further based on the differences in the structural feature indexes of the matching corner points in the tibia CT slices of the current patient and each historical patient in images with different scale transformations and the matching situation of the corner points in the tibia CT slices of the current patient and each historical patient, the stable distribution degree of each corner point in each tibia CT slice of the current patient is obtained, accurately reflecting the matching situation of each corner point in each tibia CT slice of the current patient with the corner points in the tibia CT slices of historical patients, and further showing the characteristics of each corner point in each tibia CT slice of the current patient. Then, according to the structural feature index and stable distribution degree of each corner point in each tibia CT slice of the current patient, the type of each corner point in each tibia CT slice of the current patient is accurately obtained, the various parts of the tibia of the current patient are accurately determined, and then the three-dimensional tibial osteotomy model of the current patient is accurately constructed. At the same time, the blood vessels in the surrounding tissues of the tibia are accurately displayed, avoiding damage to the blood vessels in the tissues during the treatment of the tibia of the current patient, effectively improving the treatment effect of the tibia of the current patient, and ensuring the stability of the tibial treatment surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0048] Figure 1 It is a schematic flowchart of a method for generating a personalized three-dimensional tibial osteotomy model based on CT scan data provided by an embodiment of the present invention;

[0049] Figure 2 It is a flowchart of a method for obtaining a structural feature index provided by an embodiment of the present invention;

[0050] Figure 3 It is a flowchart of a method for obtaining a stable distribution degree provided by an embodiment of the present invention;

[0051] Figure 4Structural diagram of a personalized tibial osteotomy three-dimensional model generation system provided by an embodiment of the present invention;

[0052] Figure 5 Schematic diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0053] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following combines the accompanying drawings and preferred embodiments to detail the specific implementation manners, structures, features, and effects of the personalized tibial osteotomy three-dimensional model generation method based on CT scan data proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0055] The following specifically describes the specific solution of the personalized tibial osteotomy three-dimensional model generation method provided by the present invention with reference to the accompanying drawings.

[0056] Embodiment 1:

[0057] The present invention proposes a personalized tibial osteotomy three-dimensional model generation method based on CT scan data. Please refer to Figure 1 , which shows a schematic flowchart of a personalized tibial osteotomy three-dimensional model generation method provided by an embodiment of the present invention. The method includes the following steps:

[0058] Step S1: Obtain a preset number of tibial CT slices of historical patients and the current patient.

[0059] Specifically, in order to accurately construct the tibial osteotomy three-dimensional model of the current patient, and considering the blood vessels in the surrounding tissues of the tibia at the same time, to avoid causing significant damage to the blood vessels in the tissues by subsequently obtaining the osteotomy angle and distraction distance, resulting in a large amount of bleeding and affecting the treatment of the tibia. Therefore, the blood vessels in the surrounding tissues of the tibia should also be accurately displayed in the tibial osteotomy three-dimensional model constructed in this embodiment. It is known that the structural distributions of the human tibia are similar. Therefore, in this embodiment, the tibial structures of the current patient and historical patients are compared and analyzed to accurately identify the blood vessels in the surrounding tissues of the current patient's tibia, so as to avoid damaging the blood vessels during the surgical treatment of the current patient's tibia.

[0060] In order to accurately construct a three-dimensional model of the tibial osteotomy for the current patient, in this embodiment, a preset number of historical patients' and the current patient's tibial CT slices are obtained from the hospital database. Among them, the preset number is set to 1000 in this embodiment, and the implementer can set the size of the preset number according to the actual situation, which is not limited here. It should be noted that the tibial CT slice is a slice of the tibial cross-section. Therefore, the outer contour of the tibia in the tibial CT slice generally appears as a circle.

[0061] Step S2: Perform scale transformation on each tibial CT slice through the Gaussian pyramid, and obtain the structural feature index of each corner point in each tibial CT slice according to the appearance and distribution of each corner point in the tibial CT slice in different scale-transformed images.

[0062] It is known that when constructing a three-dimensional model of the tibial osteotomy for the current patient, it is necessary to align the tibial CT slices of the current patient. After alignment, the corner points in each tibial CT slice can be obtained through the SIFT (Scale-Invariant Feature Transform) corner detection algorithm for matching, and then the voxel blocks can be divided through the matched corner points, and finally a three-dimensional model of the tibial osteotomy for the current patient is constructed. Among them, the SIFT (Scale-Invariant Feature Transform) corner detection algorithm is a well-known technology and will not be elaborated here. However, the abnormal parts of the tibia and the blood vessels in the tissues around the tibia cannot be determined in the three-dimensional model of the tibial osteotomy for the current patient constructed by the above method. Therefore, in this embodiment, it is necessary to analyze each corner point in the tibial CT image of the current patient to determine the type of each corner point, so as to accurately and completely construct the three-dimensional model of the tibial osteotomy for the current patient in the follow-up.

[0063] It is known that in a tibia CT slice, the tibia contour is the most obvious, and the corner points on the edge line corresponding to the tibia contour are sparsely distributed. At the same time, the edge line corresponding to the tibia contour is relatively long. Inside the tibia in the tibia CT slice, it is a complex honeycomb shape, and the corner points are relatively densely distributed, but the edge line where the corner points are located is relatively short. In the tibia CT slice, the corner points corresponding to the blood vessels in the tissue are distributed disorderly, and the edge line where the corner points are located is also relatively short. In order to accurately identify the type of each corner point in the tibia CT slice, in this embodiment, first, a Gaussian pyramid is used to perform scale transformation on each tibia CT slice, so as to more accurately analyze the characteristics of each corner point subsequently. When a certain corner point in a certain tibia CT slice appears more times in the different scale transformation images of this tibia CT slice, it indicates that this corner point is more obvious in this tibia CT slice, and this corner point is more likely to be a corresponding point of the tibia contour. At the same time, when the local corner point density distribution of the corresponding corner points of this corner point in the different scale transformation images of this tibia CT slice is smaller, and the number of corner points on the edge line corresponding to the corner points is less and the edge line is shorter, it also indicates that this corner point is more likely to be a corresponding point of the tibia contour. Among them, the Gaussian pyramid is a well-known technology and will not be elaborated here. It should be noted that in this embodiment, the scale factor k in the Gaussian pyramid is set to 1.5 and the parameter sigma is set to 1.6. Implementers can set the size of the scale factor k and the parameter sigma according to the actual situation, and no limitation is made here. Therefore, in this embodiment, according to the appearance and distribution of each corner point in each tibia CT slice in different scale transformation images, the structural feature index of each corner point in each tibia CT slice is obtained. Among them, the larger the structural feature index, the more likely the corresponding corner point is to be a corresponding point of the tibia contour.

[0064] Preferably, in a feasible implementation manner of this embodiment, for the method of obtaining the structural feature index, please refer to Figure 2 , which shows a flowchart of a method for obtaining a structural feature index provided by this embodiment. The method includes the following steps:

[0065] Step S201: For any tibia CT slice and any corner point in this tibia CT slice, according to the appearance of this corner point in the different scale transformation images corresponding to this tibia CT slice, obtain the obviousness of this corner point.

[0066] It is known that when the gray value of a certain corner point in a certain tibia CT slice is larger, the number of times this corner point appears in the different scale transformation images of this tibia CT slice is more. Among them, the corner point with a larger gray value is more obvious and more likely to be a corresponding point of the tibia contour. Therefore, in this embodiment, according to the appearance of this corner point in the different scale transformation images corresponding to this tibia CT slice, the obviousness of this corner point is obtained. The greater the obviousness, the more likely this corner point is to be a corresponding point of the tibia contour.

[0067] Preferably, in an implementable manner of this embodiment, the method for obtaining the obviousness degree is as follows: obtaining the number of images in which the corner point exists in the differently scaled transformed images corresponding to the tibial CT slice as the first number; taking the ratio of the first number to the total number of all differently scaled transformed images corresponding to the tibial CT slice as the obviousness degree of the corner point.

[0068] Step S202: Obtain the distribution characteristic value of the corner point according to the distribution condition of the corner point in the differently scaled transformed images corresponding to the tibial CT slice.

[0069] During the process of performing different scale transformations on the tibial CT slice through the Gaussian pyramid, there is a situation where some corner points disappear. Therefore, in this embodiment, only the scale images in which the corner point exists in the differently scaled transformed images corresponding to the tibial CT slice are analyzed. By analyzing the distribution condition of the corner point in the corresponding scale images where it exists, the distribution characteristic value of the corner point is obtained. Among them, the larger the distribution characteristic value, the more likely the corner point is the corresponding point of the tibial contour.

[0070] Preferably, in an implementable manner of this embodiment, the method for obtaining the distribution characteristic value is as follows: for any scale image in the differently scaled transformed images corresponding to the tibial CT slice, taking the corner point corresponding to the corner point in this scale image as the target corner point (the target corner point is the corner point that exists in this scale image of the corner point), and constructing a local area of the target corner point with the preset length as the radius centered on the target corner point; in this embodiment, the preset length is set to 10 pixels, and the implementer can set the size of the preset length according to the actual situation, which is not limited here. It should be noted that only the area located within this scale image is analyzed for the local area of the target corner point. Taking the ratio of the number of corner points existing in the local area to the area of the local area as the density distribution degree of the corner point in this scale image; the smaller the density distribution degree, the more likely the corner point is the corresponding point of the tibial contour because the corner points on the tibial contour are sparsely distributed;

[0071] In order to more accurately analyze the characteristics of the corner point, and further analyze the edge line where the target corner point is located. For any edge line passing through the target corner point, taking the ratio of the number of corner points on this edge line to the number of all pixel points on this edge line as the first characteristic value of this edge line; the smaller the first characteristic value, the more likely this edge line is the edge line corresponding to the tibial contour. For overall analysis, further taking the summation result of the first characteristic values of all edge lines passing through the target corner point as the position analysis value of the corner point in this scale image; the smaller the position analysis value, the more likely the corner point is the corresponding point of the tibial contour;

[0072] Furthermore, the product of the density distribution degree of the corner point in the scale image and the position analysis value is used as the overall reference value of the corner point in the scale image; the smaller the overall reference value, the more likely the corner point is the corresponding point of the tibia contour; in order to more accurately analyze the possibility that the corner point is the corresponding point of the tibia contour, the result of negative correlation and normalization after adding the overall reference values of the corner point in all different scale transformation images corresponding to the tibia CT slice is used as the distribution feature value of the corner point.

[0073] Among them, the calculation formula of the distribution feature value is: ; in the formula, is the distribution feature value of the i-th corner point in the a-th tibia CT slice; N is the total number of different scale transformation images corresponding to the a-th tibia CT slice; is the number of corner points existing in the local area corresponding to the target corner point of the i-th corner point in the n-th scale image of the a-th tibia CT slice; is the area of the local area corresponding to the target corner point of the i-th corner point in the n-th scale image of the a-th tibia CT slice; is the density distribution degree of the i-th corner point in the n-th scale image of the a-th tibia CT slice; is the number of edge lines passed by the i-th corner point in the n-th scale image of the a-th tibia CT slice corresponding to the target corner point; is the number of corner points on the m-th edge line passed by the i-th corner point in the n-th scale image of the a-th tibia CT slice corresponding to the target corner point; is the number of all pixel points on the m-th edge line passed by the i-th corner point in the n-th scale image of the a-th tibia CT slice corresponding to the target corner point; is the first eigenvalue of the m-th edge line passed by the i-th corner point in the n-th scale image of the a-th tibia CT slice corresponding to the target corner point; is the position analysis value of the i-th corner point in the n-th scale image of the a-th tibia CT slice; is the overall reference value of the i-th corner point in the n-th scale image of the a-th tibia CT slice; exp is the exponential function with the natural constant as the base.

[0074] It should be noted that if the corner point does not exist in a certain scale transformation image corresponding to the tibia CT slice, it is default that the overall reference value of the corner point in this scale image is 0.

[0075] Step S203: The product of the obviousness of the corner point and the distribution feature value is used as the structural feature index of the corner point.

[0076] The greater the known distinctness and the greater the distribution characteristic value, both can indicate that the corresponding corner point is more likely to be the corresponding point of the tibia contour. Furthermore, in this embodiment, the product of the distinctness of the corner point and the distribution characteristic value is used as the structural feature index of the corner point.

[0077] Thus far, the structural feature index of each corner point is obtained.

[0078] Step S3: According to the differences in the structural feature indices of the corner points matched in the different-scale transformed images of the tibial CT slices of the current patient and each historical patient, and the matching conditions of the corner points in the tibial CT slices of the current patient and each historical patient, obtain the stable distribution degree of each corner point in each tibial CT slice of the current patient.

[0079] Specifically, by analyzing the structural feature indices of each corner point in each tibial CT slice of the current patient, it can only be evaluated that the corner point corresponding to the larger structural feature index is more likely to be the corresponding point of the tibia contour, and the corner point corresponding to the smaller structural feature index is more likely to be the corresponding point of the blood vessel in the external tissue of the tibia, the internal corresponding point of the tibia, or the noise corresponding point. It is known that the abnormal parts of the tibia are reflected on the tibia contour. At this time, the structural feature indices corresponding to the corner points of the normal parts and the abnormal parts in the tibia contour are both larger. Therefore, the abnormal parts of the tibia of the current patient cannot be accurately identified through the structural feature indices of the corner points; at the same time, in order to accurately identify the corner points corresponding to the blood vessels in the external tissue of the tibia, it is also necessary to avoid the interference of the noise corresponding corner points and the internal corner points of the tibia.

[0080] Considering that the characteristics of the corner points corresponding to the blood vessels in the surrounding tissues of the tibia of the current patient are similar to those of the corner points corresponding to the blood vessels in the surrounding tissues of the tibia of a large number of historical patients because the human body structure distribution is similar; at the same time, the characteristics of the corner points of the normal parts in the tibia contour of the current patient are also similar to those of the corner points of the normal parts in the tibia contours of a large number of historical patients; because there must be obvious differences between the abnormal parts in the tibia contour of the current patient and the tibia contours of a large number of historical patients, and at the same time, due to the diversity of the honeycomb structure inside the tibia, there are also obvious differences in the characteristics of the internal corner points of the tibia of the current patient and the internal corner points of the tibia of historical patients; it is known that noise is random, and thus there are obvious differences between the noise corresponding corner points in the tibial CT slices of the current patient and the corner points in the tibial CT slices of historical patients.

[0081] Therefore, in this embodiment, first, based on the differences in the structural feature indexes of the corner points matched in the tibial CT slices of the current patient and each historical patient in different scale-transformed images, the overall matching situation of each tibial CT slice of the current patient and each historical patient is analyzed. Then, according to the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient, the stable distribution situation of each corner point in each tibial CT slice of the current patient is determined, indirectly indicating the matching situation of each corner point in the tibial CT slice of the current patient with the corner points in the tibial CT slice of the historical patient, which is beneficial to accurately distinguishing the types of each corner point in the tibial CT slice of the current patient in the subsequent process. Therefore, in this embodiment, based on the differences in the structural feature indexes of the corner points matched in the tibial CT slices of the current patient and each historical patient in different scale-transformed images and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient, the stable distribution degree of each corner point in each tibial CT slice of the current patient is obtained. Among them, the greater the stable distribution degree, the more matched the corresponding corner point in the tibial CT slice of the current patient is with the corner point in the tibial CT slice of the historical patient, and the more likely the corresponding corner point in the tibial CT slice of the current patient is the corresponding point of the normal part in the tibial contour of the current patient or the corresponding point of the blood vessel in the tissue.

[0082] Preferably, in an implementable manner of this embodiment, for the method of obtaining the stable distribution degree, please refer to Figure 3 , which shows a flowchart of a method for obtaining the stable distribution degree provided in this embodiment. The method includes the following steps:

[0083] Step S301: Based on the differences in the structural feature indexes of the corner points matched in the tibial CT slices of the current patient and each historical patient in different scale-transformed images, obtain the matching degree of each tibial CT slice of the current patient and each historical patient.

[0084] The greater the matching degree, the more consistent the corner point distribution characteristics in the corresponding tibial CT slice of the current patient and the corresponding historical patient are. Among them, corner point matching is a well-known technology and will not be elaborated here.

[0085] Preferably, in an implementable manner of this embodiment, the method for obtaining the matching degree is as follows: Each tibial CT slice of each patient is numbered in the same specified order; wherein, each tibial CT slice of a patient corresponds to a unique number; in this embodiment, the specified order is set from top to bottom, and the numbering order is 1, 2, 3... The implementer can set the specified order, numbering, and numbering order according to the actual situation, which is not limited herein. For any historical patient and any number, the tibial CT slice of the current patient with this number is used as the first image, and the tibial CT slice of this historical patient with this number is used as the second image; for any scale, the corner points in the first image that exist in the image corresponding to the first image after this scale transformation are used as the first corner points, and the corner points in the second image that exist in the image corresponding to the second image after this scale transformation are used as the second corner points; the matching pairs obtained by matching the first corner points with the second corner points are all used as reference matching pairs; wherein, a reference matching pair contains two corner points, namely a first corner point and a second corner point; for any reference matching pair, the result of negatively correlating and normalizing the absolute value of the difference between the structural feature indexes of the two corner points in this reference matching pair is used as the matching analysis value of this reference matching pair; the larger the matching analysis value, the greater the matching degree between the two corner points in this reference matching pair. It should be noted that in this embodiment, performs negative correlation and normalization processing on the absolute value of the difference between the structural feature indexes of the two corner points in this reference matching pair, where exp is the exponential function with the natural constant as the base, and X represents the absolute value of the difference between the structural feature indexes of the two corner points in this reference matching pair;

[0086] In order to analyze the matching situation between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number under this scale transformation, and then the sum result of the matching analysis values of all reference matching pairs is used as the reference matching value between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number under this scale transformation; the larger the reference matching value, the better the matching effect between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number under this scale transformation; in order to obtain the matching situation between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number, and then obtain the reference matching values between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number under each different scale transformation, the largest reference matching value is used as the matching degree between the tibial CT slice of the current patient and the tibial CT slice of this historical patient with this number.

[0087] Step S302: Obtain the stable distribution degree of each corner point in each tibial CT slice of the current patient according to the matching degree and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient.

[0088] In order to accurately analyze whether each corner point in each tibial CT slice of the current patient corresponds to a corner point of a stable distribution structure in the human body, in this embodiment, the stability distribution degree of each corner point in each tibial CT slice of the current patient is obtained according to the matching degree and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient. The greater the stability distribution degree, the more likely the corresponding corner point is a corner point corresponding to a normal part of the tibial contour or a blood vessel in the tissue.

[0089] Preferably, in an implementable manner of this embodiment, the method for obtaining the stability distribution degree is as follows: for any corner point in the tibial CT slice with this label of the current patient, the corner points matched by this corner point in the tibial CT slices with this label of all historical patients are all used as reference corner points; it should be noted that there may not be reference corner points in the tibial CT slices with this label of all historical patients. In order to accurately analyze the stability distribution of this corner point, the cumulative result of the matching degree between the current patient and the tibial CT slices with this label of the historical patients corresponding to each reference corner point is used as the stability analysis value of this corner point; then, the result of normalizing the ratio of the stability analysis value to the total number of all historical patients is used as the stability distribution degree of this corner point. It should be noted that in this embodiment, the ratio of the stability analysis value to the total number of all historical patients is normalized by the norm normalization function.

[0090] Thus, the stability distribution degree of each corner point in each tibial CT slice of the current patient is accurately obtained.

[0091] Step S4: According to the structural feature index and the stability distribution degree of each corner point in each tibial CT slice of the current patient, obtain the type of each corner point in each tibial CT slice of the current patient, and construct a three-dimensional tibial osteotomy model of the current patient.

[0092] It is known that the greater the structural feature index, the more likely the corresponding corner point is a point corresponding to the tibial contour; the smaller the structural feature index, the more likely the corresponding corner point is a blood vessel corresponding point in the external tissue of the tibia, an internal corresponding point of the tibia, or a noise corresponding point; the greater the stability distribution degree, the more likely the corresponding corner point is a point corresponding to a normal part of the tibial contour or a blood vessel in the tissue; the smaller the stability distribution degree, the more likely the corresponding corner point is a point corresponding to an abnormal part of the tibial contour, an internal corresponding point of the tibia, or a noise corresponding point. Therefore, in this embodiment, the type of each corner point in each tibial CT slice of the current patient is obtained according to the structural feature index and the stability distribution degree of each corner point in each tibial CT slice of the current patient.

[0093] Preferably, in an implementable manner of this embodiment, the method for obtaining the type of each corner point in each tibial CT slice of the current patient is as follows: for any corner point in any tibial CT slice of the current patient, when the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the normal part of the tibial contour; when the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is the diseased abnormal part of the tibial contour; when the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the blood vessel in the tissue; when the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is an interference point, where the interference point includes the corner point corresponding to the noise and the corner point inside the tibia, and the interference point is not considered when constructing the three-dimensional tibial osteotomy model of the current patient. In this embodiment, both the preset structural feature index threshold and the preset stable distribution degree threshold are set to 0.6. The implementer can set the sizes of the preset structural feature index threshold and the preset stable distribution degree threshold according to the actual situation, and no limitation is made here.

[0094] After determining the type of each corner point in each tibial CT slice of the current patient, the corner points in different tibial CT slices of the current patient are matched according to the type of the corner point to obtain the diseased abnormal part information in the tibia of the current patient and the blood vessel information in the tissue, and then the matched corner points are used as the center points of the quadtree for voxel block division to accurately construct the three-dimensional tibial osteotomy model of the current patient. The constructed three-dimensional tibial osteotomy model includes the blood vessels in the tissue around the tibia, so as to avoid the blood vessels in the tissue during the subsequent treatment of the tibia of the current patient, ensure the stable progress of the tibial treatment operation of the current patient, and improve the treatment effect of the tibia. Among them, the quadtree is a well-known technology and will not be elaborated here.

[0095] In summary, this embodiment obtains the tibial CT slices of the patient; obtains the structural feature index of the corner point according to the appearance and distribution of the corner point in the different-scale transformed images of the tibial CT slice; obtains the stable distribution degree of the corner point according to the matching situation of the corner points in the tibial CT slices of the current patient and the historical patients; obtains the type of each corner point in the tibial CT slice of the current patient according to the structural feature index and the stable distribution degree, and constructs the three-dimensional tibial osteotomy model of the current patient. The present invention constructs the three-dimensional tibial osteotomy model on the basis of obtaining the type of each corner point in the tibial CT slice of the current patient, which is beneficial to accurately identify each part of the tibia of the current patient, avoid damaging the blood vessels in the tissue during the tibial treatment process, and effectively improve the treatment effect of the tibia.

[0096] Example 2:

[0097] The present invention also provides a personalized tibial osteotomy three-dimensional model generation system based on CT scan data. Please refer to Figure 4 , which shows the structural diagram of a personalized tibial osteotomy three-dimensional model generation system provided by an embodiment of the present invention. The system includes: a tibial CT slice acquisition module 10, a structural feature index acquisition module 20, a stable distribution degree acquisition module 30, and a tibial osteotomy three-dimensional model construction module 40.

[0098] The tibial CT slice acquisition module 10 is used to acquire tibial CT slices of a preset number of historical patients and the current patient.

[0099] The structural feature index acquisition module 20 is used to perform scale transformation on each tibial CT slice through a Gaussian pyramid, and obtain the structural feature index of each corner point in each tibial CT slice according to the appearance and distribution of each corner point in different scale-transformed images.

[0100] The stable distribution degree acquisition module 30 is used to obtain the stable distribution degree of each corner point in each tibial CT slice of the current patient according to the difference in the structural feature indexes of the matching corner points in the tibial CT slices of the current patient and each historical patient in different scale-transformed images, and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient.

[0101] The tibial osteotomy three-dimensional model construction module 40 is used to obtain the type of each corner point in each tibial CT slice of the current patient according to the structural feature index and stable distribution degree of each corner point in each tibial CT slice of the current patient, and construct the tibial osteotomy three-dimensional model of the current patient.

[0102] It should be noted that: for the system provided in the above embodiment, only the above-mentioned division of each functional module is used for illustration. In actual application, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the personalized tibial osteotomy three-dimensional model generation system based on CT scan data provided in the above embodiment and the embodiment of the personalized tibial osteotomy three-dimensional model generation method based on CT scan data belong to the same concept. For the specific implementation process, please refer to the method embodiment, which will not be elaborated here.

[0103] Example 3:

[0104] The present invention also provides a personalized tibial osteotomy three-dimensional model generation device based on CT scan data. The device includes a memory and a processor. Among them, the memory stores executable program code, and the processor is used to call and execute the executable program code to execute a personalized tibial osteotomy three-dimensional model generation method provided by an embodiment of the present application. The device may specifically be a chip, a component or a module. The chip may include a connected processor and a memory. Among them, the memory is used to store instructions. When the processor calls and executes the instructions, the chip can execute a personalized tibial osteotomy three-dimensional model generation method provided by the above embodiment.

[0105] In addition, an embodiment of the present application also protects a computer device. Please refer to Figure 5 , the computer device includes a memory 401, a processor 402, and a computer program 403 stored in the memory 401 and running on the processor 402. Among them, when the processor 402 executes the computer program 403, the computer device can execute any one of the personalized tibial osteotomy three-dimensional model generation methods introduced above.

[0106] Example 4:

[0107] This embodiment also provides a computer-readable storage medium. Computer program code is stored in the computer-readable storage medium. When the computer program code runs on a computer, the computer is enabled to execute the above-related method steps to implement a personalized tibial osteotomy three-dimensional model generation method provided by the above embodiment.

[0108] Example 5:

[0109] This embodiment also provides a computer program product. When the computer program product runs on a computer, the computer is enabled to execute the above-related steps to implement a personalized tibial osteotomy three-dimensional model generation method provided by the above embodiment.

[0110] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding method provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method provided above, and will not be elaborated here.

[0111] It should be noted that the above-mentioned order of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the differences between each embodiment and other embodiments are emphasized.

Claims

1. A method for generating a personalized three-dimensional model of tibial osteotomy based on CT scan data, characterized in that The method includes the following steps: Obtain tibial CT slices of a preset number of historical patients and the current patient; Perform scale transformation on each tibial CT slice through a Gaussian pyramid, and obtain the structural feature index of each corner point in each tibial CT slice according to the appearance and distribution of each corner point in different scale-transformed images of each tibial CT slice; Obtain the stable distribution degree of each corner point in each tibial CT slice of the current patient according to the difference in the structural feature indexes of the matching corner points in the tibial CT slices of the current patient and each historical patient in different scale-transformed images and the matching situation of the corner points in the tibial CT slices of the current patient and each historical patient; Obtain the type of each corner point in each tibial CT slice of the current patient according to the structural feature index and stable distribution degree of each corner point in each tibial CT slice of the current patient, and construct a three-dimensional tibial osteotomy model of the current patient; The method for obtaining the structural feature index is: For any tibial CT slice and any corner point in the tibial CT slice, obtain the distinctness of the corner point according to the appearance of the corner point in different scale-transformed images corresponding to the tibial CT slice; Obtain the distribution characteristic value of the corner point according to the distribution of the corner point in different scale-transformed images corresponding to the tibial CT slice; Take the product of the distinctness of the corner point and the distribution characteristic value as the structural feature index of the corner point; The method for obtaining the stable distribution degree is: Obtain the matching degree of each tibial CT slice of the current patient and each historical patient according to the difference in the structural feature indexes of the matching corner points in the tibial CT slices of the current patient and each historical patient in different scale-transformed images; Number each tibial CT slice of each patient in the same specified order; where each tibial CT slice of a patient corresponds to a unique number; For any corner point in any numbered tibial CT slice of the current patient, obtain the corner points that match the corner point in the same numbered tibial CT slices of all historical patients as reference corner points; Take the cumulative result of the matching degrees of the current patient and the corresponding historical patients of each reference corner point in the same numbered tibial CT slices as the stable analysis value of the corner point; Take the normalized result of the ratio of the stable analysis value to the total number of all historical patients as the stable distribution degree of the corner point.

2. The personalized tibial osteotomy three-dimensional model generation method based on CT scan data according to claim 1, characterized in that The method for obtaining the distinctness is: Obtain the number of images in which the corner point exists in different scale-transformed images corresponding to the tibial CT slice as the first number; Take the ratio of the first number to the total number of all different scale-transformed images corresponding to the tibial CT slice as the distinctness of the corner point.

3. The personalized tibial osteotomy three-dimensional model generation method based on CT scan data according to claim 1, characterized in that, The method for obtaining the distribution characteristic value is: For any scale image in different scale-transformed images corresponding to the tibial CT slice, take the corner point corresponding to the corner point in the scale image as the target corner point, and construct a local area of the target corner point with the preset length as the radius centered on the target corner point; Take the ratio of the number of corner points in the local area to the area of the local area as the density distribution degree of the corner point in the scale image; For any edge line passing through the target corner point, the ratio of the number of corner points on the edge line to the number of all pixel points on the edge line is used as the first eigenvalue of the edge line; The summation result of the first eigenvalues of all edge lines passing through the target corner point is used as the position analysis value of the corner point in the image at this scale; The product of the density distribution degree of the corner point in the image at this scale and the position analysis value is used as the overall reference value of the corner point in the image at this scale; The result of negative correlation and normalization after summing the overall reference values of the corner point in all different scale transformation images corresponding to the tibia CT slice is used as the distribution characteristic value of the corner point.

4. A personalized tibial osteotomy three-dimensional model generation method based on CT scan data according to claim 1, characterized in that The method for obtaining the matching degree is as follows: For any historical patient and any label, the tibia CT slice with the any label of the current patient is used as the first image, and the tibia CT slice with the any label of the historical patient is used as the second image; For any scale, the corner points in the first image that exist in the image corresponding to the first image after passing through this scale transformation are all used as the first corner points, and the corner points in the second image that exist in the image corresponding to the second image after passing through this scale transformation are all used as the second corner points; The matching pairs obtained by matching the first corner points and the second corner points are all used as reference matching pairs; For any reference matching pair, the result of negative correlation and normalization of the difference in the structural feature indexes of the two corner points in the reference matching pair is used as the matching analysis value of the reference matching pair; The summation result of the matching analysis values of all reference matching pairs is used as the reference matching value of the tibia CT slices with the any label of the current patient and the historical patient under this scale transformation; Obtain the reference matching values of the tibia CT slices with the any label of the current patient and the historical patient under each different scale transformation, and use the maximum reference matching value as the matching degree of the tibia CT slices with the any label of the current patient and the historical patient.

5. A method for generating a personalized tibial osteotomy three-dimensional model based on CT scan data according to claim 1, characterized in that The method for obtaining the type of each corner point in each tibia CT slice of the current patient is as follows: For any corner point in any tibia CT slice of the current patient, when the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the normal part of the tibia contour; When the structural feature index of the corner point is greater than the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is the abnormal part of the tibia contour with disease; When the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is greater than the preset stable distribution degree threshold, the type of the corner point is the blood vessel in the tissue; When the structural feature index of the corner point is less than or equal to the preset structural feature index threshold and the stable distribution degree is less than or equal to the preset stable distribution degree threshold, the type of the corner point is the interference point.

6. The personalized tibial osteotomy three-dimensional model generation method based on CT scan data according to claim 1, wherein, The method for constructing the three-dimensional model of tibial osteotomy of the current patient is as follows: Match the corner points in different tibial CT slices of the current patient according to the types of corner points, and use the matched corner points as the center points of the quadtree to divide the voxel blocks, so as to construct a three-dimensional model of tibial osteotomy for the current patient.

7. The personalized tibial osteotomy three-dimensional model generation method based on CT scan data according to claim 1, wherein, The method for obtaining the corner points is as follows: obtain the corner points in the tibial CT slices through the SIFT corner detection algorithm.

Images