A method and system for segmenting the fibular head and tibia head in DR knee joint anteroposterior radiographs for quality control

Abstract

The present invention discloses a method and system for segmenting the fibula and tibia in a DR knee joint anteroposterior film in quality control, which relates to the field of DR image recognition technology; the method includes obtaining at least one DR knee joint image, and preprocessing the DR knee joint image by a threshold method; using a fuzzy C-means clustering method to segment the fibula and tibia in the preprocessed DR knee joint image; determining the edge of the overlapping part of the fibula and tibia by edge detection; obtaining the lateral edge of the fibula; and judging whether the overlap amount meets the standard based on the edges of the two sides of the fibula and the edges of the overlapping bone tissue. The present invention does not require manual operation, and can completely automatically segment the fibula and tibia, and subsequently uses the overlap amount of the fibula and tibia to determine whether the positioning meets the requirements. The overall process greatly saves the time of judging whether it is a DR knee joint anteroposterior film, and significantly improves the efficiency of image quality control and evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of DR image recognition, and in particular to a method and system for segmenting the fibular head and tibia head in DR knee joint anteroposterior radiographs for quality control. Background Art

[0002] With the continuous development of my country's medical industry, various examination and treatment equipment are constantly improving. Imaging technology plays a vital role in diagnostic examinations across various clinical disciplines. The quality of imaging directly affects the accuracy of diagnostic examination results, and in turn, the accuracy of treatment measures. Therefore, ensuring imaging quality control is a key clinical priority for hospitals. Furthermore, the rapid development of regional and cloud imaging this year has further increased the demand for image quality. Due to the massive volume of images, manual quality control is time-consuming and labor-intensive, making automated image quality control crucial.

[0003] DR knee radiographs require minimal overlap between the fibular head and tibia. This ensures the image is taken from the frontal view, minimizing obstruction between bones and facilitating easy interpretation. Traditional manual identification is time-consuming and laborious, and with the rapid growth of imaging data, manual methods are no longer feasible. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method and system for segmenting the fibula and tibia in DR knee joint AP radiographs for quality control, which can automatically find the fibula and tibia, obtain the overlapping parts between them, and determine whether the overlapping amount meets the standard.

[0005] In a first aspect, an embodiment of the present invention provides a method for segmenting the fibular head and tibia head in a DR knee joint AP radiograph for quality control, comprising:

[0006] Acquire at least one DR knee joint image, and preprocess the DR knee joint image by thresholding;

[0007] The fibula and tibia in the preprocessed DR knee joint images were segmented using the fuzzy C-means clustering method.

[0008] Determine the edge of the overlapping portion of the fibula and the tibia by edge detection;

[0009] obtaining a lateral edge of the fibula;

[0010] Whether the overlap amount meets the standard is determined based on the edges of both sides of the fibula and the edges of the bone tissue of the overlapping portion.

[0011] Furthermore, the DR knee joint image is preprocessed by thresholding, specifically:

[0012] The range of the DR knee joint image is uniformly linearly mapped to the range of [-1024, 4096];

[0013] The classification threshold is adaptively obtained using the OTSU method to divide the DR knee joint image into bone tissue and non-bone tissue;

[0014] Perform vertical projection on the binarized DR knee joint image to obtain a cumulative histogram;

[0015] If there is a zero region in the middle of the histogram, the DR knee joint image is divided into left and right parts at the midpoint of the zero region to obtain a left knee joint anteroposterior radiograph and a right knee joint anteroposterior radiograph.

[0016] Furthermore, horizontal projection was performed on the DR knee joint images for which the left and right knee joint AP radiographs were not obtained to obtain a cumulative histogram;

[0017] If there is a zero region in the middle of the histogram, the DR knee joint image is divided into upper and lower parts at the midpoint of the zero region to obtain an anteroposterior radiograph of the left knee joint and an anteroposterior radiograph of the right knee joint;

[0018] Then the left knee joint anteroposterior film and the right knee joint anteroposterior film are rotated 90 degrees and saved.

[0019] Furthermore, a horizontal line is drawn at one quarter and three quarters of the height of the binarized DR knee joint image, and the ranges of the high-value points of the two are compared. The larger range is the fibula-tibia side. If the fibula-tibia side is at the upper part, the DR knee joint image is rotated 180 degrees.

[0020] Furthermore, the fuzzy C-means clustering method is used to segment the fibula and tibia in the preprocessed DR knee joint image, specifically:

[0021] Classify each pixel of the bone tissue to obtain the membership u1 of the tibia and the membership u2 of the fibula. The corresponding objective function is:

[0022]

[0023] Where i, j are class labels; u ij Represents sample x i The degree of membership to class j; x is a sample with 2-dimensional features; C j is the center of cluster j and also has dimension 2.

[0024] Furthermore, an iterative method is used to solve the optimal problem of the objective function, specifically:

[0025] Set the accuracy, fuzzy index and maximum number of iterations of the objective function;

[0026] Initialize the membership matrix u1j 、u 2j and cluster centers c1 and c2, where the cluster centers are the left and right points of the horizontal line that determines the tibia side of the fibula, c1 is the initial center point of the tibia, and c2 is the initial center point of the fibula;

[0027] Update the fuzzy partition matrix and cluster centers:

[0028]

[0029]

[0030] If the change in the objective function J is less than the preset accuracy, the iteration ends; otherwise, the above formula is used to update again.

[0031] Furthermore, the edge of the overlapping portion of the fibula and the tibia is determined by edge detection, specifically:

[0032] Perform seed growth in both the upper and lower directions on the inner edge points of the fibula and tibia, growing downward to the bottom of the DR knee joint image and upward to the intersection position. This position is the overlapping intersection point, recorded as X1;

[0033] The two grown point sets are used to perform parabolic fitting to obtain the medial direction equations J(x) and f(x) of the fibula and tibia respectively;

[0034] At the intersection position, the respective direction equations continue to grow until there is no suitable edge point within the equation range with a deviation of n pixels;

[0035] By moving toward the stopping point of the equation growth, the edge is grown upward using the template until no new points can be found, where the template uses:

[0036] or

[0037] Furthermore, the outer edge of the fibula is obtained, specifically:

[0038] The obtained overlapping edges constitute a closed area. The average membership v is obtained through the membership of the points in the closed area. The average membership v and the template are used to grow the edge upward to find new points. The membership of the new points is required to be within the range of [v-0.05, v+0.05]. The new points are recorded as overlapping bone tissue.

[0039] Furthermore, based on the edges of the fibula on both sides and the edges of the overlapping bone tissue, it is determined whether the overlap amount meets the standard, specifically:

[0040] According to the edge lines on both sides of the fibula and the edge lines of the overlapping bone tissue, the distance of the maximum horizontal position is taken as the transverse diameter of the fibula; the intersection of this position and the edge line of the tibia is the distance of the overlapping part. If the distance of the overlapping part is less than 1 / 2 of the transverse diameter, the overlap amount is qualified.

[0041] In a second aspect, an embodiment of the present invention provides a system for segmenting the fibular head and tibia head in a quality control DR knee joint AP radiograph, comprising:

[0042] a preprocessing module, acquiring at least one DR knee joint image and preprocessing the DR knee joint image by a threshold method;

[0043] The segmentation module uses the fuzzy C-means clustering method to segment the fibula and tibia in the preprocessed DR knee joint image;

[0044] an overlap acquisition module, for determining the edge of the overlapped portion of the fibula and the tibia by edge detection;

[0045] a lateral edge acquisition module, for acquiring the lateral edge of the fibula;

[0046] The judgment module judges whether the overlap amount meets the standard based on the edges of the fibula on both sides and the edges of the overlapping bone tissue.

[0047] In the third aspect, an embodiment of the present invention provides a device for segmenting and processing the fibula and tibia heads in DR knee joint AP films during quality control, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0048] In a fourth aspect, an embodiment of the present invention provides a computer-readable medium having a processor-executable non-volatile program code, wherein the program code enables the processor to execute the steps of any one of the methods described in the first aspect.

[0049] The present invention provides a method and system for segmenting the fibula and tibia in DR knee AP radiographs for quality control. This method automatically segments the fibula and tibia without manual intervention, and then uses the overlap between the fibula and tibia to determine whether the image is positioned correctly. This process significantly reduces the time required to determine whether a DR knee AP radiograph is correct, significantly improving the efficiency of image quality control and evaluation.

[0050] Other features and advantages disclosed in the present invention will be described in the following description, or some features and advantages can be inferred or determined without doubt from the description, or can be learned by practicing the above-mentioned technology disclosed in the present invention.

[0051] In order to make the above-mentioned objects, features and advantages disclosed in the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 A schematic structural diagram of a processing device provided in an embodiment of the present invention;

[0054] Figure 2 A flow chart of a method for segmenting the fibular head and tibia head in a quality control DR knee joint AP radiograph provided by an embodiment of the present invention;

[0055] Figure 3 DR knee joint image preprocessing flow chart provided in an embodiment of the present invention;

[0056] Figure 4 A flow chart for determining the edge of the overlapping portion of the fibula and tibia provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It will be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions relevant to the present invention, not all structures.

[0058] Before discussing the exemplary embodiments in more detail, it should be noted that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the steps as sequential processes, many of the steps can be performed in parallel, concurrently, or simultaneously. In addition, the order of the steps can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the accompanying drawings.

[0059] Example 1:

[0060] First, refer to Figure 1 The processing device 100 for implementing the embodiments of the present invention is described below. The processing device can be used to execute the methods of various embodiments of the present invention.

[0061] like Figure 1As shown, the processing device 100 includes one or more processors 102, one or more memories 104, an input device 106, an output device 108, and a data collector 110, and these components are interconnected via a bus system 112 and / or other forms of connection mechanisms (not shown). It should be noted that Figure 1 The components and structure of the processing device 100 shown are merely exemplary and non-limiting. The processing device may also have other components and structures as needed.

[0062] The processor 102 can be implemented in at least one hardware form of a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic array (PLA) and an ASIC (Application Specific Integrated Circuit). The processor 102 can be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and can control other components in the processing device 100 to perform desired functions.

[0063] The memory 104 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 102 may execute the program instructions to implement the client functions and / or other desired functions in the embodiments of the present invention (implemented by the processor) described below. Various applications and various data may also be stored in the computer-readable storage medium, such as various data used and / or generated by the application.

[0064] The input device 106 may be a device used by a user to input instructions, and may include one or more of a keyboard, a mouse, a microphone, a touch screen, and the like.

[0065] The output device 108 may output various information (eg, images or sounds) to the outside (eg, a user), and may include one or more of a display, a speaker, and the like.

[0066] The data collector 110 is used for data collection, wherein the data collected by the data collector is the original data or target data of the current target. Then, the data collector can also store the original data or target data in the memory 104 for use by other components.

[0067] For example, the processing device used to implement the method for segmenting the fibula and tibia heads in the DR knee joint AP radiographs in quality control according to an embodiment of the present invention may be an intelligent terminal such as a server, a smart phone, a tablet computer, or a computer.

[0068] Example 2:

[0069] See also Figure 2 The method for segmenting the fibular head and tibia head in a DR knee joint AP radiograph during quality control is shown, and is applied to a mobile terminal. The method can be executed by the processing device provided in the aforementioned embodiment, and specifically includes the following steps:

[0070] S110. Acquire at least one DR knee joint image and preprocess the DR knee joint image by thresholding;

[0071] Specifically, in order to be compatible with DR knee joint images of various manufacturers in various shooting ranges, preprocessing is performed here. If the range is large and includes both legs, segmentation is performed to obtain a unilateral knee joint image.

[0072] S120. Segment the fibula and tibia in the preprocessed DR knee joint image using the fuzzy C-means clustering method;

[0073] It should be noted that the fuzzy C-means clustering method (FCM) is used to segment the fibula and tibia. Because the fibula and tibia overlap at the fibula head, it is unreasonable to use other segmentation methods to classify each point as either tibia or fibula, so the fuzzy method is used.

[0074] S130. Determine the edge of the overlapping portion of the fibula and tibia by edge detection;

[0075] For example, the values ​​of a DR knee joint image in the range [-1024, 4096] are mapped to [0, 255] using the window width and window level method. Then, a set of all edge points is obtained using the Canny edge detection method. Using the lateral line of the fibula and tibia, the outer edge points of the fibula and tibia can be determined. The edge point between the two outer edge points represents the medial side of the tibia and fibula. The fact that the tibia is much thicker than the fibula can be used to determine which two edge points are within the tibia and which two are within the fibula.

[0076] S140. Obtaining the lateral edge of the fibula;

[0077] For example, the overlapping edges obtained in step S130 form a closed region. The average value v is obtained from the membership of the points within the closed region. Since the edge growth just now did not obtain a complete overlapping region, the average membership v and the template are used to grow the edge upward to find new points. The membership of the new points must be within the range of [v-0.05, v+0.05] until no new points can be added. These new points are recorded as the overlapping bone tissue.

[0078] S150. Determine whether the overlap amount meets the standard based on the edges of the fibula on both sides and the edges of the overlapping bone tissue.

[0079] For example, based on the edge lines on both sides of the fibula and the edge lines of the overlapping bone tissue, the distance of the maximum horizontal position is taken as the transverse diameter of the fibula; the intersection of this transverse diameter and the edge line of the tibia is the distance of the overlapping part. If the distance of the overlapping part is less than 1 / 2 of the transverse diameter, the overlap amount is qualified.

[0080] The present invention provides a method for segmenting the fibula and tibia in DR knee AP radiographs for quality control. This method automatically segments the fibula and tibia without manual intervention. The overlap between the fibula and tibia is then used to determine whether the image is positioned correctly. This process significantly reduces the time required to determine whether a DR knee radiograph is correct, significantly improving the efficiency of image quality control and evaluation.

[0081] See also Figure 3 The DR knee joint image preprocessing flow chart shown in FIG. preprocesses the DR knee joint image by a threshold method. The specific implementation method may be:

[0082] S111. uniformly linearly map the range of the DR knee joint image to the range of [-1024, 4096];

[0083] For example, some manufacturers' equipment needs to apply a custom mapping table LUT when generating DR knee joint images. In this step, various mapping tables LUT are used according to the protocol, using 16-bit signed data, and the value range can be uniformly linearly mapped to the range of [-1024, 4096].

[0084] S112. Using the otsu method to adaptively obtain a classification threshold, the DR knee joint image is divided into bone tissue and non-bone tissue;

[0085] For example, here, since bone marrow and osteoporosis have low density and relatively low value, they are also classified as non-bone tissue.

[0086] S113. Performing vertical projection on the binarized DR knee joint image to obtain a cumulative histogram;

[0087] S114. If there is a zero region in the middle of the histogram, the DR knee joint image is separated into left and right sections at the midpoint of the zero section to obtain a left knee joint AP film and a right knee joint AP film.

[0088] It should be noted that for DR knee joint images without left and right knee joint AP radiographs, a horizontal projection is performed to obtain a cumulative histogram. If there is a zero-value region in the middle of the histogram, the DR knee joint image is split vertically at the midpoint of the zero-value region to obtain left and right knee joint AP radiographs. This is an image of the leg bone taken in a horizontal orientation. The left and right knee joint AP radiographs are then rotated 90 degrees and saved. Subsequent analysis is based on the vertical orientation of the leg bone images.

[0089] S115. Draw a horizontal line at one quarter and three quarters of the height of the binarized DR knee joint image, and compare the ranges of the high-value points of the two. The larger range is the fibula-tibia side. If the fibula-tibia side is at the upper part, rotate the DR knee joint image 180 degrees.

[0090] Preferably, the fuzzy C-means clustering method is used to segment the fibula and tibia in the preprocessed DR knee joint image, specifically:

[0091] Classify each pixel of the bone tissue to obtain the membership u1 of the tibia and the membership u2 of the fibula. The corresponding objective function is:

[0092]

[0093] Where i, j are class labels; u ij Represents sample x i The degree of membership to class j; x is a sample with 2-dimensional features; C j is the center of cluster j and also has 2 dimensions;

[0094] The iterative method is used to solve the optimal problem of the objective function, specifically:

[0095] S121. Set the precision of the objective function to 1e-5, the fuzzy index to 2, and the maximum number of iterations to 100;

[0096] S122. Initialize membership matrix u 1j 、u 2j and cluster centers c1 and c2, where the cluster centers are the left and right points of the horizontal line that determines the tibia side of the fibula, c1 is the initial center point of the tibia, and c2 is the initial center point of the fibula;

[0097] Update the fuzzy partition matrix and cluster centers:

[0098]

[0099]

[0100] S123. If the change in the objective function J is less than the preset accuracy, the iteration ends; otherwise, return to S122 and update again.

[0101] Generally, the fuzzy C-means clustering method is used. After the iteration stops, the maximum membership of each pixel is used to determine which category the point belongs to. Due to the overlap of bones, this method needs to further judge the pixels with a maximum membership within [0.5-0.7]. This part may belong to overlapping bone tissue.

[0102] See also Figure 4 The flowchart for determining the edge of the overlapping portion of the fibula and tibia is shown; the edge of the overlapping portion of the fibula and tibia is determined by edge detection. The specific implementation method may be:

[0103] S131. Perform seed growth in both the upper and lower directions on the inner edges of the fibula and tibia, growing downward to the bottom of the DR knee joint image and upward to the intersection. This location is the overlapping intersection, denoted as X1.

[0104] S132. Using the two grown point sets, perform parabolic fitting to obtain the medial direction equations J(x) and f(x) of the fibula and tibia, respectively;

[0105] S133. The equations at the intersection positions continue to grow until there are no suitable edge points within the equation range with a deviation of n pixels;

[0106] S134. Using the template, grow the edge upwards until no new point is found, by moving toward the stopping point of the equation growth. The template uses:

[0107] or

[0108] They are selected according to their original directions.

[0109] Example 3:

[0110] Regarding the method for segmenting the fibula and tibia heads in the quality control DR knee joint anteroposterior radiograph provided in the second embodiment, an embodiment of the present invention provides a device for segmenting the fibula and tibia heads in the quality control DR knee joint anteroposterior radiograph, which is applied to a mobile terminal and includes:

[0111] a preprocessing module, acquiring at least one DR knee joint image and preprocessing the DR knee joint image by a threshold method;

[0112] The segmentation module uses the fuzzy C-means clustering method to segment the fibula and tibia in the preprocessed DR knee joint image;

[0113] an overlap acquisition module, for determining the edge of the overlapped portion of the fibula and the tibia by edge detection;

[0114] a lateral edge acquisition module, for acquiring the lateral edge of the fibula;

[0115] The judgment module judges whether the overlap amount meets the standard based on the edges of the fibula on both sides and the edges of the overlapping bone tissue.

[0116] The device provided in this embodiment has the same implementation principle and technical effects as those of the aforementioned embodiments. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0117] In addition, this embodiment provides a device for segmenting and processing the fibula head and tibia head in the DR knee joint AP film during quality control. The device includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements the method for segmenting the fibula head and tibia head in the DR knee joint AP film during quality control provided in the above-mentioned embodiment 2.

[0118] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the segmentation method can refer to the corresponding process in the aforementioned embodiment, and will not be repeated here.

[0119] Furthermore, this embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method provided in the above-mentioned embodiment 2 are executed.

[0120] The computer program product of a method and system for segmenting the fibular head and tibia head in a DR knee joint anteroposterior radiograph in quality control provided by an embodiment of the present invention includes a computer-readable storage medium storing a program code, wherein the instructions included in the program code can be used to execute the method described in the previous method embodiment. For specific implementation, please refer to the method embodiment and will not be repeated here. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or partly contributed to the prior art or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0121] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed by the present invention, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph, characterized in that: include: Acquire at least one DR knee joint image and preprocess the DR knee joint image using a threshold method, specifically: The range of the DR knee joint image is uniformly linearly mapped to the range of [-1024, 4096]; The classification threshold is adaptively obtained using the OTSU method to divide the DR knee joint image into bone tissue and non-bone tissue; Perform vertical projection on the binarized DR knee joint image to obtain a cumulative histogram; If there is a zero region in the middle of the histogram, the DR knee joint image is divided into left and right sections at the midpoint of the zero region to obtain a left knee joint anteroposterior radiograph and a right knee joint anteroposterior radiograph; The fibula and tibia in the preprocessed DR knee joint images were segmented using the fuzzy C-means clustering method. The edge of the overlapping portion of the fibula and tibia is determined by edge detection, specifically: seed growth is performed on the inner edge points of the fibula and tibia in the upper and lower directions, growing downward to the bottom of the DR knee joint image and upward to the intersection position, which is the overlapping intersection point, recorded as X1; parabolic fitting is performed using the two grown point sets to obtain the inner side direction equations J(x) and f(x) of the fibula and tibia respectively; the direction equations at the intersection position continue to grow until there is no suitable edge point within the equation range with a deviation of n pixels; through the stopping point of the direction equation growth, the edge is grown upward using a template until no new point is found, wherein the template adopts: or obtaining a lateral edge of the fibula; Whether the overlap amount meets the standard is determined based on the edges of both sides of the fibula and the edges of the bone tissue of the overlapping portion.

2. The method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph according to claim 1, characterized in that: For DR knee joint images that do not have left and right knee joint AP radiographs, horizontal projection is performed to obtain a cumulative histogram; If there is a zero region in the middle of the histogram, the DR knee joint image is divided into upper and lower parts at the midpoint of the zero region to obtain an anteroposterior radiograph of the left knee joint and an anteroposterior radiograph of the right knee joint; Then the left knee joint anteroposterior film and the right knee joint anteroposterior film are rotated 90 degrees and saved.

3. The method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph according to claim 1, characterized in that: Draw a horizontal line at one quarter and three quarters of the height of the binarized DR knee joint image, and compare the ranges of the high-value points of the two. The larger range is the fibula-tibia side. If the fibula-tibia side is at the top, rotate the DR knee joint image 180 degrees.

4. The method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph according to claim 1, characterized in that: The fuzzy C-means clustering method is used to segment the fibula and tibia in the preprocessed DR knee joint image, specifically: Classify each pixel of bone tissue to get the membership u1 of the tibia and the membership u2 of the fibula. The corresponding objective function is: Where i, j are class labels; u ij Represents sample x i The degree of membership to class j; x is a sample with 2-dimensional features; C j is the center of cluster j and also has dimension 2.

5. The method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph according to claim 4, characterized in that: The iterative method is used to solve the optimal problem of the objective function, specifically: Set the accuracy, fuzzy index and maximum number of iterations of the objective function; Initialize the membership matrix u 1j 、u 2j and cluster centers c1 and c2, where the cluster centers are the left and right points of the horizontal line that determines the tibia side of the fibula, c1 is the initial center point of the tibia, and c2 is the initial center point of the fibula; Update the fuzzy partition matrix and cluster centers: If the change in the objective function J is less than the preset accuracy, the iteration ends; otherwise, the above formula is used to update again.

6. The method for segmenting the fibular head and tibia head in a quality control DR knee joint anteroposterior radiograph according to claim 1, characterized in that: Obtain the lateral edge of the fibula, specifically: The obtained overlapping edges constitute a closed area. The average membership v is obtained through the membership of the points in the closed area. The average membership v and the template are used to grow the edge upward to find new points. The membership of the new points is required to be within the range of [v-0.05, v+0.05]. The new points are recorded as overlapping bone tissue.

7. The method for segmenting the fibular head and tibia head in a quality control DR knee joint AP radiograph according to claim 1, characterized in that: Based on the edges of the fibula on both sides and the edges of the overlapping bone tissue, it is determined whether the overlap meets the standard, specifically: According to the edge lines on both sides of the fibula and the edge lines of the overlapping bone tissue, the distance at the maximum horizontal position is taken as the transverse diameter of the fibula; the intersection of this transverse diameter and the edge line of the tibia is the distance of the overlapping part. If the distance of the overlapping part is less than 1 / 2 of the transverse diameter, the overlap amount is qualified.

8. A system for segmenting the fibular head and tibia head in DR knee joint anteroposterior radiographs for quality control, characterized in that: include: A preprocessing module is provided for obtaining at least one DR knee joint image and preprocessing the DR knee joint image by means of a threshold value method: the range of the DR knee joint image is uniformly linearly mapped to the range of [-1024, 4096]; a classification threshold is adaptively obtained using the OTSU method to classify the DR knee joint image into bone tissue and non-bone tissue; a binarized DR knee joint image is vertically projected to obtain a cumulative histogram; if there is a zero region in the middle of the histogram, the DR knee joint image is separated into left and right anteroposterior radiographs at the midpoint of the zero region to obtain left and right knee joint anteroposterior radiographs; The segmentation module uses the fuzzy C-means clustering method to segment the fibula and tibia in the preprocessed DR knee joint image; The overlap acquisition module determines the edge of the overlapping portion of the fibula and tibia through edge detection. Specifically, the following steps are performed: seed growth is performed on the inner edge points of the fibula and tibia in the upper and lower directions, growing downward to the bottom of the DR knee joint image and upward to the intersection position. This position is the overlapping intersection point, recorded as X1; parabolic fitting is performed using the two grown point sets to obtain the inner side direction equations J(x) and f(x) of the fibula and tibia respectively; the direction equations at the intersection position continue to grow until there is no suitable edge point within the equation range of n pixels; through the stopping point of the direction equation growth, the edge is grown upward using a template until no new point is found, where the template adopts: or a lateral edge acquisition module, for acquiring the lateral edge of the fibula; The judgment module judges whether the overlap amount meets the standard based on the edges of the fibula on both sides and the edges of the overlapping bone tissue.

Images