A spinal region segmentation system for osteoporosis identification

Abstract

The present invention relates to the field of image data processing technology, and in particular to a spinal region segmentation system for osteoporosis identification, which obtains a first sliding window by acquiring distribution parameters of a bone scan image and a sliding window, and compares the local and overall distribution parameters, obtains the window weight of the first sliding window according to the difference in distribution parameters, obtains a center window and a neighborhood window in combination with the positional relationship of the sliding window, further obtains the confidence of the center window, and uses the window weight of the center window and the weight factor of the corresponding neighborhood window to adjust the weight factor of the center window to obtain the weighted confidence of the center window, and obtains the spinal region according to the size of the weighted confidence. The present invention accurately extracts each spinal region in the bone scan image by extracting and analyzing the grayscale features of windows at different positions in the bone scan image.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a spinal region segmentation system for osteoporosis identification. Background Art

[0002] Currently, osteoporosis is usually determined by doctors using professional bone density measurements. However, this method has some limitations: first, the results need to be interpreted by professional doctors, which may involve subjectivity and the risk of misjudgment; second, these methods usually require the use of expensive instruments and equipment, which limits their application in the field of health care.

[0003] In the process of detecting and identifying the spine area of ​​CT images using image processing technology, existing algorithms usually directly use the threshold segmentation method for detection. However, when the spine area is obtained by simple threshold segmentation, since there are regions formed by multiple grayscale values ​​in the bone scan image collected by the CT machine, it is easy to be interfered by non-bone areas when the image is segmented using the determined threshold, resulting in difficulty in further specific analysis of different parts of the bone area in the image after threshold segmentation, resulting in low accuracy of spine area segmentation, which is not conducive to subsequent osteoporosis identification based on the spine area, thereby affecting the diagnosis result. Summary of the invention

[0004] The present invention provides a spinal region segmentation system for osteoporosis identification to solve the existing problems.

[0005] The present invention provides a spinal region segmentation system for osteoporosis identification using the following technical solutions:

[0006] The present invention provides a spine region segmentation system for osteoporosis identification, the system comprising the following modules:

[0007] Image acquisition module: used to obtain bone scanning images;

[0008] Sliding window module: used to obtain the distribution parameters of the bone scan image through the grayscale values ​​of all pixels in the bone scan image; construct a sliding window of a preset size and traverse the bone scan image according to a preset step length, obtain the distribution parameters of the sliding window during the traversal process, and obtain the first sliding window and the second sliding window according to the size of the distribution parameters of the sliding window, and the distribution parameters of the sliding window are obtained in the same way as the distribution parameters of the bone scan image; obtain the window weight of the first sliding window according to the difference between the distribution parameters of the first sliding window and the bone scan image;

[0009] Spine determination module: used to select a sliding window at any traversal position as the center window, obtain the neighborhood window of the center window according to the positional relationship between the sliding window and the center window; determine whether the sliding window is the first sliding window or the second sliding window, assign a preset weight factor to the sliding window according to the determination result, and obtain the confidence of the center window according to the weight factors of the center window and the neighborhood window; adjust the weight factor of the center window using the window weight of the center window and the weight factor of the corresponding neighborhood window to obtain the weighted confidence of the center window; obtain the spine area according to the size of the weighted confidence.

[0010] Furthermore, the distribution parameters of the bone scan image are obtained by using the grayscale values ​​of all pixels in the bone scan image, including:

[0011] The mean and variance of the grayscale values ​​of all pixels in the bone scan image are obtained and recorded as the first parameter and the second parameter of the bone scan image respectively, and the ratio of the second parameter to the first parameter is recorded as the distribution parameter of the bone scan image.

[0012] Further, the step of constructing a sliding window of a preset size and traversing the bone scan image according to a preset step length, obtaining distribution parameters of the sliding window during the traversal process, and obtaining the first sliding window and the second sliding window according to the size of the distribution parameters of the sliding window includes:

[0013] First, a sliding window of size A×A is constructed. The sliding window traverses the bone scan image with a step size of a to obtain the second parameter and distribution parameter of the sliding window, where A and a are both preset hyperparameters.

[0014] Then, if the distribution parameter of the sliding window is less than the distribution parameter of the bone scan image, the sliding window is recorded as the first sliding window; if the distribution parameter of the sliding window is greater than or equal to the distribution parameter of the bone scan image, the sliding window is recorded as the second sliding window.

[0015] Further, obtaining the window weight of the first sliding window according to the difference between the distribution parameters of the first sliding window and the bone scan image includes:

[0016] The specific calculation method of the window weight of any first sliding window is:

[0017]

[0018] Where W represents the window weight of the first sliding window; C ′ represents the distribution parameter of the first sliding window; C represents the distribution parameter of the bone scan image; δ ′ Represents the second parameter of the first sliding window; exp() represents an exponential function with a natural constant as the base.

[0019] Furthermore, the sliding window at any traversal position is selected as the center window, and the neighborhood window of the center window is obtained according to the positional relationship between the sliding window and the center window, including:

[0020] For the sliding windows at all positions during the sliding traversal process, select any sliding window as the center window, and obtain the sliding windows at the positions corresponding to the 8 neighborhoods of the center window, which are recorded as the neighborhood windows of the center window.

[0021] Further, the determining whether the sliding window is the first sliding window or the second sliding window and assigning a preset weight factor to the sliding window according to the determination result includes:

[0022] When the central window is the first sliding window, the weight factor T1 is assigned to the central window. Otherwise, when the central window is not the first sliding window, that is, when the central window is the second sliding window, the weight factor T0 is assigned to the central window. When any neighborhood window is the first sliding window, the weight factor T2 is assigned to the corresponding neighborhood window, and when the neighborhood window is the second window, the weight factor T0 is assigned to the neighborhood window, where T0, T1 and T2 are all preset hyperparameters.

[0023] Further, the step of obtaining the confidence of the center window according to the weight factors of the center window and the neighborhood windows includes:

[0024] First, the set formed by any central window and the corresponding neighboring windows under the 8-neighborhood is recorded as a window set, and the sequence number of the central window in the window set is recorded as 0. For the neighboring windows under the 8-neighborhood, the sequence numbers of the neighboring windows are set from 1 to 8 in the order from left to right and from top to bottom;

[0025] Then, the specific calculation method of the confidence of the central window in the window set is:

[0026]

[0027] Among them, P represents the confidence of the center window; T i Represents the weight factor of the i-th sliding window in the window set; T1 and T2 are both preset hyperparameters.

[0028] Furthermore, the weight factor of the central window is adjusted by using the window weight of the central window and the weight factor of the corresponding neighboring window to obtain the weighted confidence of the central window, and the specific steps include the following:

[0029] The specific calculation method of the weighted confidence of the center window is:

[0030]

[0031] Among them, S represents the weighted confidence of the center window; Pi represents the confidence of the i-th sliding window in the window set; G represents the preset hyperparameter; P represents the confidence of the center window; W′ is the window weight of the center window.

[0032] Furthermore, obtaining the spine region according to the weighted confidence comprises:

[0033] The weighted confidences of all first sliding windows in the bone scan image are obtained, and the first sliding windows whose weighted confidences are greater than the hyperparameter G are recorded as the spine region.

[0034] The beneficial effect of the technical solution of the present invention is that since the osteoporotic area is formed by multiple small connected domains, the system's sensitivity to the grayscale value distribution characteristics in the local area of ​​the bone scan image is improved by traversing the bone scan image using a sliding window, and each spinal region in the bone scan image is obtained by performing feature extraction and weighted analysis on the sliding window at different positions during the traversal process, thereby avoiding the influence of non-bone areas in the image on the spinal region segmentation results, enhancing the accuracy and reliability of the spinal region segmentation results, and further improving the detection accuracy of the osteoporotic area. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 This is a module flow chart of a spinal region segmentation system for osteoporosis identification according to the present invention. DETAILED DESCRIPTION

[0037] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation, structure, features and effects of a spinal region segmentation system for osteoporosis identification proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0038] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0039] The following is a detailed description of a specific solution of a spinal region segmentation system for osteoporosis identification provided by the present invention in conjunction with the accompanying drawings.

[0040] See also Figure 1 , which shows a module flow chart of a spine region segmentation system for osteoporosis identification provided by an embodiment of the present invention, the system includes the following modules:

[0041] Image acquisition module: used to obtain bone scanning images.

[0042] The CT machine is used to collect bone scan images of the patient.

[0043] At this point, a bone scan image is obtained.

[0044] Sliding window module: used to traverse the bone scan image through a sliding window and obtain the window weight of the sliding window according to the grayscale values ​​of all pixels in the sliding window.

[0045] It should be noted that the grayscale value boundary of the black background and the bone can be distinguished based on the grayscale value distribution of the overall image. The image is traversed using a sliding window and the grayscale values ​​within the sliding window are counted. After the confidence is calculated, the central window is weighted according to the confidence of the neighboring window. If it is greater than a certain threshold, it can be assessed as the spine part.

[0046] It should be noted that it is obvious from the image that the average grayscale value of the spine is low and the grayscale value does not change much. The distribution is relatively concentrated, and the corresponding variance is relatively small. The coefficient of variation of the spine area can reflect whether the change of the grayscale value of the spine area is stable. By observing the spine area image, it can be found that the grayscale value of the spine area is more stable than that of other bone parts. Therefore, the coefficient of variation of the grayscale value of the spine area is lower than the coefficient of variation of the global grayscale value. The coefficient of variation of the area in the sliding window is compared with the global coefficient of variation. The area with a coefficient of variation lower than the global coefficient of variation indicates that the bone grayscale value in this area changes less, and the bones of the spine meet such characteristics and may be the target area.

[0047] Step (1), obtaining the mean and variance of the grayscale values ​​of all pixels in the bone scan image as the first parameter and the second parameter of the bone scan image, respectively, and recording the ratio of the second parameter to the first parameter as the distribution parameter of the bone scan image.

[0048] Step (2), first, construct a sliding window of size A×A, the sliding window traverses the bone scan image with a step size of a, and obtains the distribution parameters of the sliding window according to the grayscale values ​​of all pixels in the sliding window, where A and a are both preset hyperparameters.

[0049] It should be noted that the size of the sliding window A×A is preset to 3×3 based on experience, and the step length a is 1, which can be adjusted according to actual conditions and is not specifically limited in this embodiment.

[0050] It should be noted that in the process of calculating the distribution parameters of the sliding window, the mean of the grayscale values ​​in the sliding window reflects the overall grayscale of the pixel points of the bone part in the sliding window, and the variance of the grayscale values ​​in the sliding window reflects the discreteness of the grayscale values ​​of the bone part in the region, so as to further compare and analyze the distribution parameters of the sliding window obtained by the variance and mean with the distribution parameters of the bone scan image.

[0051] Then, the distribution parameters of the sliding window are compared with the distribution parameters of the bone scan image. If the distribution parameters of the sliding window are smaller than the distribution parameters of the bone scan image, the sliding window is recorded as the first sliding window; if the distribution parameters of the sliding window are greater than or equal to the distribution parameters of the bone scan image, the sliding window is recorded as the second sliding window.

[0052] Finally, according to the difference between the distribution parameters of the first sliding window and the bone scan image and the second parameter corresponding to the first sliding window, the window weight of any first sliding window is obtained. The specific calculation method is:

[0053]

[0054] Where W represents the window weight of the first sliding window; C ′ represents the distribution parameter of the first sliding window; C represents the distribution parameter of the bone scan image; δ ′ Represents the second parameter of the first sliding window; exp() represents an exponential function with a natural constant as the base.

[0055] It should be noted that the second parameter of the first sliding window, i.e., the variance of the grayscale values ​​of all pixels in the first sliding window, reflects the degree of discreteness of the grayscale values ​​of the pixels in the sliding window; in addition, since the grayscale value of the vertebrae is more discrete and the overall grayscale value is larger, the grayscale value distribution of the non-vertebrae area is more uniform and the overall grayscale value is smaller. Therefore, when the distribution parameter of the area corresponding to the sliding window is close to the distribution parameter of the bone scan image, the sliding window is more likely to contain the vertebrae bone area.

[0056] At this point, the window weight of the first sliding window is obtained.

[0057] Spine determination module: used to obtain the confidence of the sliding window by combining the positional relationship between the sliding windows; and to adjust the weight factor of the center window through the window weight of the center window and the weight factor of the neighboring window corresponding to the center window to obtain the weighted confidence of the center window.

[0058] Step 1: First, for the sliding windows at all positions during the sliding traversal process, select any sliding window as the center window, and obtain the sliding window at the position corresponding to the 8 neighborhoods of the center window, which are recorded as the neighborhood windows of the center window.

[0059] It should be noted that when obtaining the neighborhood window of the center window, the problem of exceeding the boundary may occur. In this embodiment, a sliding window identical to the center window is copied as the neighborhood window corresponding to the center window when exceeding the boundary.

[0060] Then, when the central window is the first sliding window, the weight factor T1 is assigned to the central window. Conversely, when the central window is not the first sliding window, that is, when the central window is the second sliding window, the weight factor T0 is assigned to the central window. When any neighborhood window is the first sliding window, the weight factor T2 is assigned to the corresponding neighborhood window, and when the neighborhood window is the second window, the weight factor T0 is assigned to the neighborhood window, where T0, T1 and T2 are all preset hyperparameters.

[0061] It should be noted that when the central window is the first sliding window, it means that the bone part of the area corresponding to the central window is likely to be the spine part, which belongs to the target area that needs to be analyzed. Therefore, the weight factor of the central window needs to be assigned a high value, and when the central window is not the first sliding window, the weight factor value of the central window is set to 0; in addition, since the neighborhood window is close to the corresponding central window in spatial position, when the neighborhood window is the first sliding window, the weight factor of the neighborhood window is assigned a higher value, and when the neighborhood window is not the first sliding window, the weight factor value of the neighborhood window is set to 0.

[0062] It should be noted that the order of weight factors should be T1>T2>T0. According to experience, weight factor T1=2, weight factor T2=1, and weight factor T0=0 are preset. They can be adjusted according to actual conditions and are not specifically limited in this embodiment.

[0063] Finally, the set formed by any central window and the corresponding neighborhood windows under the 8-neighborhood is recorded as a window set, and the sequence number of the central window in the window set is recorded as 0. For the neighborhood windows under the 8-neighborhood, the sequence numbers of the neighborhood windows are set from 1 to 8 in order from left to right and from top to bottom. According to the weight factor of the sliding window in any window set, the confidence of the central window in the window set is obtained. The specific calculation method is:

[0064]

[0065] Among them, P represents the confidence of the center window; T i Represents the weight factor of the i-th sliding window in the window set; T1 and T2 are both preset hyperparameters.

[0066] It should be noted that the confidence reflects the probability that the center window is the spinal region. Since the distribution of the spinal region is usually compact, if the neighborhood window of the center window is also part of the spinal region, it means that the center window is more likely to contain the spinal region. In addition, the center window has a greater impact on the judgment result, so the weight of the center window is greater than that of the corresponding neighborhood window. A probability calculation is performed through the center window and the neighborhood window to obtain the confidence corresponding to when the center window contains the spinal region.

[0067] The confidence of all sliding windows in the bone scan image is obtained by the confidence acquisition method.

[0068] Step 2: It should be noted that in order to avoid special conditions in the areas corresponding to some sliding windows, for example: although the area corresponding to the sliding window is the spine area, the grayscale value in the area changes greatly, or although the area corresponding to the sliding window is not the spine area, the grayscale value in the area changes little, which leads to misjudgment of the detection results. Therefore, it is necessary to weight the confidence of the center window according to the confidence of each center window combined with the confidence of the neighboring window.

[0069] First, when the center window is the first sliding window, the weight factor of the center window is adjusted by combining the window weight of the center window and the weight factor of the neighboring window corresponding to the center window to obtain the weighted confidence of the center window. The specific calculation method is:

[0070]

[0071] Among them, S represents the weighted confidence of the center window; P i represents the confidence of the i-th sliding window in the window set; G represents the preset hyperparameter; P represents the confidence of the center window; W′ is the window weight of the center window.

[0072] It should be noted that, since the distance between the neighborhood window and the center window is close, the confidence of the neighborhood window will affect the confidence of the center window. In addition, since the bone distribution of the human body is usually concentrated, this embodiment weights the confidence of the center window by the confidence of the neighborhood window to prevent errors caused by special circumstances and improve the robustness of the confidence. If the confidence of the neighborhood window is greater than the hyperparameter G, it means that the area corresponding to the neighborhood window is very likely to be the spinal area, and the area corresponding to the center window may also be the spinal area, that is, the confidence of the center window is enhanced to a certain extent; and if the confidence of the neighborhood window is less than or equal to the hyperparameter G, it means that the possibility of the area corresponding to the neighborhood window being the spinal area is small, so the confidence of the center window will be weakened to a certain extent.

[0073] It should be noted that the hyperparameter G is preset to 0.8 based on experience and can be adjusted according to actual conditions, and is not specifically limited in this embodiment.

[0074] Then, the weighted confidences of all the first sliding windows in the bone scan image are obtained, and the first sliding windows whose weighted confidences are greater than the hyperparameter G are recorded as the spine region.

[0075] At this point, various spinal regions that can be used to identify osteoporosis are obtained in the bone scan image.

[0076] After obtaining each spinal region in the bone scan image in this embodiment, a subsequent osteoporosis recognition process may be performed. As a specific implementation method, the osteoporosis region is obtained according to the grayscale value of the pixel point in the spinal region.

[0077] It should be noted that the spine region may contain a background region, so the spine portion needs to be determined based on the grayscale value of the pixel points in the spine region.

[0078] First, if the grayscale value of the pixel in the spinal area is greater than the preset grayscale threshold, it means that the pixel belongs to the bone part, otherwise it is the background part. The grayscale value of the pixel in the bone scan image whose grayscale value is greater than the grayscale threshold remains unchanged, and the grayscale value of the pixel below the grayscale threshold is set to 0, then the bone area is formed by the pixels whose grayscale value is not 0.

[0079] Then, the connected domain detection is performed on the bone area to obtain the osteoporosis area formed by multiple connected domains, and the multiple connected domains are visually marked with different colors.

[0080] It should be noted that normal bones have uniform density, so the grayscale value changes smoothly in the CT image, forming a complete connected domain. However, when the bones have osteoporosis problems, the bone density changes and is no longer uniform, so the grayscale value changes in the CT image are no longer smooth, so more connected domains are obtained.

[0081] It should be noted that the exp(-x) model used in this embodiment is only used to indicate that the negative correlation and the result of the constraint model output are in the interval (0,1). In specific implementation, it can be replaced with other models with the same purpose. This embodiment only uses the exp(-x) model as an example for description without making specific limitations on it, where x refers to the input of the model.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A spine region segmentation system for osteoporosis identification, characterized in that: The system includes the following modules: Image acquisition module: used to obtain bone scanning images; Sliding window module: used to obtain the distribution parameters of the bone scan image through the grayscale values ​​of all pixels in the bone scan image; construct a sliding window of a preset size and traverse the bone scan image according to a preset step length, obtain the distribution parameters of the sliding window during the traversal process, and obtain the first sliding window and the second sliding window according to the size of the distribution parameters of the sliding window, and the distribution parameters of the sliding window are obtained in the same way as the distribution parameters of the bone scan image; obtain the window weight of the first sliding window according to the difference between the distribution parameters of the first sliding window and the bone scan image; Spine determination module: used to select a sliding window at any traversal position as the center window, obtain the neighborhood window of the center window according to the positional relationship between the sliding window and the center window; determine whether the sliding window is the first sliding window or the second sliding window, assign a preset weight factor to the sliding window according to the determination result, and obtain the confidence of the center window according to the weight factors of the center window and the neighborhood window; adjust the weight factor of the center window using the window weight of the center window and the weight factor of the corresponding neighborhood window to obtain the weighted confidence of the center window; obtain the spine area according to the size of the weighted confidence; The step of adjusting the weight factor of the central window by using the window weight of the central window and the weight factor of the corresponding neighboring window to obtain the weighted confidence of the central window includes: The specific calculation method of the weighted confidence of the center window is: in, represents the weighted confidence of the center window; Indicates the first The confidence of the sliding window; Represents the preset hyperparameters; Indicates the confidence of the center window; is the window weight of the center window; The step of constructing a sliding window of a preset size and traversing the bone scan image according to a preset step length, obtaining distribution parameters of the sliding window during the traversal, and obtaining the first sliding window and the second sliding window according to the size of the distribution parameters of the sliding window comprises: First, build size The sliding window is The bone scan image is traversed for the step length to obtain the second parameter and distribution parameter of the sliding window, where and All are preset hyperparameters; Then, if the distribution parameter of the sliding window is less than the distribution parameter of the bone scan image, the sliding window is recorded as the first sliding window; if the distribution parameter of the sliding window is greater than or equal to the distribution parameter of the bone scan image, the sliding window is recorded as the second sliding window.

2. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The distribution parameters of the bone scan image obtained by the grayscale values ​​of all pixels in the bone scan image include: The mean and variance of the grayscale values ​​of all pixels in the bone scan image are obtained and recorded as the first parameter and the second parameter of the bone scan image respectively, and the ratio of the second parameter to the first parameter is recorded as the distribution parameter of the bone scan image.

3. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The step of obtaining the window weight of the first sliding window according to the difference between the distribution parameters of the first sliding window and the bone scan image comprises: The specific calculation method of the window weight of any first sliding window is: in, represents the window weight of the first sliding window; represents the distribution parameter of the first sliding window; Represents the distribution parameters of the bone scan image; represents a second parameter of the first sliding window; Represents an exponential function with a natural constant as base.

4. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The sliding window at any traversal position is selected as the center window, and the neighborhood window of the center window is obtained according to the positional relationship between the sliding window and the center window, including: For the sliding windows at all positions during the sliding traversal process, select any sliding window as the center window, and obtain the sliding windows at the positions corresponding to the 8 neighborhoods of the center window, which are recorded as the neighborhood windows of the center window.

5. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The determining whether the sliding window is the first sliding window or the second sliding window and assigning a preset weight factor to the sliding window according to the determination result includes: When the center window is the first sliding window, the weight factor Assign the weight factor to the center window. On the contrary, when the center window is not the first sliding window, that is, the center window is the second sliding window, Assign to the center window; when any neighborhood window is the first sliding window, the weight factor Assign the corresponding neighborhood window, and when the neighborhood window is the second window, the weight factor Assign a neighborhood window, where , and These are all preset hyperparameters.

6. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The step of obtaining the confidence of the central window according to the weight factors of the central window and the neighborhood windows includes: First, the set formed by any central window and the corresponding neighboring windows under the 8-neighborhood is recorded as a window set, and the sequence number of the central window in the window set is recorded as 0. For the neighboring windows under the 8-neighborhood, the sequence numbers of the neighboring windows are set from 1 to 8 in the order from left to right and from top to bottom; Then, the specific calculation method of the confidence of the central window in the window set is: in, Indicates the confidence of the center window; Indicates the first The weight factor of the sliding window; and These are all preset hyperparameters.

7. A spine region segmentation system for osteoporosis identification according to claim 1, characterized in that: The step of obtaining the spine region according to the weighted confidence comprises: Get the weighted confidence of all first sliding windows in the bone scan image, and set the weighted confidence greater than the hyperparameter The first sliding window is recorded as the spine area.

Images