Orthopedic patient wound nursing method based on visual image processing

Abstract

The invention discloses an orthopedic patient wound nursing method based on visual image processing, and relates to the technical field of wound nursing, and the method comprises the following steps: constructing a patient information database for orthopedic patients based on basic information of the patients, and recording the basic information of the patients and wound information, acquiring daily images of the orthopedic patient after wound suture through image acquisition equipment, and recording the daily images in a patient information database; for a wound image in a patient information database, feature pixels of the wound image are extracted and projected into a blank image, and initial wound line pixels and initial thread end line pixels are calibrated for an initial image; numbering the thread ends and recording initial thread end line pixels and the number of the pixels and initial wound line pixels and the number of the pixels; the wound healing state is monitored through visual image processing, the stitch removing opportunity is dynamically judged, the individual difference of patients is adapted, and the wound nursing efficiency of the orthopedic patients is improved.

Description

Technical Field

[0001] This invention relates to the field of wound care technology, and more specifically, to a method for wound care of orthopedic patients based on visual image processing. Background Technology

[0002] In orthopedic clinical treatment, post-suture care and suture removal timing are key factors affecting wound healing quality and reducing the risk of complications. Orthopedic patients have significant individual differences due to age, physical condition, wound location, and degree of injury, and their wound healing speed and recovery status exhibit distinct individual characteristics. Current methods for determining the timing of suture removal in wound care mainly rely on clinical experience to set fixed suture removal cycles. These methods do not fully consider the impact of individual patient differences on wound healing. Some patients, due to their good physical condition and minor wound injuries, heal quickly. If sutures are removed at a fixed time, problems such as embedded suture ends and tissue adhesions may occur, increasing the difficulty of subsequent treatment. On the other hand, some patients, due to old age, frailty, underlying diseases, or wound infections, heal slowly. Premature suture removal may lead to wound dehiscence, bleeding, secondary injury, and infection risks. Furthermore, under the current care model, patients need to frequently visit the hospital for wound check-ups. Medical staff rely on visual observation of the wound to assist in determining the timing of suture removal. This method is highly subjective, has low accuracy, and cannot monitor the dynamic changes in wound healing in real time. It is difficult to detect abnormalities in the suture end area in a timely manner, resulting in problems with low practicality and functionality. Summary of the Invention

[0003] In response to the problems in related technologies, this invention proposes a method for wound care of orthopedic patients based on visual image processing, in order to overcome the aforementioned technical problems existing in the existing related technologies.

[0004] Therefore, the specific technical solution adopted by the present invention is as follows: A method for wound care of orthopedic patients based on visual image processing, the method comprising the following steps: S1. For orthopedic patients, a patient information database is built based on the patient's basic information to record the patient's basic information and wound information. Daily images of orthopedic patients' wounds after suturing are collected by image acquisition equipment and recorded in the patient information database. S2. For wound images in the patient information database, extract feature pixels from the wound images and project them onto a blank image. Label the initial wound line pixels and initial line head pixels for the initial image. Number the line heads and record the initial line head pixels and the number of pixels, as well as the initial wound line pixels and the number of pixels. S3. Define the surrounding skin tissue area in the initial line of the initial image. Based on the number of pixels of the initial line and the number of pixels of the surrounding tissue, and combined with the ratio change rate of different numbered lines in the subsequent captured images, determine the embedding degree of different numbered lines. At the same time, mark the number of detachable lines daily according to the number of pixels of the wound lines in the area of ​​different numbered lines.

[0005] In a preferred embodiment, S1 includes the following steps: S11. Collect basic information and wound information of orthopedic patients, including patient name, contact information, age, wound location, wound length, surgery date, suture date and expected suture removal date. Establish a patient information database through MySQL and create files based on patient name to record the collected information. S12. Collect daily images of the patient's wound suture location using an image acquisition device, and upload the images with the date annotation to the corresponding patient file in the patient information database.

[0006] In a preferred embodiment, S2 includes the following steps: S21. Extract the first patient wound image from the current patient file in the patient information database as the initial image, and identify the wound and suture ends through the edge detection algorithm; S22. Project the identified wound lines and thread ends onto a blank image. For the initial image after projection, record the initial wound line pixel set and the initial thread end pixel set. Then, number the thread ends of the current patient's wound sequentially and record them in the corresponding patient file.

[0007] In a preferred embodiment, S21 includes the following steps: S211. For the first patient wound image in the current patient file in the patient information database, mark it as the initial image and perform color space conversion. At the same time, reduce skin texture noise through Gaussian smoothing filter to obtain the preprocessed initial image I. S212. For the preprocessed initial image I, the horizontal and vertical gradients of the initial image I are calculated using the Sobel operator. Local maxima are preserved and edges are refined using non-maximum suppression. Edge points of the initial image are identified using double thresholding and an edge map is output. ; S213, Regarding the obtained edge map Extract all contours to obtain a contour set. , where each contour Composed of continuous edge pixels, the contour is determined to identify wounds and suture ends, specifically including the following steps: The profile is calculated using the sum of Euclidean distances. The distance between the middle and edge pixels is used to obtain the length. Calculate the aspect ratio of the minimum bounding rectangle. ,in Calculate the curvature features based on the width and height of the rectangle, respectively. ,in Represents the curvature of each point on the outline; when and At that time, Marked as candidate wound edges, where These represent the length threshold and the high curvature threshold, respectively. when and or ,and At that time, Marked as candidate suture ends, among which These represent the short length threshold, aspect ratio threshold, and low curvature threshold, respectively. S214. For each edge pixel of the candidate wound edge on the contour, take a neighborhood of width d on both sides along the normal direction, convert the two sides to the HSV color space, extract the histograms p and q of the H channel, and calculate the histogram difference using Bach distance. ,when If the difference exceeds the difference threshold, the current contour is confirmed as the wound edge; Extract the minimum bounding rectangle region corresponding to the contour of the candidate suture thread, and calculate the average saturation of this region in the HSV color space. ,in Let N be the saturation value of pixel k, and N be the number of pixels in the region. The directional variance is also calculated. ,in These represent the gradient direction and the mean gradient direction, respectively. Less than the saturation threshold and If the value is less than the direction threshold, then the current contour is confirmed as a stitching thread.

[0008] In a preferred embodiment, S22 includes the following steps: S221. Add all pixels whose current contour is confirmed to be the edge of the wound to the wound edge mask. Add all pixels that confirm the current outline is the start of the stitch line to the stitch line mask. Create a solid color image with the same dimensions as the original captured image, and then mask it. , All pixels with a value of 1 are projected onto a solid color image in different colors; S222, Record the wound edge mask obtained from the initial image. All pixels with a value of 1 represent the initial set of wound line pixels, recording the suture mask obtained from the initial image. All pixels with a value of 1 in the set are the initial line start pixel set. Connectivity component analysis is performed to separate spatially connected pixel regions, which are then numbered sequentially and recorded in the corresponding patient files.

[0009] In a preferred embodiment, step S3 includes the following steps: S31. Define the surrounding skin tissue region in the initial line of the initial image. With each pixel of the line as the center, expand outward by a fixed pixel distance and exclude the line pixel itself to obtain the set of surrounding region pixels as the surrounding tissue pixels. S32. Based on the ratio of the initial number of thread head pixels to the number of surrounding tissue pixels, and based on the rate of change of the proportion of different numbered thread heads in subsequent captured images, the degree of thread head embedding is determined. At the same time, based on the daily concentration of wound line pixels in different thread head areas compared with the initial concentration of wound line pixels, the number of detachable thread heads is marked daily.

[0010] In a preferred embodiment, S31 includes the following steps: S311, Initial line pixel set for different numbers ,in Create a binary mask representing the set of all pixel coordinates of the line ending with ID id in the initial image. if and only if pixel coordinates hour Otherwise, it is 0; right Obtain by performing morphological dilation operations The expanding structured element is a disk with radius r. The pixels in the middle are to The set of pixels whose Euclidean distance to any pixel is less than or equal to r is subtracted from the line head body region from the dilated region to obtain the surrounding tissue region mask. ; S312, Extract the mask of the surrounding tissue region. The coordinates of all pixels with a value of 1 constitute the initial peripheral organization pixel set. ,in .

[0011] In a preferred embodiment, S32 includes the following steps: S321, Initial number of pixels for each line head ID Initial pixel count of the surrounding tissue To obtain the initial ratio The number of pixels at each line end on day t is obtained by processing the image on day t using the same process as S21-S22. 1. Calculate the pixel set of the wound lines to obtain the line proportion on day t. ,in The number of skin tissue pixels re-identified by color segmentation within the initial surrounding tissue area; Calculate the current rate of change. Based on the risk warning threshold embedded in the thread end and embedded confirmation threshold Determine the different thread ends with different numbers each day: when This indicates that the visibility of the thread ends has not changed significantly, and there are no obvious signs of embedding. when This indicates that the visibility of the thread ends has begun to decrease, and the tissue is starting to cover the thread ends; when This indicates that the thread end is embedded; S322. Combining the pixel set of the wound lines in the initial image and the pixel set of the wound lines on day t, the wound region is divided into different line regions based on a fixed radius, and the number of pixels is counted to calculate the wound contraction rate. ,in The initial number of image wound pixels for the line with the ID number, given a fixed radius. The number of wound pixels on day t under a fixed radius for the line with the ID number; S323. For all thread ends under day t, mark them as satisfying... and The wire end ID serves as the detachable wire end. This represents the wound healing threshold.

[0012] The beneficial effects of this invention are as follows: This invention collects wound data from different orthopedic patients, sets up initial images after wound suturing, and combines them with images uploaded by patients daily to determine the condition of the suture ends. Based on the wound contraction in different suture end areas, it intelligently determines whether different suture ends are ready for removal, making it easier to flexibly select the suture removal date based on the wound growth of different patients, thus enhancing functionality. This invention determines the appropriate time for suture removal based on the individual patient's condition. It can provide early warning of the risk of suture embedding due to excessively rapid healing, prompting timely suture removal. At the same time, it can also identify insufficient wound strength due to slow healing, avoiding premature suture removal and thus effectively preventing secondary injury and infection risks. This invention abandons the traditional one-size-fits-all fixed suture removal time. By combining visual image processing to continuously monitor the objective changes of the patient's wound, it determines the timing of suture removal, avoiding premature or delayed suture removal due to individual differences in condition. This improves the patient's actual medical experience and enhances the convenience of wound care. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of a method for wound care of orthopedic patients based on visual image processing according to an embodiment of the present invention. Detailed Implementation

[0015] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0016] According to an embodiment of the present invention, a method for wound care of orthopedic patients based on visual image processing is provided.

[0017] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments: Example 1:

[0018] like Figure 1 As shown, according to an embodiment of the present invention, a method for wound care of orthopedic patients based on visual image processing includes the following steps: S1. For orthopedic patients, a patient information database is built based on the patient's basic information to record the patient's basic information and wound information. Daily images of orthopedic patients' wounds after suturing are collected by image acquisition equipment and recorded in the patient information database. S11. Collect basic information and wound information of orthopedic patients, including patient name, contact information, age, wound location, wound length, surgery date, suture date and expected suture removal date. Establish a patient information database through MySQL and create files based on patient name to record the collected information. S12. Collect daily images of the patient's wound suture location using an image acquisition device, and upload the images with the date annotation to the corresponding patient file in the patient information database.

[0019] It should be noted that during the process of acquiring images of the patient's suture location using an image acquisition device, the image acquisition device needs to be kept at a fixed distance from the wound plane and the image acquisition should be performed with fixed parameters to ensure the consistency of subsequent images. Image acquisition devices include mobile phone cameras, cameras, etc. The fixed distance needs to be determined based on the current size of the wound and the actual image acquisition device. Taking a mobile phone camera as an example, a bracket needs to be used to ensure that the lens is kept at a constant distance from the wound plane, the main camera of the device is locked and the automatic lens switching function is disabled, and the lens axis is as perpendicular as possible to the wound plane when taking wound images.

[0020] S2. For wound images in the patient information database, extract feature pixels from the wound images and project them onto a blank image. Label the initial wound line pixels and initial line head pixels for the initial image. Number the line heads and record the initial line head pixels and the number of pixels, as well as the initial wound line pixels and the number of pixels. S21. Extract the first patient wound image from the current patient file in the patient information database as the initial image, and identify the wound and suture ends through the edge detection algorithm; S211. For the first patient wound image in the current patient file in the patient information database, mark it as the initial image and perform color space conversion. At the same time, reduce skin texture noise through Gaussian smoothing filter to obtain the preprocessed initial image I. It should be noted that color space conversion can highlight features of the original image, including LAB color space, HSV color space, and grayscale processing. The specific choice needs to be based on the actual wound condition. Wounds and sutures usually have high contrast in the chroma channel. The LAB color space separates the luminance (L) and chroma (A, B) channels to facilitate subsequent feature extraction. The HSV color space can separate tissue and sutures through saturation and chroma.

[0021] S212. For the preprocessed initial image I, the horizontal and vertical gradients of the initial image I are calculated using the Sobel operator. Local maxima are preserved and edges are refined using non-maximum suppression. Edge points of the initial image are identified using double thresholding and an edge map is output. ; It should be noted that the Sobel operator calculates the gradients in the horizontal and vertical directions of the initial image I. The algorithm calculates the gradient magnitude and direction, and uses non-maximum suppression to compare the gradient magnitude of neighboring pixels along the gradient direction to preserve local maxima. It sets high and low thresholds, marks pixels with gradient magnitudes greater than the high threshold as strong edge points, suppresses pixels with gradient magnitudes less than the low threshold, and retains pixels in between only when connecting strong edges. The high and low thresholds need to be set in advance based on empirical methods. In actual use, the characteristics of the actual image and the expected edge detection are observed, and the thresholds are adjusted in combination with the edge detection results to obtain the high and low thresholds.

[0022] S213, Regarding the obtained edge map Extract all contours to obtain a contour set. , where each contour Composed of continuous edge pixels, the contour is determined to identify wounds and suture ends, specifically including the following steps: The profile is calculated using the sum of Euclidean distances. The distance between the middle and edge pixels is used to obtain the length. Calculate the aspect ratio of the minimum bounding rectangle. ,in Calculate the curvature features based on the width and height of the rectangle, respectively. ,in Represents the curvature of each point on the outline; It should be noted that on a two-dimensional image plane, wound edges typically appear as long, curved, irregular curves. Furthermore, due to the irregularity of human tissue and the tearing nature of wounds, wound boundaries are usually curved and rough. Sutures, being linear foreign bodies, appear as short straight segments or arc segments with very little curvature due to the physical properties of lines and the imaging angle; their circumscribed rectangles are typically long and narrow. Represents the curvature of each point on the contour. The average curvature at each point, , where (x, y) are the coordinates of the contour point, and the point represents the derivative with respect to the arc length.

[0023] when and At that time, Marked as candidate wound edges, where These represent the length threshold and the high curvature threshold, respectively. when and or ,and At that time, Marked as candidate suture ends, among which These represent the short length threshold, aspect ratio threshold, and low curvature threshold, respectively. It should be noted that the length threshold defines the minimum length required for a contour to be considered "long enough to be a wound edge," and contours shorter than this value will not be considered as candidate wound edges. The high curvature threshold defines the minimum average curvature required for a contour to be considered "curved enough to be a wound edge," the short length threshold defines the maximum length required for a contour to be considered "short enough to be a thread end," and the aspect ratio threshold is used to determine the elongated shape of the contour. The representation is very broad. The representation is very narrow, and the low curvature threshold is used to define the maximum average curvature required for a contour to be considered "sufficiently straight and possibly a thread end". By collecting historical data under different wound length ranges, the preset thresholds under different ranges are determined by the Raida criterion, and the current patient's wound length is combined to divide into the corresponding length range to obtain the preset threshold.

[0024] S214. For each edge pixel of the candidate wound edge on the contour, take a neighborhood of width d on both sides along the normal direction, convert the two sides to the HSV color space, extract the histograms p and q of the H channel, and calculate the histogram difference using Bach distance. ,when If the difference exceeds the difference threshold, the current contour is confirmed as the wound edge; It should be noted that, The Bach distance is an indicator of the similarity between two probability distributions; a larger value indicates a greater difference between the two distributions. The H-value distributions on both sides of the wound edge are significantly different, with one side biased towards red and the other towards skin tone. Therefore, the calculated... It will be very big, when If the difference is greater than the difference threshold, it means that the colors on both sides of the line do indeed belong to different categories, which meets the definition of a wound edge.

[0025] Extract the minimum bounding rectangle region corresponding to the contour of the candidate suture thread, and calculate the average saturation of this region in the HSV color space. ,in Let N be the saturation value of pixel k, and N be the number of pixels in the region. The directional variance is also calculated. ,in These represent the gradient direction and the mean gradient direction, respectively. Less than the saturation threshold and If the value is less than the direction threshold, then the current contour is confirmed as a stitching thread.

[0026] It should be noted that suture lines are usually made of materials with relatively pure colors. In the HSV color space, high-purity, vivid colors have high saturation, while white, gray, or black objects have very low saturation. That is, the saturation value inside the suture line area will be relatively pure. At the same time, the suture line usually appears as a thin, long structure with a consistent direction in the image. The image gradient direction describes the direction of edge brightness change at each pixel point. The direction variance measures the degree of dispersion of the gradient direction of all pixels in the region. The smaller the variance, the more likely that the edge direction of all points is almost the same. The specific difference threshold, direction threshold, and saturation threshold need to be set by consulting experts in the field and combining empirical methods.

[0027] S22. Project the identified wound lines and thread ends onto a blank image. For the initial image after projection, record the initial wound line pixel set and the initial thread end pixel set. Sequentially number the thread ends of the current patient's wound and record them in the corresponding patient file. S221. Add all pixels whose current contour is confirmed to be the edge of the wound to the wound edge mask. Add all pixels that confirm the current outline is the start of the stitch line to the stitch line mask. Create a solid color image with the same dimensions as the original captured image, and then mask it. , All pixels with a value of 1 are projected onto a solid color image in different colors; It should be noted that the specific projection color can be selected according to the actual situation.

[0028] S222, Record the wound edge mask obtained from the initial image. All pixels with a value of 1 represent the initial set of wound line pixels, recording the suture mask obtained from the initial image. All pixels with a value of 1 in the set are the initial line start pixel set. Connectivity component analysis is performed to separate spatially connected pixel regions, which are then numbered sequentially and recorded in the corresponding patient files.

[0029] It should be noted that connected component analysis can group all adjacent pixels with the same pixel value in a binary image into one group. When performing connected component analysis, 8-adjacency is used to ensure that pixels that are diagonally connected can also be identified as the same line. The number of elements in the pixel set is the number of pixels.

[0030] Example 2:

[0031] S3. Define the surrounding skin tissue area in the initial line of the initial image. Based on the number of pixels of the initial line and the number of pixels of the surrounding tissue, and combined with the ratio change rate of different numbered lines in the subsequent captured images, determine the embedding degree of different numbered lines. At the same time, mark the number of detachable lines daily according to the number of pixels of the wound lines in the area of ​​different numbered lines. S31. Define the surrounding skin tissue region in the initial line of the initial image. With each pixel of the line as the center, expand outward by a fixed pixel distance and exclude the line pixel itself to obtain the set of surrounding region pixels as the surrounding tissue pixels. S311, Initial line pixel set for different numbers ,in Create a binary mask representing the set of all pixel coordinates of the line ending with ID id in the initial image. if and only if pixel coordinates hour Otherwise, it is 0; right Obtain by performing morphological dilation operations The expanding structured element is a disk with radius r. The pixels in the middle are to The set of pixels whose Euclidean distance to any pixel is less than or equal to r is subtracted from the line head body region from the dilated region to obtain the surrounding tissue region mask. ; S312, Extract the mask of the surrounding tissue region. The coordinates of all pixels with a value of 1 constitute the initial peripheral organization pixel set. ,in .

[0032] It should be noted that by extracting the pixel set of the tissue around the thread end and combining it with the pixel set of thread ends with different numbers, the embedding situation of the subsequent thread ends can be analyzed and determined, thereby determining the thread disassembly status. The radius r can be set based on empirical methods according to actual needs.

[0033] S32. Based on the ratio of the initial number of thread head pixels to the number of surrounding tissue pixels, and based on the rate of change of the proportion of different numbered thread heads in subsequent captured images, the degree of thread head embedding is determined. At the same time, based on the daily concentration of wound line pixels in different thread head areas compared with the initial concentration of wound line pixels, the number of detachable thread heads is marked daily. S321, Initial number of pixels for each line head ID Initial pixel count of the surrounding tissue To obtain the initial ratio The number of pixels at each line end on day t is obtained by processing the image on day t using the same process as S21-S22. 1. Calculate the pixel set of the wound lines to obtain the line proportion on day t. ,in The number of skin tissue pixels re-identified by color segmentation within the initial surrounding tissue area; It should be noted that, for The number of skin pixels re-identified daily within the original fixed coordinate range through color segmentation can reflect the dynamic process of tissue growth and coverage.

[0034] Calculate the current rate of change. Based on the risk warning threshold embedded in the thread end and embedded confirmation threshold Determine the different thread ends with different numbers each day: when This indicates that the visibility of the thread ends has not changed significantly, and there are no obvious signs of embedding. when This indicates that the visibility of the thread ends has begun to decrease, and the tissue is starting to cover the thread ends; when This indicates that the thread end is embedded; It should be noted that, , It is usually set to -0.2 or -0.4, but can also be adjusted based on experience according to the actual situation. , Adjustments will be made.

[0035] S322. Combining the pixel set of the wound lines in the initial image and the pixel set of the wound lines on day t, the wound region is divided into different line regions based on a fixed radius, and the number of pixels is counted to calculate the wound contraction rate. ,in The initial number of image wound pixels for the line with the ID number, given a fixed radius. The number of wound pixels on day t under a fixed radius for the line with the ID number; It should be noted that, since wounds grow at different locations at different rates, by dividing the wounds based on the suture end area at a fixed radius, the wound recovery status of the suture end areas can be analyzed separately, which will facilitate the subsequent determination of the conditions for suture removal.

[0036] S323. For all thread ends under day t, mark them as satisfying... and The wire end ID serves as the detachable wire end. This represents the wound healing threshold.

[0037] It should be noted that the wound healing threshold can be used to determine the healing status of wounds with different suture ends. The wound healing threshold needs to be set based on experience and in consultation with experts in the field, taking into account the location of the wound. By outputting the number of suture ends that can be removed each day, patients can easily monitor the wound growth status and make their own suture removal schedule and plan when caring for their wounds at home. This avoids delays in suture removal and secondary injury caused by embedded suture ends due to different individual constitutions and wound growth rates.

[0038] In summary, this invention collects wound data from different orthopedic patients, sets initial images after wound suturing, and combines these with daily images uploaded by patients to determine the condition of the suture ends. It intelligently determines whether different suture ends are ready for removal based on wound contraction in different suture end areas. This allows for flexible selection of suture removal dates based on each patient's individual wound growth. By determining the appropriate timing for suture removal according to each patient's condition, it can warn of the risk of suture embedding due to excessively rapid healing, prompting timely suture removal. Simultaneously, it can identify insufficient wound strength due to slow healing, preventing premature suture removal and thus effectively preventing secondary injury and infection risks.

[0039] 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 spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for wound care of orthopedic patients based on visual image processing, characterized in that, The method includes the following steps: S1. For orthopedic patients, a patient information database is built based on the patient's basic information to record the patient's basic information and wound information. Daily images of orthopedic patients' wounds after suturing are collected by image acquisition equipment and recorded in the patient information database. S2. For wound images in the patient information database, extract feature pixels from the wound images and project them onto a blank image. Label the initial wound line pixels and initial line head pixels for the initial image. Number the line heads and record the initial line head pixels and the number of pixels, as well as the initial wound line pixels and the number of pixels. S3. Define the surrounding skin tissue area in the initial line of the initial image. Based on the number of pixels of the initial line and the number of pixels of the surrounding tissue, and combined with the ratio change rate of different numbered lines in the subsequent captured images, determine the embedding degree of different numbered lines. At the same time, mark the number of detachable lines daily according to the number of pixels of the wound lines in the area of ​​different numbered lines.

2. The method for wound care of orthopedic patients based on visual image processing according to claim 1, characterized in that, S1 includes the following steps: S11. Collect basic information and wound information of orthopedic patients, including patient name, contact information, age, wound location, wound length, surgery date, suture date and expected suture removal date. Establish a patient information database through MySQL and create files based on patient name to record the collected information. S12. Collect daily images of the patient's wound suture location using an image acquisition device, and upload the images with the date annotation to the corresponding patient file in the patient information database.

3. The method for wound care of orthopedic patients based on visual image processing according to claim 1, characterized in that, S2 includes the following steps: S21. Extract the first patient wound image from the current patient file in the patient information database as the initial image, and identify the wound and suture ends through the edge detection algorithm; S22. Project the identified wound lines and thread ends onto a blank image. For the initial image after projection, record the initial wound line pixel set and the initial thread end pixel set. Then, number the thread ends of the current patient's wound sequentially and record them in the corresponding patient file.

4. The method for wound care of orthopedic patients based on visual image processing according to claim 3, characterized in that, S21 includes the following steps: S211. For the first patient wound image in the current patient file in the patient information database, mark it as the initial image and perform color space conversion. At the same time, reduce skin texture noise through Gaussian smoothing filter to obtain the preprocessed initial image I. S212. For the preprocessed initial image I, the horizontal and vertical gradients of the initial image I are calculated using the Sobel operator. Local maxima are preserved and edges are refined using non-maximum suppression. Edge points of the initial image are identified using double thresholding and an edge map is output. ; S213, Regarding the obtained edge map Extract all contours to obtain a contour set. , where each contour Composed of continuous edge pixels, the contour is determined to identify wounds and suture ends, specifically including the following steps: The profile is calculated using the sum of Euclidean distances. The distance between the middle and edge pixels is used to obtain the length. Calculate the aspect ratio of the minimum bounding rectangle. ,in Calculate the curvature features based on the width and height of the rectangle, respectively. ,in Represents the curvature of each point on the outline; when and At that time, Marked as candidate wound edges, where These represent the length threshold and the high curvature threshold, respectively. when and or ,and At that time, Marked as candidate suture ends, among which These represent the short length threshold, aspect ratio threshold, and low curvature threshold, respectively. S214. For each edge pixel of the candidate wound edge on the contour, take a neighborhood of width d on both sides along the normal direction, convert the two sides to the HSV color space, extract the histograms p and q of the H channel, and calculate the histogram difference using Bach distance. ,when If the difference exceeds the difference threshold, the current contour is confirmed as the wound edge; Extract the minimum bounding rectangle region corresponding to the contour of the candidate suture thread, and calculate the average saturation of this region in the HSV color space. ,in Let N be the saturation value of pixel k, and N be the number of pixels in the region. The directional variance is also calculated. ,in These represent the gradient direction and the mean gradient direction, respectively. Less than the saturation threshold and If the value is less than the direction threshold, then the current contour is confirmed as a stitching thread.

5. A method for wound care of orthopedic patients based on visual image processing according to claim 4, characterized in that, S22 includes the following steps: S221. Add all pixels whose current contour is confirmed to be the edge of the wound to the wound edge mask. Add all pixels that confirm the current outline is the start of the stitch line to the stitch line mask. Create a solid color image with the same dimensions as the original captured image, and then mask it. , All pixels with a value of 1 are projected onto a solid color image in different colors; S222, Record the wound edge mask obtained from the initial image. All pixels with a value of 1 represent the initial set of wound line pixels, recording the suture mask obtained from the initial image. All pixels with a value of 1 in the set are the initial line start pixel set. Connectivity component analysis is performed to separate spatially connected pixel regions, which are then numbered sequentially and recorded in the corresponding patient files.

6. A method for wound care of orthopedic patients based on visual image processing according to claim 5, characterized in that, S3 includes the following steps: S31. Define the surrounding skin tissue region in the initial line of the initial image. With each pixel of the line as the center, expand outward by a fixed pixel distance and exclude the line pixel itself to obtain the set of surrounding region pixels as the surrounding tissue pixels. S32. Based on the ratio of the initial number of thread head pixels to the number of surrounding tissue pixels, and based on the rate of change of the proportion of different numbered thread heads in subsequent captured images, the degree of thread head embedding is determined. At the same time, based on the daily concentration of wound line pixels in different thread head areas compared with the initial concentration of wound line pixels, the number of detachable thread heads is marked daily.

7. A method for wound care of orthopedic patients based on visual image processing according to claim 6, characterized in that, S31 includes the following steps: S311, Initial line pixel set for different numbers ,in Create a binary mask representing the set of all pixel coordinates of the line ending with ID id in the initial image. if and only if pixel coordinates hour Otherwise, it is 0; right Obtain by performing morphological dilation operations The expanding structured element is a disk with radius r. The pixels in the middle are to The set of pixels whose Euclidean distance to any pixel is less than or equal to r is subtracted from the line head body region from the dilated region to obtain the surrounding tissue region mask. ; S312, Extract the mask of the surrounding tissue region. The coordinates of all pixels with a value of 1 constitute the initial peripheral organization pixel set. ,in .

8. A method for wound care of orthopedic patients based on visual image processing according to claim 7, characterized in that, S32 includes the following steps: S321, Initial number of pixels for each line head ID Initial pixel count of the surrounding tissue To obtain the initial ratio The number of pixels at each line end on day t is obtained by processing the image on day t using the same process as S21-S22.

1. Calculate the pixel set of the wound lines to obtain the line proportion on day t. ,in The number of skin tissue pixels re-identified by color segmentation within the initial surrounding tissue area; Calculate the current rate of change. Based on the risk warning threshold embedded in the thread end and embedded confirmation threshold Determine the different thread ends with different numbers each day: when This indicates that the visibility of the thread ends has not changed significantly, and there are no obvious signs of embedding. when This indicates that the visibility of the thread ends has begun to decrease, and the tissue is starting to cover the thread ends; when This indicates that the thread end is embedded; S322. Combining the pixel set of the wound lines in the initial image and the pixel set of the wound lines on day t, the wound region is divided into different line regions based on a fixed radius, and the number of pixels is counted to calculate the wound contraction rate. ,in The initial number of image wound pixels for the line with the ID number, given a fixed radius. The number of wound pixels on day t under a fixed radius for the line with the ID number; S323. For all thread ends under day t, mark them as satisfying... and The wire end ID serves as the detachable wire end. This represents the wound healing threshold.

Images