Pressure sore staging judgment method and device based on image recognition

Abstract

The invention provides a pressure sore staging judgment method and device based on image recognition, and the method comprises the steps: segmenting a pressure sore image from a collected image of an affected part; identifying pressure sore features from the pressure sore image; wherein the pressure sore characteristics comprise pressure sore color, tissue type and area and skin integrity; the tissue types at least comprise spots, serous vesicles, granulation tissues, fat, slough, eschar, fascia muscles, tendons, ligaments, cartilages and bones; and determining stages of the pressure sores according to the identified features of the pressure sores. According to the invention, the pressure sore staging can be automatically identified based on the pressure sore features in the pressure sore image, and the accuracy and efficiency of pressure sore staging identification are improved.

Description

Technical Field

[0001] The present application relates to the technical field of medical data processing, and in particular to a method and device for determining the stages of pressure sores based on image recognition. Background Art

[0002] Pressure injury, commonly known as pressure sore, is the ulceration and necrosis of soft tissue caused by long-term pressure on local tissue, blood circulation disorder, continuous ischemia, hypoxia and malnutrition. Pressure sores often occur in the elderly, the weak and malnourished anemic patients, diabetics, obese people, people with unclear consciousness and impaired consciousness, long-term bedridden or paralyzed patients, edema patients, fever patients, pain patients, incontinence patients, restricted activities and advanced tumor patients. In practice, a large number of patients suffer from pressure sores.

[0003] When pressure sores are found on the patient's skin surface, necessary treatment and care should be provided in a timely manner according to the severity of the pressure sores to relieve the patient's pain. In 2023, the National Pressure Ulcer Advisory Committee (NPUAP) divided the stages of pressure injuries into: stage 1, stage 2, stage 3, stage 4, unclear stage, and deep tissue injury. Accurately judging the stage of a patient's pressure sores is of great guiding significance for the subsequent treatment of pressure sores. At present, the stage of pressure sores is generally judged based on the work experience of medical staff, which requires a lot of material and human resources, and there is an urgent need to improve the efficiency of pressure sore staging judgment. Summary of the invention

[0004] In view of this, the purpose of the embodiments of the present application is to provide a method and device for determining pressure ulcer staging based on image recognition, which can automatically identify the pressure ulcer staging based on the pressure ulcer features in the pressure ulcer image, thereby improving the accuracy and efficiency of pressure ulcer staging identification.

[0005] In a first aspect, an embodiment of the present application provides a method for determining the stages of pressure sores based on image recognition, the method comprising: Segmenting a pressure sore image from the acquired affected area image; Identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue type includes at least plaque, serous blister, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone; Based on the identified pressure ulcer characteristics, the stage of the pressure ulcer is determined.

[0006] In a possible implementation, segmenting the pressure sore image from the acquired affected area image includes: Performing semantic segmentation on the affected area image, and assigning a semantic label to each pixel in the affected area image; wherein the semantic label includes healthy skin and pressure sore skin; Target detection is performed on pixels with a semantic label of pressure ulcer skin to detect the location of each pressure ulcer, and a segmentation mask is used to represent the area where each pressure ulcer is located, thereby segmenting at least one pressure ulcer image.

[0007] In a possible implementation, identifying pressure sore features from the pressure sore image includes: Normalizing the RGB value of each pixel in the affected area image to obtain a normalized RGB value of each pixel; Generating a color histogram of the pressure sore image according to the normalized RGB value of each pixel in the pressure sore image; The color of the pressure sore is determined according to the normalized RGB values ​​of the pixels labeled as healthy skin in the affected area image and the color histogram of the pressure sore image.

[0008] In a possible implementation, identifying pressure sore features from the pressure sore image further includes: Performing grayscale processing on the RGB value of each pixel in the pressure sore image to obtain a grayscale value of each pixel; Using an artificial neural network model, based on the gray value of each pixel, identifying the tissue type from the pressure sore image, as well as the pixel proportion of each tissue type in the pressure sore image; For each tissue type, determine the area of ​​the tissue type according to the pixel ratio of the tissue type in the pressure sore image and the total area of ​​the pressure sore; The skin integrity is determined based on the area of ​​each tissue type and the total area of ​​the pressure ulcer.

[0009] In a possible implementation, the total area of ​​the pressure sores is calculated in the following manner: the total area of ​​the pressure sores is determined according to the size of the pressure sore image and the distance between the acquisition device of the affected part image and the affected part.

[0010] In a possible implementation, determining the stage of the pressure sore according to the identified pressure sore characteristics includes: The pressure sore feature is input into a classifier, and the classifier determines the stage of the pressure sore based on the matching degree of the input pressure sore feature with the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury.

[0011] In a possible implementation, determining the stage of the pressure sore according to the identified pressure sore characteristics further includes: The Kmeans clustering method was used with the K value set to 6 to cluster the pressure sore characteristics and the preset pressure sore characteristics of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and the stage of pressure sores was determined based on the clustering results.

[0012] In a second aspect, an embodiment of the present application provides a device for determining the stages of pressure sores based on image recognition, the device comprising: A pressure sore image segmentation module, used for segmenting a pressure sore image from the collected affected area image; A pressure sore feature extraction module, used to identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue type includes at least plaque, serous blister, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone; The pressure sore staging identification module is used to determine the stage of pressure sores based on the identified pressure sore characteristics.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory through the bus, and the processor executes the machine-readable instructions to perform the steps of the pressure ulcer staging determination method based on image recognition as described in any one of the first aspects.

[0014] In a fourth aspect, an embodiment of the present application 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 for determining the pressure ulcer staging based on image recognition as described in any one of the first aspects are executed.

[0015] The pressure sore staging judgment method and device based on image recognition provided in the embodiments of the present application can automatically identify the pressure sore staging based on the pressure sore features in the pressure sore image, thereby improving the accuracy and efficiency of pressure sore staging recognition.

[0016] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A flow chart of a method for determining pressure sores by staging based on image recognition provided in an embodiment of the present application is shown; Figure 2 A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application is shown; Figure 3 A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application is shown; Figure 4 A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application is shown; Figure 5 A schematic diagram of the structure of a device for determining pressure sores by staging based on image recognition provided in an embodiment of the present application is shown; Figure 6 A schematic diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0019] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0020] Pressure injury, commonly known as pressure sore, is the ulceration and necrosis of soft tissue caused by long-term pressure on local tissue, blood circulation disorder, continuous ischemia, hypoxia and malnutrition. Pressure sores often occur in the elderly, the weak and malnourished anemic patients, diabetics, obese people, people with unclear consciousness and impaired consciousness, long-term bedridden or paralyzed patients, edema patients, fever patients, pain patients, incontinence patients, restricted activities and advanced tumor patients. In practice, a large number of patients suffer from pressure sores.

[0021] When pressure sores are found on the patient's skin surface, necessary treatment and care should be provided in a timely manner according to the severity of the pressure sores to relieve the patient's pain. In 2023, the National Pressure Ulcer Advisory Committee (NPUAP) divided the stages of pressure injuries into: stage 1, stage 2, stage 3, stage 4, unclear stage, and deep tissue injury. Accurately judging the stage of a patient's pressure sores is of great guiding significance for the subsequent treatment of pressure sores. At present, the stage of pressure sores is generally judged based on the work experience of medical staff, which requires a lot of material and human resources, and there is an urgent need to improve the efficiency of pressure sore staging judgment.

[0022] Based on the above problems, an embodiment of the present application provides a method and device for determining pressure sore staging based on image recognition, which can automatically identify the pressure sore staging based on the pressure sore features in the pressure sore image, thereby improving the accuracy and efficiency of pressure sore staging identification.

[0023] The defects existing in the above solutions are the results obtained by the inventor after practice and careful research. Therefore, the process of discovering the above problems and the solutions proposed by this application for the above problems below should be the contributions made by the inventor to this application during the application process.

[0024] The technical solutions in the present application will be clearly and completely described below in conjunction with the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0025] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0026] To facilitate understanding of this embodiment, a method for determining pressure ulcer stages based on image recognition disclosed in an embodiment of the present application is first introduced in detail.

[0027] See also Figure 1 As shown, Figure 1 A flowchart of a method for determining the stages of pressure sores based on image recognition provided in an embodiment of the present application, the method comprising the following steps: S101, segmenting a pressure sore image from the collected affected area image.

[0028] The image of the patient's skin surface is collected using a smart phone, a digital camera, a tablet or other collection device. The collected image is called an affected area image. The affected area image includes both the patient's healthy skin and pressure sores. In order to reduce the amount of image data processing, the affected area image is segmented and the pressure sore image is segmented from the affected area image. The affected area image may include multiple pressure sores. Therefore, multiple pressure sore images can be segmented from the affected area image, and the number of pressure sore images is consistent with the number of pressure sores contained in the affected area image.

[0029] S102. Identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue types include at least plaques, serous blisters, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone.

[0030] Pressure sore images contain pressure sore features that can indicate the stage of pressure sores, such as pressure sore color, tissue type, tissue area, and skin integrity. Tissue type refers to the spots, serous blisters, and eschars on the surface of the pressure sore, as well as the exposed granulation tissue, fat, slough, fascia muscle, tendon, ligament, cartilage, and bone of the pressure sore; tissue area refers to the total area of ​​each tissue type; the skin on the pressure sore may be damaged, and the more damaged it is, the more serious the pressure sore is; skin integrity refers to the area ratio of intact skin in the pressure sore.

[0031] S103. Determine the stage of the pressure sore based on the identified characteristics of the pressure sore.

[0032] The stages of pressure sores include stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury. Stage 1 features: non-blanchable erythema, intact skin without damage; stage 2 features: partial skin loss with exposed dermis, active wound bed, pink or red moist, and may also appear as intact or broken serous blisters; stage 3 features: full-thickness skin loss, often with fat, granulation tissue and involuted edges, visible slough and / or eschar; stage 4 features: full-thickness skin and tissue loss, visible or directly accessible fascia, muscle, tendon, ligament, cartilage or bone, visible slough and / or eschar, often with involuted edges, sinus tracts and / or undermining; unstageable features: full-thickness skin and tissue loss, the extent of tissue loss cannot be confirmed due to being covered by slough and / or eschar; deep tissue injury features: intact or broken local skin with persistent non-blanchable dark red, maroon or purple, or epidermal separation presenting a black wound bed or blood-filled blisters

[0033] The pressure sore features identified from the pressure sore images were matched with the common pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and the stage with the highest matching degree was used as the stage of the pressure sore.

[0034] The pressure sore staging judgment method based on image recognition provided in the embodiment of the present application can automatically identify the pressure sore staging based on the pressure sore features in the pressure sore image, thereby improving the accuracy and efficiency of pressure sore staging identification.

[0035] For further information, see Figure 2 As shown, Figure 2 A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application, wherein step S101: segmenting a pressure sore image from a collected affected area image comprises: S1011. Perform semantic segmentation on the affected area image, and assign a semantic label to each pixel in the affected area image; wherein the semantic labels include healthy skin and pressure ulcer skin.

[0036] Semantic segmentation is to assign a semantic label (category) to each pixel in the image, but does not distinguish between object instances in a single category. In the embodiment of the present application, two types of semantic labels are set for pixels, namely healthy skin and pressure sore skin. Here, if the image of the affected area includes two pressure sores, the semantic label of each pixel of the two pressure sores is pressure sore skin. The use of semantic labels cannot distinguish between different pressure sores, but can only distinguish between healthy skin and pressure sore skin.

[0037] S1012. Perform target detection on pixels with a semantic label of pressure sore skin, detect the location of each pressure sore, use a segmentation mask to represent the area where each pressure sore is located, and segment at least one pressure sore image.

[0038] Target detection is to detect the location (Localization) of an object in an image and determine the category (Classification) of the object. In the embodiment of the present application, step S1011 has identified the skin category (healthy skin, pressure sore skin) in the affected area image, and step S1012 only needs to distinguish the object instances of the pressure sore image, and only uses the "detect object location" function in target detection to perform target detection on pixels with the semantic label of pressure sore skin, identify pixels belonging to different pressure sores, and then segment different pressure sore images.

[0039] Optionally, target detection is used to detect the position of each pixel in the affected area image and determine the object category. If the object category determined by target detection is consistent with the semantic label of semantic segmentation, the pixel classification is determined to be correct. For example, for a certain pixel, the semantic label of the pixel is pressure sore skin, and the object category of target detection is also pressure sore skin, then the pixel is determined to be a pressure sore pixel. The category to which the same pixel belongs is determined twice to improve the accuracy of pressure sore image segmentation.

[0040] For further information, see Figure 3 As shown, Figure 3 A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application, wherein step S102: identifying pressure sore features from the pressure sore image includes: S10211. Normalize the RGB value of each pixel in the affected area image to obtain a normalized RGB value of each pixel.

[0041] Normalizing the pixel values ​​of an image can adjust the pixel values ​​to a similar range. The RGB (color system) pixel value is a 1×3 dimensional vector. Each pixel of RGB has 3 components (also called 3 channels), so RGB can be used to represent color images. Specifically, the following formula is used for normalization:

[0042] output=[input-min(input)] / [max(input)-min(input)] Among them, output is the normalized RGB value of the pixel, input is the RGB value of the pixel, max(input) is the maximum RGB value of each pixel in the affected area image, and min(input) is the minimum RGB value of each pixel in the affected area image.

[0043] S10212. Generate a color histogram of the pressure sore image according to the normalized RGB value of each pixel in the pressure sore image.

[0044] Color histogram is a global color feature of an image. Given a pressure ulcer image (f ij )m×n, f ij Represents the RGB value of the image at the pixel point (i, j), m×n represents the size of the image, the color set contained in the image is C, and the ordinate value of each abscissa point c in the color histogram is h c , the color histogram is generated using the following formula: S10213. Determine the color of the pressure sore according to the normalized RGB values ​​of the pixels labeled as healthy skin in the affected area image and the color histogram of the pressure sore image.

[0045] Since the color of the pressure sore will be affected by the light and sunlight when the acquisition device collects the pressure sore image, the color histogram may be distorted. In order to improve the accuracy of the pressure sore color, the color histogram is corrected using the normalized RGB value of the pixels of healthy skin. Specifically, the average value of the normalized RGB value of the pixels labeled as healthy skin in the affected area image is calculated as the healthy skin reference value, and the calculation formula in step S1023 is used to calculate the h corresponding to the healthy skin reference value. c , and use the calculated result as the correction value h of the color histogram c0 , according to the color histogram h c With h c0 The difference between the two values ​​determines the color of the pressure sore.

[0046] For further information, see Figure 4 As shown, Figure 4A flowchart of another method for determining pressure sore stages based on image recognition provided in an embodiment of the present application, wherein step S102: identifying pressure sore features from the pressure sore image, further comprises: S10221. Perform grayscale processing on the RGB value of each pixel in the pressure sore image to obtain a grayscale value of each pixel.

[0047] Grayscale values ​​divide white and black into several levels according to the degree of lightness and darkness, which is called grayscale. Grayscale values ​​represent the degree of lightness and darkness. The value range is generally [0,255]. The higher the grayscale value, the brighter it is.

[0048] The pressure sore image is grayed to obtain a grayscale image. The pressure sore image includes three channels, while the grayscale image only contains one channel, and the channel value represents the grayscale value. In the grayscale process, the RGB value of each pixel is combined into a single grayscale value, which can effectively reduce the matrix dimension, reduce the amount of data processing, and greatly improve the operation speed.

[0049] Specifically, any of the following methods can be used to grayscale the RGB values: weighted average method: perform weighted average of the red, green and blue color channels according to different weights to obtain a grayscale value; average method: add the values ​​of the red, green and blue color channels and divide them by 3 to obtain a grayscale value; maximum method: take the maximum value of the red, green and blue color channels as the grayscale value; minimum method: take the minimum value of the red, green and blue color channels as the grayscale value.

[0050] S10222. Using an artificial neural network model, based on the grayscale value of each pixel, identify the tissue type from the pressure sore image, as well as the pixel ratio of each tissue type in the pressure sore image.

[0051] Grayscale images of tissue types of pressure sores at various stages are used as training samples to train the initial artificial neural network model, and an artificial neural network model for identifying pressure sore tissue types from grayscale images is obtained. The grayscale value of each pixel of the pressure sore image (i.e., grayscale image) is input into the trained artificial neural network model, and the model outputs the recognition result of the pressure sore tissue type, and also outputs the pixel ratio of the tissue type in the pressure sore image. Here, the pixel ratio is the ratio of the number of pixels of the tissue type to the number of pixels of the pressure sore image.

[0052] S10223. For each tissue type, determine the area of ​​the tissue type according to the pixel ratio of the tissue type in the pressure sore image and the total area of ​​the pressure sore.

[0053] The area of ​​each tissue type identified by the artificial neural network model was calculated separately, and the product of the total area of ​​the pressure sore and the pixel ratio of the tissue type in the pressure sore image was taken as the area of ​​the tissue type.

[0054] S10224. Determine the skin integrity based on the area of ​​each tissue type and the total area of ​​the pressure sore.

[0055] The ratio of the sum of the areas of each tissue type in the pressure sore image to the total area of ​​the pressure sore was used as the calculation result of the skin integrity.

[0056] It should be noted that the total area of ​​the pressure sore is calculated in the following manner: the total area of ​​the pressure sore is determined according to the size of the pressure sore image and the distance between the acquisition device of the affected part image and the affected part. Specifically, the scale of the pressure sore image is determined according to the distance between the acquisition device of the affected part image and the affected part, and the total area of ​​the pressure sore is calculated based on the determined scale and the size of the pressure sore image (the total area of ​​the pressure sore image).

[0057] Furthermore, in the method for determining the stage of pressure sores based on image recognition provided in the embodiment of the present application, the step S103: determining the stage of pressure sores according to the identified pressure sore features, includes: The pressure sore feature is input into a classifier, and the classifier determines the stage of the pressure sore based on the matching degree of the input pressure sore feature with the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury.

[0058] The classifier calculates the matching degree between the input pressure sore feature and the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and takes the stage with the highest matching degree as the stage of the input pressure sore feature. Optionally, the pressure sore feature is a multidimensional vector, and the classifier calculates the distance between the input pressure sore feature and the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, respectively. The smaller the distance, the higher the matching degree, and the stage with the smallest distance is taken as the stage of the input pressure sore feature.

[0059] Furthermore, in the method for determining the stage of pressure sores based on image recognition provided in the embodiment of the present application, the step S103: determining the stage of pressure sores according to the identified pressure sore features, further includes: The Kmeans clustering method was used with the K value set to 6 to cluster the pressure sore characteristics and the preset pressure sore characteristics of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and the stage of pressure sores was determined based on the clustering results.

[0060] There are 6 stages of pressure sores. The characteristics of pressure sores in different stages are quite different. The characteristics of pressure sores can be divided into 6 categories. By judging the classification of pressure sore characteristics, the stage of the pressure sore characteristics can be judged. The preset pressure sore characteristics of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury are very typical characteristics of each stage. The Kmeans clustering method is used to divide the characteristics of pressure sores to be staged and the preset pressure sore characteristics of the 6 stages of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury into 6 categories. The pressure sore characteristics to be staged belong to the stage to which the preset pressure sore characteristics of the stage are classified into the same category.

[0061] Based on the same inventive concept, the embodiment of the present application also provides a pressure sore staging judgment device based on image recognition corresponding to the pressure sore staging judgment method based on image recognition. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned pressure sore staging judgment method based on image recognition in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0062] See also Figure 5 As shown, Figure 5 A schematic diagram of a device for determining the stages of pressure sores based on image recognition provided in an embodiment of the present application, the device comprising: The pressure sore image segmentation module 501 is used to segment the pressure sore image from the acquired affected area image; The pressure sore feature extraction module 502 is used to identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue type includes at least plaque, serous blister, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone; The pressure sore staging identification module 503 is used to determine the stage of the pressure sore according to the identified pressure sore characteristics.

[0063] In a possible implementation, the pressure sore image segmentation module 501, when segmenting the pressure sore image from the acquired affected part image, includes: Performing semantic segmentation on the affected area image, and assigning a semantic label to each pixel in the affected area image; wherein the semantic label includes healthy skin and pressure sore skin; Target detection is performed on pixels with a semantic label of pressure ulcer skin to detect the location of each pressure ulcer, and a segmentation mask is used to represent the area where each pressure ulcer is located, thereby segmenting at least one pressure ulcer image.

[0064] In a possible implementation, the pressure sore feature extraction module 502, when identifying the pressure sore feature from the pressure sore image, includes: Normalizing the RGB value of each pixel in the affected area image to obtain a normalized RGB value of each pixel; Generating a color histogram of the pressure sore image according to the normalized RGB value of each pixel in the pressure sore image; The color of the pressure sore is determined according to the normalized RGB values ​​of the pixels labeled as healthy skin in the affected area image and the color histogram of the pressure sore image.

[0065] In a possible implementation, the pressure sore feature extraction module 502, when identifying the pressure sore feature from the pressure sore image, further includes: Performing grayscale processing on the RGB value of each pixel in the pressure sore image to obtain a grayscale value of each pixel; Using an artificial neural network model, based on the gray value of each pixel, identifying the tissue type from the pressure sore image, as well as the pixel proportion of each tissue type in the pressure sore image; For each tissue type, determine the area of ​​the tissue type according to the pixel ratio of the tissue type in the pressure sore image and the total area of ​​the pressure sore; The skin integrity is determined based on the area of ​​each tissue type and the total area of ​​the pressure ulcer.

[0066] In a possible implementation, the pressure sore feature extraction module 502 is further used to determine the total area of ​​the pressure sore according to the size of the pressure sore image and the distance between the acquisition device of the affected part image and the affected part.

[0067] In a possible implementation, the pressure sore stage identification module 503, when determining the stage of the pressure sore according to the identified pressure sore characteristics, includes: The pressure sore feature is input into a classifier, and the classifier determines the stage of the pressure sore based on the matching degree of the input pressure sore feature with the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury.

[0068] In a possible implementation, the pressure sore staging identification module 503, when determining the stage of the pressure sore according to the identified pressure sore characteristics, further includes: The Kmeans clustering method was used with the K value set to 6 to cluster the pressure sore characteristics and the preset pressure sore characteristics of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and the stage of pressure sores was determined based on the clustering results.

[0069] The pressure sore staging judgment device based on image recognition provided in the embodiment of the present application can automatically identify the pressure sore staging based on the pressure sore features in the pressure sore image, thereby improving the accuracy and efficiency of pressure sore staging identification.

[0070] See also Figure 6 As shown, Figure 6 A schematic diagram of an electronic device provided in an embodiment of the present application, the electronic device 600 includes: a processor 601, a memory 602 and a bus 603, the memory 602 stores machine-readable instructions executable by the processor 601, when the electronic device is running, the processor 601 communicates with the memory 602 via the bus 603, the processor 601 executes the machine-readable instructions to perform the steps of the above-mentioned method for determining pressure ulcer staging based on image recognition.

[0071] Specifically, the above-mentioned memory 602 and processor 601 can be general-purpose memory and processor, which are not specifically limited here. When the processor 601 runs the computer program stored in the memory 602, it can execute the above-mentioned pressure sore staging judgment method based on image recognition.

[0072] Corresponding to the above-mentioned method for determining the stage of pressure sores based on image recognition, an embodiment of the present application further 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 above-mentioned method for determining the stage of pressure sores based on image recognition are executed.

[0073] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0074] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0075] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0076] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for 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 application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0077] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for determining the stages of pressure sores based on image recognition, characterized in that: The method comprises: Segmenting a pressure sore image from the acquired affected area image; Identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue type includes at least plaque, serous blister, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone; Based on the identified pressure ulcer characteristics, the stage of the pressure ulcer is determined.

2. The method for determining the stages of pressure sores based on image recognition according to claim 1, characterized in that: The step of segmenting the pressure sore image from the acquired affected area image comprises: Performing semantic segmentation on the affected area image, and assigning a semantic label to each pixel in the affected area image; wherein the semantic label includes healthy skin and pressure sore skin; Target detection is performed on pixels with a semantic label of pressure ulcer skin to detect the location of each pressure ulcer, and a segmentation mask is used to represent the area where each pressure ulcer is located, thereby segmenting at least one pressure ulcer image.

3. The method for determining the stages of pressure sores based on image recognition according to claim 2, characterized in that: The identifying the pressure sore feature from the pressure sore image comprises: Normalizing the RGB value of each pixel in the affected area image to obtain a normalized RGB value of each pixel; Generating a color histogram of the pressure sore image according to the normalized RGB value of each pixel in the pressure sore image; The color of the pressure sore is determined according to the normalized RGB values ​​of the pixels labeled as healthy skin in the affected area image and the color histogram of the pressure sore image.

4. The method for determining the stages of pressure sores based on image recognition according to claim 1, characterized in that: The identifying the pressure sore feature from the pressure sore image further includes: Performing grayscale processing on the RGB value of each pixel in the pressure sore image to obtain a grayscale value of each pixel; Using an artificial neural network model, based on the gray value of each pixel, identifying the tissue type from the pressure sore image, as well as the pixel proportion of each tissue type in the pressure sore image; For each tissue type, determine the area of ​​the tissue type according to the pixel ratio of the tissue type in the pressure sore image and the total area of ​​the pressure sore; The skin integrity is determined based on the area of ​​each tissue type and the total area of ​​the pressure ulcer.

5. The method for determining the stages of pressure sores based on image recognition according to claim 4 is characterized in that: The total area of ​​the pressure sores is calculated in the following manner: the total area of ​​the pressure sores is determined according to the size of the pressure sore image and the distance between the acquisition device of the affected part image and the affected part.

6. The method for determining the stages of pressure sores based on image recognition according to claim 1, characterized in that: Determining the stage of the pressure sore based on the identified pressure sore characteristics includes: The pressure sore feature is input into a classifier, and the classifier determines the stage of the pressure sore based on the matching degree of the input pressure sore feature with the preset pressure sore features of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury.

7. The method for determining the stages of pressure sores based on image recognition according to claim 1, characterized in that: Determining the stage of the pressure sore according to the identified pressure sore characteristics also includes: The Kmeans clustering method was used with the K value set to 6 to cluster the pressure sore characteristics and the preset pressure sore characteristics of stage 1, stage 2, stage 3, stage 4, unstageable, and deep tissue injury, and the stage of pressure sores was determined based on the clustering results.

8. A device for determining the stages of pressure sores based on image recognition, characterized in that: The device comprises: A pressure sore image segmentation module, used to segment a pressure sore image from the collected affected area image; A pressure sore feature extraction module, used to identify pressure sore features from the pressure sore image; wherein the pressure sore features include pressure sore color, tissue type and area, and skin integrity; the tissue type includes at least plaque, serous blister, granulation tissue, fat, slough, eschar, fascia muscle, tendon, ligament, cartilage, and bone; The pressure sore staging identification module is used to determine the stage of pressure sores based on the identified pressure sore characteristics.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the pressure ulcer staging determination method based on image recognition as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of the method for determining pressure ulcer staging based on image recognition as described in any one of claims 1 to 7.

Images