A method, system and device for dynamic monitoring of wound surface for burn treatment

Abstract

The present application relates to a method, system and device for dynamic monitoring of wounds used in burn treatment, which belongs to the field of dynamic monitoring technology, wherein the method includes obtaining an initial burn image set of each target part of the patient; judging the degree of burns of the patient based on all the initial burn image sets to obtain key parts; periodically obtaining stage wound images of each key part of the patient; evaluating the recovery of the wound based on all stage wound images and the initial burn images; when the evaluation result is suspected infection or scar hyperplasia, triggering an adverse reaction prediction mechanism, and triggering an alarm based on the prediction result. The present application can more comprehensively track and monitor the dynamic changes of burn wounds of different patients in the subsequent recovery process, and promptly discover problems such as infection and poor healing of the wound.

Description

Technical Field

[0001] The present application relates to the field of dynamic monitoring technology, and in particular to a method, system and device for dynamic monitoring of wound surfaces used in burn treatment. Background Art

[0002] Burns are a common and serious accidental injury that not only causes great physical pain to patients, but may also lead to long-term disability and psychological trauma. In traditional burn treatment, the attending physician relies on clinical observation and experience to judge the severity of the patient's burns when the patient seeks medical treatment for burns, and gives a corresponding treatment plan based on the judgment; however, in reality, the complexity of burn treatment lies in the slow recovery process and uncertainty, and the treatment plan needs to be adjusted in real time according to the patient's recovery. For patients with moderate and above burns, their burn wounds are large and deep, and the skin's barrier function is severely damaged. At the same time, the patient's immunity will drop significantly after the burn, all of which make the patient extremely vulnerable to invasion by pathogens such as bacteria and fungi.

[0003] At present, medical institutions mainly equip hospitalized patients with professional nursing staff to regularly observe the recovery of their wounds (regularly changing dressings, keeping the wounds clean, promoting wound healing, etc.); however, wound infection and other conditions are almost imperceptible to the naked eye in the early stage, and their development speed will be very rapid. If they are not discovered and treated in time, they may cause the patient's wound to deteriorate and deepen, prolong the patient's hospitalization time, and even endanger the patient's life. This also leads to the nursing work of burn patients relying more on the experience and professional level of the nursing staff. This method of relying on manual observation has certain limitations. On the one hand, it is impossible for nursing staff to stay by the patient's side for uninterrupted observation at all times, which is prone to monitoring loopholes; on the other hand, there are differences in experience and professional level of different nursing staff, which may lead to inaccurate judgments on wound infection and other conditions, thereby affecting the treatment effect of patients. Summary of the invention

[0004] In order to improve the problem of not being able to monitor the wound surface in real time and having low accuracy by observing the wound surface condition by nursing staff to achieve dynamic monitoring of the wound surface, the present application provides a method, system and device for dynamic monitoring of the wound surface for burn treatment.

[0005] In a first aspect, the present application provides a method for dynamic monitoring of wounds for burn treatment, which adopts the following technical scheme: a method for dynamic monitoring of wounds for burn treatment, comprising: obtaining an initial burn image set of each target part of a patient, wherein the target part corresponds to the initial burn image set one by one, and the initial burn image set includes a plurality of initial burn images of the same target part at different shooting angles; judging the degree of burns of the patient according to all the initial burn image sets, and obtaining key parts according to the burn judgment result, wherein the burn judgment result includes mild burns, moderate burns and severe burns, and the key parts are the target parts corresponding to the burn judgment result of moderate burns or severe burns; periodically obtaining stage wound images of each of the key parts of the patient; evaluating the recovery of the wound according to all the stage wound images and the initial burn images to obtain an evaluation result, wherein the evaluation result includes suspected infection or scar hyperplasia; when the evaluation result is suspected infection or scar hyperplasia, triggering an adverse reaction prediction mechanism, and triggering an alarm according to the prediction result.

[0006] By adopting the above technical solution, the present application mainly realizes dynamic monitoring of burn wounds through four aspects: acquisition of burn images of various parts, determination of burn degree, monitoring and evaluation of wound recovery, and prediction and early warning of adverse reactions. The present application first obtains the initial burn image set of each target part of the patient, and determines the severity of burns of each target part according to the initial burn images at different angles to obtain the key parts, which can not only accurately determine the degree of burns of the patient and provide a reliable basis for subsequent treatment, but also enable medical resources to be more targeted in key areas to improve treatment efficiency. Then, the present application periodically obtains stage wound images of key parts, and evaluates the wound recovery in combination with the initial burn images, so as to timely grasp the changing trend of the patient's wound and provide a reference for adjusting the treatment plan. In addition, when the wound recovery is found to be different (the recovery speed is significantly slow, and it is increasingly similar to the adverse reaction image, etc.) (that is, when the evaluation result is suspected infection or scar hyperplasia), the adverse reaction prediction mechanism is triggered, which helps to remind the nursing staff to take intervention measures or adjust the treatment plan as soon as the wound is about to have an adverse reaction, thereby reducing the risk of complications for patients.

[0007] Compared with the current static evaluation and diagnosis and treatment methods, this application can more comprehensively track and monitor the dynamic changes of different patients' burn wounds in the subsequent recovery process, and promptly detect problems such as wound infection and poor healing. At the same time, this application obtains an initial burn image set from different angles, conducts a comprehensive analysis to determine the extent of the burn, avoids the differences in manual subjective judgment and the dependence on the experience of nursing staff, and makes the determination of the extent of the burn more accurate; periodically obtains stage wound images of key areas and conducts evaluations, which can continuously and dynamically grasp the recovery of the wound, unlike manual observation, which may be intermittent and unstable. When suspected infection or scar hyperplasia occurs, the adverse reaction prediction mechanism is triggered and an alarm is sounded, which provides a guarantee for timely intervention, while manual observation may miss the best time for intervention due to negligence or lack of experience;

[0008] In summary, this application effectively overcomes the problems of inaccurate monitoring and high professional requirements caused by inconsistent manual observation standards, and improves the scientificity and effectiveness of dynamic monitoring during the recovery process of burn wounds.

[0009] In a specific implementation manner, determining the burn extent of the patient based on all the initial burn image sets includes:

[0010] Performing image preprocessing on each of the initial burn images;

[0011] Dividing all the preprocessed initial burn images into a plurality of detection areas, and extracting a first feature matrix of all the detection areas corresponding to each of the initial burn images;

[0012] Obtaining initial wound surface information in each of the initial burn images by using contour recognition technology;

[0013] A burn determination result is obtained according to the initial wound surface information.

[0014] By adopting the above technical solution, the initial burn image is preprocessed, which can effectively remove interference factors such as noise in the image, make the image clearer and more accurate, and provide a better basis for subsequent image analysis and processing; by dividing the image into several detection areas and extracting the first feature matrix, the burn situation can be analyzed from multiple local areas, the characteristics of the burn can be captured more carefully, and the accuracy of the burn degree determination can be improved; by using contour recognition technology to obtain the initial wound surface information, key information such as the range and shape of the burn can be determined more accurately, further providing a strong basis for the determination of the burn degree.

[0015] In a specific implementation scheme, the initial wound surface information includes the shape of the wound surface, a number of the detection areas covered by the wound surface, and all corresponding first feature matrices; and obtaining the burn determination result according to the initial wound surface information includes:

[0016] Taking all the detection areas covered by the wound surface in the initial wound surface information as burn areas, and taking the first feature matrix corresponding to all the burn areas as an initial feature matrix;

[0017] Obtaining color feature values ​​according to the initial feature matrix and a plurality of preset burn feature matrices, wherein the burn feature matrix is ​​a plurality of feature matrices corresponding to different burn determination results respectively;

[0018] Obtaining a depth feature value according to the shape of the wound surface in the initial wound surface information;

[0019] Calculate the area characteristic value according to the detection area covered by the wound in the initial wound information;

[0020] Obtaining a burn determination result according to the color feature value, the area feature value, and the depth feature value;

[0021] The step of obtaining color feature values ​​according to the initial feature matrix and a plurality of preset burn feature matrices includes:

[0022] Comparing the initial characteristic matrix with a plurality of preset burn characteristic matrices respectively for phase difference, and obtaining a plurality of first phase differences corresponding to each of the initial characteristic matrices, wherein the first phase difference is used to reflect the difference between the two characteristic matrices;

[0023] A color feature value is obtained according to all the first phase differences corresponding to each of the initial feature matrices, and the color feature value corresponds to each of the initial feature matrices one by one.

[0024] By adopting the above technical scheme, by taking the detection area covered by the wound as the burn area and extracting the corresponding feature matrix, and combining the preset feature matrix corresponding to different burn judgment results to obtain the color feature value, it is possible to more accurately reflect the corresponding relationship between the color feature of the burn and the different burn judgment results; the depth feature value is obtained according to the shape of the wound, and the three-dimensional features of the burn wound are considered, so that the judgment result is more in line with the actual situation; the area feature value is obtained by calculating the detection area covered by the wound, which intuitively reflects the range of the burn and provides an important basis for the judgment of the degree of burn; by comparing the initial feature matrix with several preset burn feature matrices, the difference between the two feature matrices can be accurately quantified, thereby reflecting the closeness of the color of the burn wound to the color corresponding to the different burn judgment results; for each initial feature matrix, the corresponding first difference is calculated, so that the evaluation of the color feature is more detailed and comprehensive, reducing the uncertainty caused by subjective judgment, thereby improving the accuracy of the color feature value. The present application comprehensively considers the three key feature values ​​of color, area and depth to determine the burn judgment result, evaluates the burn situation from multiple dimensions, and improves the accuracy and comprehensiveness of the judgment.

[0025] In a specific implementation scheme, the calculation formula of the first phase difference is as follows:

[0026]

[0027]

[0028]

[0029]

[0030] in, Indicates that the initial burn image is in the detection area The initial feature matrix of represents a burn feature matrix corresponding to a certain burn determination result; express and The Euclidean distance between Represents the feature matrix Middle Line Elements of a column; Represents the feature matrix Middle Line Elements of a column; represents the first phase difference;

[0031] The calculation formula of the color feature value is as follows:

[0032]

[0033]

[0034] in, Represents color feature value; Indicates that the burn determination result is The reference value corresponding to the burn area of ​​the burn; It means that the initial feature matrix and the burn determination result are The first phase difference between the burn feature matrices corresponding to the burns; express The minimum value in .

[0035] In a specific implementation manner, obtaining a burn determination result according to the color characteristic value, the area characteristic value, and the depth characteristic value comprises:

[0036] Obtaining a burn assessment value according to the color feature value, the area feature value, and the depth feature value, wherein the burn assessment value is used to reflect the severity of the burn at the target part of the patient;

[0037] When the burn assessment value is less than a preset first assessment threshold, the burn determination result is a mild burn;

[0038] When the burn assessment value is greater than or equal to the first assessment threshold and less than a preset second assessment threshold, the burn determination result is moderate burn;

[0039] When the burn assessment value is greater than or equal to the second assessment threshold, the burn determination result is severe burn; and the first assessment threshold is less than the second assessment threshold.

[0040] In a specific implementation scheme, the evaluation results include normal, suspected infection or scar hyperplasia; the wound recovery is evaluated based on all the wound images at the stages and the initial burn image, and the evaluation results include:

[0041] Dividing all the wound surface images at the said stage into a plurality of detection areas, and extracting a second feature matrix of all the detection areas corresponding to each wound surface image at the said stage;

[0042] Obtaining stage wound surface information according to each stage wound surface image, taking all the detection areas covered by the wound surface in the stage wound surface information as burn areas, and taking the second feature matrix corresponding to all the burn areas as the stage feature matrix;

[0043] Calculating the average value of all the stage feature matrices corresponding to the wound surface image of each stage, and using the average value as the stage average feature matrix;

[0044] Calculating an average value of all the initial feature matrices corresponding to each of the initial burn images, and using the average value as an initial average feature matrix;

[0045] Calculating in chronological order a first similarity between the initial average feature matrix and a preset standard feature matrix, and a second similarity between the average feature matrix of each stage and the standard feature matrix, wherein the standard feature matrix is ​​used to reflect the gray value distribution of the initial healing image of the burn wound surface;

[0046] When the second similarity is greater than the first similarity, and all the second similarities increase in sequence, the evaluation result is that everything is normal;

[0047] Otherwise, the evaluation result is suspected infection or scarring.

[0048] By adopting the above technical scheme, the stage wound image is divided into several detection areas and the second feature matrix is ​​extracted. Similar to the processing method of the initial burn image, the wound changes can be analyzed locally and in detail to improve the accuracy of the assessment; the average value of the stage feature matrix is ​​calculated to obtain the stage average feature matrix, and the average value of the initial feature matrix is ​​calculated to obtain the initial average feature matrix. Through averaging processing, the overall wound characteristics can be more stably reflected; the similarity between the initial average feature matrix and the standard feature matrix is ​​calculated, and a standard reflecting the gray value distribution of the initial healing image of the burn wound is introduced, so that the assessment has an objective reference; the similarity is calculated and compared in chronological order, which can dynamically monitor the changing trend in the wound recovery process, promptly discover abnormal conditions in the wound recovery process, and provide early warning for taking corresponding treatment measures.

[0049] In a specific implementation scheme, the prediction result includes the possibility of adverse reaction development and everything is normal; when the evaluation result is suspected infection or scar hyperplasia, the adverse reaction prediction mechanism is triggered, and the alarm is triggered according to the prediction result, including:

[0050] Acquire a number of historical adverse reaction diagnosis images, and obtain an adverse reaction reference matrix corresponding to each type based on all the adverse reaction diagnosis images of the same type;

[0051] Calculate the third similarity between the average feature matrix of each stage and each adverse reaction reference matrix in chronological order;

[0052] If the third similarity increases, the prediction result is that there is a possibility of adverse reaction development, and an alarm is directly triggered;

[0053] If the third similarity decreases, the prediction result is normal and no alarm is triggered;

[0054] Otherwise, obtaining the patient's physical sign changes, including changes in body temperature and heart rate and respiration;

[0055] When the changes in the vital signs meet the alarm conditions, an alarm is triggered, and the alarm conditions refer to that the body temperature rises and is about to exceed the normal range or the heart rate and breathing gradually accelerate.

[0056] By adopting the above technical solution, the adverse reaction reference matrix is ​​obtained by acquiring historical adverse reaction diagnosis images, which provides a reliable reference basis for judging the current patient's wound condition; the third similarity between the stage average feature matrix and the adverse reaction reference matrix is ​​calculated in chronological order, which can dynamically monitor the trend of the wound to the adverse reaction. If the third similarity increases, it is predicted that there is a possibility of adverse reaction development and an alarm is directly triggered, so as to promptly remind medical staff to take measures; if the third similarity decreases, it is predicted that everything is normal, and no alarm is triggered at this time, thus avoiding unnecessary interference and improving the accuracy of the alarm; when the similarity situation is unclear, the judgment is made in combination with the changes in the patient's physical signs, which further improves the accuracy of the prediction, and triggers the alarm when the physical sign changes meet the alarm conditions, ensuring that the changes in the patient's condition can be discovered in time.

[0057] In a second aspect, the present application provides a smart terminal, which adopts the following technical solution:

[0058] A smart terminal, characterized in that it includes a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement a method for dynamic wound monitoring applied to burn treatment as described in the first aspect.

[0059] In a third aspect, the present application provides a wound surface dynamic monitoring system for burn treatment, which adopts the following technical solution:

[0060] A wound surface dynamic monitoring system for burn treatment, comprising:

[0061] An image acquisition device, used to periodically acquire a number of initial burn images and staged wound surface images of each target part of the patient at different angles;

[0062] The intelligent terminal as described in the second aspect is used to obtain the initial burn image and the stage wound surface image in the image acquisition device, and dynamically monitor the recovery of the patient's wound surface based on the initial burn image and the stage wound surface image.

[0063] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution:

[0064] A computer-readable storage medium, wherein at least one instruction, at least one program, code set or instruction set is stored in the readable storage medium, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement a method for dynamic monitoring of wound surface applied to burn treatment as described in the first aspect.

[0065] In summary, the present application includes at least one of the following beneficial technical effects:

[0066] 1. Compared with the current static evaluation and diagnosis and treatment methods when patients seek medical treatment and the manual observation of wound healing during treatment, this application can more comprehensively track and monitor the dynamic changes of burn wounds of different patients during the subsequent recovery process, and timely discover problems such as infection and poor healing of the wound;

[0067] 2. This application obtains the initial burn image set from different angles, and makes a comprehensive judgment on the burn degree by combining the color feature value, depth feature value and area feature value, avoiding the difference in manual subjective judgment and the dependence on the experience of nursing staff, making the burn degree judgment more accurate.

[0068] 3. Periodically obtain stage wound images of key areas and conduct evaluations, which can continuously and dynamically grasp the wound recovery status and improve the intermittent and unstable problems existing in manual observation; when suspected infection or scar hyperplasia occurs, the adverse reaction prediction mechanism is triggered and an alarm is sounded, which provides a guarantee for timely intervention and avoids the situation where manual observation misses the best time for intervention due to negligence or lack of experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a structural schematic diagram of a wound surface dynamic monitoring system applied to burn treatment according to an embodiment of the present application.

[0070] Figure 2 It is a flow chart of a method for dynamic monitoring of wound surface applied to burn treatment according to another embodiment of the present application.

[0071] Explanation of reference numerals: 100, image acquisition device; 200, intelligent terminal; 210, memory; 220, processor. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solutions and advantages of the present application clearer, the implementation mode of the present application will be further described in detail in conjunction with the accompanying drawings; the following is a further detailed description of an embodiment of a method, system and device for dynamic monitoring of wounds applied to burn treatment in the present application in conjunction with the accompanying drawings of the specification.

[0073] An embodiment of the present application discloses a wound surface dynamic monitoring system for burn treatment.

[0074] Reference Figure 1A dynamic monitoring system for wounds used in burn treatment includes an image acquisition device 100 and an intelligent terminal 200 that is communicatively connected to the image acquisition device 100; the image acquisition device 100 is used to periodically acquire a number of initial burn images and stage wound images of each target part of the patient at different angles; the image acquisition device 100 of this embodiment uses a high-resolution camera to clearly capture the appearance of the patient's burn wound, including details such as the shape, color, and texture of the wound, which helps to improve the accuracy of subsequent intelligent devices in analyzing and processing the acquired images. The intelligent terminal 200 includes a memory 210 and a processor 220, which is used to acquire the initial burn image and stage wound image in the image acquisition device 100 for processing by the processor 220, and dynamically monitor the recovery of the patient's wound according to the initial burn image and the stage wound image.

[0075] The memory 210 is used to store at least one instruction, at least one program, code set or instruction set; the processor 220 executes the following steps of the wound dynamic monitoring method applied to burn treatment when running at least one instruction, at least one program, code set or instruction set.

[0076] The implementation of the method is described in detail below in conjunction with the above system:

[0077] Another embodiment of the present application discloses a method for dynamic monitoring of wound surface applied to burn treatment.

[0078] Reference Figure 2 A method for dynamic monitoring of wound surface applied to burn treatment includes:

[0079] S100, acquiring an initial burn image set of each target part of the patient;

[0080] The target parts mainly include the neck, front torso, back torso, upper arms, forearms, hands, buttocks, thighs, calves and feet. The initial burn image set is the initial burn images taken at several different angles of the patient with a vertical body under natural light conditions.

[0081] It should be noted that in this embodiment, since the initial burn image needs to include the corresponding entire target part and the burn wound surface, the number of initial burn images corresponding to each target part is mainly determined according to the complexity of the target part. For example, for parts with relatively regular shapes and small burn areas such as arms (upper arms, forearms), legs (thighs, calves), etc., 3-5 pictures are taken from the front and side angles as an example; for parts with complex structures such as hands and burns that may involve multiple planes, 5-10 or even more pictures are taken from different angles as an example (for example: hand burns require taking pictures of the palm, back of the hand, fingers at various angles, as well as the state of joint movement, to facilitate subsequent burn assessment of hand function). Among them, the front photo is mainly used to clearly show the contrast between the burn area and the normal area to distinguish the burn area; the side photo is mainly used to assist in detecting the three-dimensional situation and depth changes of the burn wound surface.

[0082] S200, determining the burn degree of the patient according to all the initial burn image sets, and obtaining the key parts according to the burn determination results;

[0083] Among them, the burn assessment results include mild burns, moderate burns and severe burns; key areas refer to one or more target areas corresponding to the burn assessment results of moderate burns or severe burns.

[0084] Specifically, the steps of determining the burn degree of the patient according to all the initial burn image sets in S200 are as follows:

[0085] S210, performing image preprocessing on each initial burn image;

[0086] Among them, the image preprocessing operation of the present application includes denoising the image (this embodiment takes bilateral filtering technology as an example, and combines the spatial proximity and pixel value similarity of the image for filtering, which can better retain the complex texture and edge information in the initial burn image while removing noise), image enhancement (by adjusting the grayscale histogram of the image, linear stretching, histogram equalization and other methods, the contrast of the image is increased to make the burn area clearer and more visible), image normalization (size normalization, by adjusting the initial burn images of different resolutions corresponding to the same target part to a uniform size through bilinear interpolation and other methods, so as to facilitate subsequent analysis and processing; grayscale normalization, by adjusting the grayscale value range of the image to a uniform range through linear transformation and other methods to eliminate the brightness difference between different images), image cropping (removing irrelevant areas in the image and retaining only the area containing the target part. The retained area in this embodiment takes the minimum adjacency matrix of each target part as an example), etc. The above preprocessing methods are all existing technologies and will not be repeated here.

[0087] S220, dividing all the initial burn images into a number of detection areas, and extracting a first feature matrix of all the detection areas corresponding to each initial burn image;

[0088] Among them, this embodiment adopts ( as well as The value is determined according to the size of the initial burn image corresponding to each target part after preprocessing) grid, each initial burn image is divided into several detection areas of the same area, starting with the first detection area in the upper left corner of each initial burn image, and each divided detection area is named as the position of the detection area ,Right now Indicates that the detection area is located in the corresponding image Line List, From left to right, in increasing order, Increasing from top to bottom, and .

[0089] It should be noted that all feature matrices in this embodiment are two-dimensional arrays. These feature matrices are used to represent the digital image data of an image, wherein the rows of the feature matrix correspond to the height of the image, the columns correspond to the width of the image, the elements in the feature matrix correspond to the pixels of the image, and the values ​​of the elements are the grayscale values ​​of the pixels. Since the grayscale value in a grayscale image ranges from 0 to 255, the closer to white, the larger the value, and the closer to black, the smaller the value; the first feature matrix in this embodiment is For example.

[0090] S230, obtaining initial wound surface information in each initial burn image by using a contour recognition technology;

[0091] Among them, the contour recognition technology of this embodiment can be an image segmentation technology, including but not limited to a threshold-based segmentation method (in the initial burn image, if the color difference between the wound surface and the normal skin is obvious, the wound surface can be extracted from the corresponding target part through threshold segmentation), a region growing method, a watershed algorithm, etc., or a gradient / morphology-based edge detection method, or a deep learning technology (using a deep neural network, etc. to perform pixel-level classification of the initial burn image, etc.), all of which are existing technologies and will not be described in detail here; the initial wound surface information includes the shape of the wound surface, the position of the wound surface in the corresponding target part, a number of detection areas covered by the wound surface, and a first feature matrix corresponding to each detection area.

[0092] S240, obtaining a burn assessment result according to the initial wound surface information;

[0093] Specifically, S240 includes:

[0094] S241, taking all detection areas covered by the wound surface in the initial wound surface information as burn areas, and taking the first feature matrix corresponding to all the burn areas as the initial feature matrix;

[0095] Among them, the initial feature matrix is ​​mainly used to reflect the initial gray value distribution characteristics of the burn area.

[0096] S242, obtaining color feature values ​​according to the initial feature matrix and a plurality of preset burn feature matrices;

[0097] The burn feature matrix is ​​a plurality of feature matrices corresponding to different burn determination results. This embodiment takes three as examples: light pink corresponding to mild burns, purple corresponding to moderate burns, and black corresponding to severe burns, which are the corresponding feature matrices in the grayscale histogram.

[0098] Specifically, S242 includes:

[0099] S2421, comparing the initial characteristic matrix with all the burn characteristic matrices respectively to obtain a plurality of first phase differences corresponding to each initial characteristic matrix;

[0100] The first phase difference is used to reflect the difference between the two feature matrices. The smaller the first phase difference is, the more similar the two feature matrices (the initial feature matrix and the burn feature matrix) are, that is, the gray value distribution of the burn area is similar to the gray value distribution of the burn area corresponding to a certain burn determination result. The first phase difference of this embodiment is mainly obtained by the Euclidean distance calculation formula, which is:

[0101]

[0102]

[0103]

[0104]

[0105] in, Indicates that the initial burn image is in the detection area The initial feature matrix of represents the burn feature matrix corresponding to a certain burn determination result; express and The Euclidean distance between Represents the feature matrix Middle Line Elements of a column; Represents the feature matrix Middle Line Elements of a column; Represents the first phase difference, and its value range is [0,1].

[0106] S2422, obtaining color eigenvalues ​​according to all first phase differences corresponding to each initial eigenmatrix;

[0107] Among them, the color eigenvalue corresponds to each initial feature matrix one by one, and its calculation formula is as follows:

[0108]

[0109]

[0110] in, Represents color feature value; Indicates the burn assessment result is The reference value corresponding to the burn area of ​​the burn; The initial feature matrix and the burn determination result are represented as The first phase difference between the burn feature matrices corresponding to the burns; express The minimum value in .

[0111] S243, obtaining a depth feature value according to the shape of the wound surface in the initial wound surface information;

[0112] Among them, this embodiment mainly obtains the depth of the wound surface in the target part (i.e., the degree of depression) based on the initial burn image with a side shooting angle in S100; specifically, taking a horizontally placed forearm side image as an example, under normal circumstances, in the image of a healthy person, the feature matrix of each detection area is similar (the color distribution is uniform), and there is a smooth outline (horizontal line) of an arm above and below the image, but in the image of a burn patient, the feature matrix of the burn area will be greatly different from the feature matrix of other detection areas (the gray value distribution is different, that is, the first phase difference is large), and the upper and lower horizontal lines have obvious twists and turns (usually mainly concave). At this time, the specific depth of the wound surface can be obtained according to the vertical difference between the horizontal lines of the skin edges of other detection areas and the horizontal lines of the skin edges of the burn area in the horizontal direction (which can be calculated based on the side lengths of several detection areas that are different). The depth feature value of this embodiment takes the depth value with the largest value in the initial burn image as an example.

[0113] S244, calculating an area characteristic value according to the detection area covered by the wound in the initial wound information;

[0114] Among them, the area characteristic value refers to the sum of the areas of the wound surface in the corresponding target part; specifically, in S244, the area values ​​of the wound surface in the initial burn images of all angles can be calculated according to the number of detection areas (i.e., burn areas) covered by the wound surface and the side length of each detection area, and then the burn areas at different angles are weighted averaged to obtain the sum of the areas (in this embodiment, weights are allocated according to the visibility of the burn area in each initial burn image, that is, a higher weight ratio is assigned to the largest wound surface area value, and the weights of the wound surface area values ​​in the initial burn images of other angles decrease in sequence according to their area sizes).

[0115] S245, obtaining a burn determination result according to the color characteristic value, the area characteristic value, and the depth characteristic value;

[0116] The burn assessment results include mild burns, moderate burns and severe burns. Specifically, S245 includes:

[0117] S2451, obtaining a burn assessment value according to the color characteristic value, the area characteristic value and the depth characteristic value;

[0118] The burn assessment value is used to reflect the severity of the burns in the target part of the patient. When the color feature value / area feature value / depth feature value is larger, the burn assessment value is higher, that is, the burn is more severe. Specifically, the burn assessment value of this embodiment is calculated by a weighted method, and the weight of each influencing factor (color feature value, area feature value and depth feature value) is set according to its relative importance in a certain target part. For example, when the target part is the neck, considering that the neck is usually exposed to the air, the burn may have a greater impact on the patient's appearance, and the weight of the color feature value is relatively high; when the target part is the hand, due to the importance of the hand function, the weight of the depth feature value may be relatively high; when the target part is the trunk, the weight of the area feature value may be relatively prominent. The specific weight values ​​can be determined by expert evaluation, data analysis and other methods to ensure that the burn assessment value can accurately reflect the severity of the burns in different target parts.

[0119] S2452, when the burn assessment value is less than a preset first assessment threshold, the burn determination result is a mild burn;

[0120] When the burn assessment value is greater than or equal to the first assessment threshold and less than the preset second assessment threshold, the burn determination result is moderate burn;

[0121] When the burn assessment value is greater than or equal to the second assessment threshold, the burn determination result is severe burn;

[0122] Among them, the first evaluation threshold is smaller than the second evaluation threshold. In this embodiment, the values ​​of the first evaluation threshold and the second evaluation threshold need to be adjusted according to the total area of ​​the target part where burns actually occur in different patients, and are dynamic values.

[0123] S300, periodically obtain the wound surface images of each key part of the patient;

[0124] The stage wound surface images of this embodiment are photographed at an angle with the highest visibility of the wound surface (largest wound surface display area).

[0125] It should be noted that the cycle in S300 will vary depending on the severity of the burn, the overall condition of the patient, and the stage of treatment. Generally speaking, for patients with moderate or above burns, at least 1-2 wound monitoring times a day may be required in the early acute phase after burns (usually within 1 week after injury) (generally, the nursing staff can take stage wound images of each key part of the patient after completing debridement and other care for the patient). As the wound begins to heal, the risk of infection gradually decreases, and the monitoring frequency can be adjusted to once every 2-3 days. If an early warning is triggered in the subsequent S500 (suspected infection, scar hyperplasia, etc. occur during the healing process of the patient's wound), the monitoring frequency will be appropriately increased within a certain period of time until the patient's wound recovery returns to normal; if the patient's condition is stable and the wound heals well, images can be collected and analyzed at a monitoring frequency of 1-2 times a week in the later stage until the wound is completely healed. The specific monitoring frequency should be adjusted individually by the burn specialist according to the actual situation of the patient.

[0126] S400, evaluating the recovery of the wound surface according to the wound surface images at all stages and the initial burn image, and obtaining an evaluation result;

[0127] Among them, the evaluation results include everything normal, suspected infection or scar hyperplasia.

[0128] Specifically, S500 includes:

[0129] S410, dividing the wound surface images of all stages into a number of detection areas, and extracting the second feature matrix of all the detection areas corresponding to the wound surface images of each stage;

[0130] The method of dividing the grid is the same as step S220, which will not be repeated here.

[0131] S420, obtaining stage wound surface information according to the wound surface image of each stage, and taking all detection areas covered by the wound surface in the stage wound surface information as burn areas, and taking the second feature matrix corresponding to all burn areas as the stage feature matrix;

[0132] Among them, the method principle of S420 is the same as S241, which will not be repeated here; the stage wound information is similar to the initial wound information, including the shape of the wound, the position of the wound in the corresponding key area, several detection areas covered by the wound and all corresponding second feature matrices.

[0133] S430, calculating the average value of all stage feature matrices corresponding to the wound surface image of each stage, and using the average value as the stage average feature matrix;

[0134] Calculate the average value of all initial feature matrices corresponding to each initial burn image, and use the average value as the initial average feature matrix;

[0135] Among them, the stage average feature matrix represents the average value of all stage feature matrices of a wound image at a certain stage, and the initial average feature matrix represents the average value of all initial feature matrices of a certain initial burn image. Both of the above are used to reflect the grayscale value distribution of the wound surface (in this embodiment, the grayscale value is mainly used to reflect the grayscale depth of the wound surface).

[0136] S440, calculating the first similarity between the initial average feature matrix and the preset standard feature matrix, and the second similarity between the average feature matrix of each stage and the standard feature matrix in chronological order;

[0137] Among them, the standard feature matrix is ​​used to reflect the gray value distribution of the initial healing image of the burn wound. In this embodiment, the image of new light skin is taken as an example. If the burn condition of a key part is very serious, the corresponding standard feature matrix is ​​the dark skin in a scab state. Specifically, in this embodiment, the first similarity / second similarity between the two feature matrices is calculated by the cosine similarity formula (the value range is 0-1, and the larger the value, the more similar the two images are). This is a prior art and will not be repeated here.

[0138] S450, when the second similarity is greater than the first similarity, and all the second similarities increase in sequence, the evaluation result is that everything is normal;

[0139] Otherwise, the evaluation result is suspected infection or scarring.

[0140] S500, when the evaluation result is suspected infection or scar hyperplasia, the adverse reaction prediction mechanism is triggered, and an alarm is triggered according to the prediction result;

[0141] Among them, the prediction results include the possibility of adverse reactions and everything is normal; specifically, S500 includes:

[0142] S510, obtaining a number of historical adverse reaction diagnosis images, and obtaining an adverse reaction reference matrix corresponding to each type based on all the adverse reaction diagnosis images of the same type;

[0143] Among them, the adverse reaction diagnosis images of this embodiment take the wound infection diagnosis images and the scar hyperplasia diagnosis images as examples; the adverse reaction reference matrix is ​​used to reflect the grayscale distribution characteristics of a certain type of adverse reaction in the grayscale histogram, and its acquisition principle is the same as the stage average feature matrix in S430, which will not be repeated here.

[0144] S520, calculating the third similarity between the average feature matrix of each stage and each adverse reaction reference matrix in chronological order;

[0145] Among them, the calculation formula of the third similarity is the same as the first similarity and the second similarity (cosine similarity formula), which is used to reflect the similarity between the stage wound image and the adverse reaction diagnosis image.

[0146] S530, if the third similarity increases, the prediction result is that there is a possibility of adverse reaction development, and an alarm is directly triggered;

[0147] If the third similarity decreases, the prediction result is normal and no alarm is triggered;

[0148] Otherwise, obtain the patient's physical sign changes, and trigger an alarm when the physical sign changes meet the alarm conditions;

[0149] Among them, changes in physical signs mainly include changes in body temperature and heart rate and breathing that do not require additional analysis and testing; meeting the alarm conditions means that the body temperature rises and is about to exceed the normal range / the heart rate and breathing gradually accelerate.

[0150] Based on the same inventive concept mentioned above, an embodiment of the present application also discloses an intelligent terminal, which includes a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement a dynamic wound monitoring method for burn treatment as provided in the above method embodiment.

[0151] Based on the same inventive concept mentioned above, an embodiment of the present application also discloses a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set. The at least one instruction, at least one program, code set or instruction set can be loaded and executed by a processor to implement the above method embodiment, which provides a method for dynamic monitoring of wounds applied to burn treatment.

[0152] It should be understood that the "plurality" mentioned in this article refers to two or more. "And / or" describes the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0153] Those skilled in the art will appreciate that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium. The storage medium mentioned above may include, for example, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0154] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for dynamic monitoring of wound surface applied to burn treatment, characterized in that: include: Acquire an initial burn image set of each target part of the patient, wherein the target part corresponds to the initial burn image set one by one, and the initial burn image set includes a plurality of initial burn images of the same target part taken at different angles; Determine the degree of burns of the patient according to all the initial burn image sets, and obtain key parts according to the burn determination results, wherein the burn determination results include mild burns, moderate burns and severe burns, and the key parts are the target parts corresponding to the burn determination results of moderate burns or severe burns; Periodically obtain stage wound surface images of each of the key parts of the patient; Evaluating the recovery of the wound surface according to the wound surface images at all the stages and the initial burn image to obtain an evaluation result, wherein the evaluation result includes suspected infection or scar hyperplasia; When the evaluation result is suspected infection or scar hyperplasia, the adverse reaction prediction mechanism is triggered, and an alarm is triggered according to the prediction result; Determining the degree of burns of the patient according to all the initial burn image sets comprises: Performing image preprocessing on each of the initial burn images; Dividing all the preprocessed initial burn images into a plurality of detection areas, and extracting a first feature matrix of all the detection areas corresponding to each of the initial burn images; Obtaining initial wound surface information in each of the initial burn images by using a contour recognition technology, wherein the initial wound surface information includes the shape of the wound surface, a number of the detection areas covered by the wound surface, and all the corresponding first feature matrices; Obtaining a burn assessment result according to the initial wound surface information; Obtaining a burn determination result according to the initial wound surface information comprises: Taking all the detection areas covered by the wound surface in the initial wound surface information as burn areas, and taking the first feature matrix corresponding to all the burn areas as an initial feature matrix; Obtaining color feature values ​​according to the initial feature matrix and a plurality of preset burn feature matrices, wherein the burn feature matrix is ​​a plurality of feature matrices corresponding to different burn determination results respectively; Obtaining a depth feature value according to the shape of the wound surface in the initial wound surface information; Calculate the area characteristic value according to the detection area covered by the wound in the initial wound information; Obtaining a burn determination result according to the color feature value, the area feature value, and the depth feature value; The step of obtaining color feature values ​​according to the initial feature matrix and a plurality of preset burn feature matrices includes: Comparing the initial characteristic matrix with a plurality of preset burn characteristic matrices respectively for phase difference, and obtaining a plurality of first phase differences corresponding to each of the initial characteristic matrices, wherein the first phase difference is used to reflect the difference between the two characteristic matrices; A color feature value is obtained according to all the first phase differences corresponding to each of the initial feature matrices, and the color feature value corresponds to each of the initial feature matrices one by one.

2. The method for dynamic wound monitoring according to claim 1, characterized in that: The calculation formula of the first phase difference is as follows: in, Indicates that the initial burn image is in the detection area The initial feature matrix of represents a burn feature matrix corresponding to a certain burn determination result; express and The Euclidean distance between Represents the feature matrix Middle Line Elements of a column; Represents the feature matrix Middle Line Elements of a column; represents the first phase difference; The calculation formula of the color feature value is as follows: in, Represents color feature value; Indicates that the burn determination result is The reference value corresponding to the burn area of ​​the burn; It means that the initial feature matrix and the burn determination result are The first phase difference between the burn feature matrices corresponding to the burns; express The minimum value in .

3. The method for dynamic wound monitoring according to claim 2, characterized in that: The burn determination result obtained according to the color feature value, the area feature value and the depth feature value comprises: Obtaining a burn assessment value according to the color feature value, the area feature value, and the depth feature value, wherein the burn assessment value is used to reflect the severity of the burn at the target part of the patient; When the burn assessment value is less than a preset first assessment threshold, the burn determination result is a mild burn; When the burn assessment value is greater than or equal to the first assessment threshold and less than a preset second assessment threshold, the burn determination result is moderate burn; When the burn assessment value is greater than or equal to the second assessment threshold, the burn determination result is severe burn; and the first assessment threshold is less than the second assessment threshold.

4. The method for dynamic wound monitoring according to claim 1, characterized in that: The evaluation results include normal, suspected infection or scar hyperplasia; the recovery of the wound surface is evaluated based on all the wound surface images at the said stages and the initial burn image, and the evaluation results obtained include: Dividing all the wound surface images at the said stage into a plurality of detection areas, and extracting the second feature matrix of all the detection areas corresponding to each wound surface image at the said stage; Obtaining stage wound surface information according to each stage wound surface image, taking all the detection areas covered by the wound surface in the stage wound surface information as burn areas, and taking the second feature matrix corresponding to all the burn areas as the stage feature matrix; Calculating the average value of all the stage feature matrices corresponding to the wound surface image of each stage, and using the average value as the stage average feature matrix; Calculating an average value of all the initial feature matrices corresponding to each of the initial burn images, and using the average value as an initial average feature matrix; Calculating in chronological order a first similarity between the initial average feature matrix and a preset standard feature matrix, and a second similarity between the average feature matrix of each stage and the standard feature matrix, wherein the standard feature matrix is ​​used to reflect the gray value distribution of the initial healing image of the burn wound surface; When the second similarity is greater than the first similarity, and all the second similarities increase in sequence, the evaluation result is that everything is normal; Otherwise, the evaluation result is suspected infection or scarring.

5. The method for dynamic wound monitoring according to claim 4, characterized in that: The prediction results include the possibility of adverse reactions and whether everything is normal; when the evaluation result is suspected infection or scar hyperplasia, the adverse reaction prediction mechanism is triggered, and the alarm is triggered according to the prediction result, including: Acquire a number of historical adverse reaction diagnosis images, and obtain an adverse reaction reference matrix corresponding to each type based on all the adverse reaction diagnosis images of the same type; Calculate the third similarity between the average feature matrix of each stage and each adverse reaction reference matrix in chronological order; If the third similarity increases, the prediction result is that there is a possibility of adverse reaction development, and an alarm is directly triggered; If the third similarity decreases, the prediction result is normal and no alarm is triggered; Otherwise, obtaining the patient's physical sign changes, including changes in body temperature and heart rate and respiration; When the changes in the vital signs meet the alarm conditions, an alarm is triggered, wherein the alarm conditions refer to that the body temperature rises and is about to exceed the normal range or the heart rate and breathing gradually accelerate.

6. An intelligent terminal, characterized in that: It includes a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the dynamic wound monitoring method for burn treatment as described in any one of claims 1 to 5.

7. A dynamic wound monitoring system for burn treatment, using an intelligent terminal as claimed in claim 6, the dynamic wound monitoring system comprising: An image acquisition device (100) is used to periodically acquire a number of initial burn images and staged wound surface images of each target part of a patient at different angles; The intelligent terminal (200) is used to obtain the initial burn image and the stage wound surface image in the image acquisition device (100), and dynamically monitor the recovery of the patient's wound surface based on the initial burn image and the stage wound surface image.

8. A computer-readable storage medium, characterized in that: The readable storage medium stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the dynamic wound monitoring method for burn treatment as described in any one of claims 1 to 5.

Images