Method and system for evaluating burn condition based on burn area and burn depth

Abstract

The invention relates to the technical field of skin injury detection, in particular to a burn condition evaluation method and system based on burn area and burn depth, comprising the following steps: acquiring a wound depth image and calculating a height difference sequence, dividing intervals according to the sequence to form a height level, constructing a wound triangular mesh and generating a height mesh data set, according to the method, the depth difference between the wound surface and the normal skin is extracted, the height sequence is formed, the fluctuation state is continuously quantified, the hierarchy is constructed according to the height features, the depth and depth structure is clearer, the depth and depth are more accurate, and the burn area depth level combination quantity is generated. A three-dimensional grid is used for expressing a curved surface form, deformation details are accurately presented, damage distribution is presented by calculating a three-dimensional area and combining depth partitions, area proportions are summarized according to anatomical areas and combining depth grades, and therefore area measurement and calculation are more accurate, depth judgment is more objective, and overall load evaluation is more stable.

Description

Technical Field

[0001] This invention relates to the field of skin injury detection technology, and in particular to a method and system for assessing burn conditions based on burn area and burn depth. Background Technology

[0002] The field of skin injury detection technology belongs to the medical testing technology field of identifying, measuring, and judging the condition of human skin wounds. Its core matters include acquiring, quantifying, and analyzing the morphological characteristics of the wound, the degree of tissue damage, and related physiological parameters, and conducting systematic assessments through wound area measurement methods, tissue depth determination methods, and various imaging detection methods. Among them, the traditional method of assessing burn conditions based on burn area and burn depth refers to the technical topic of measuring the area and determining the depth of burn wounds. The technical matter addressed by this technical topic is the quantitative assessment of burn area and burn depth. Traditional methods generally use the nine-point method or palm area method to estimate the burn area according to the proportion of human body parts, and use visual inspection and palpation to determine the depth based on wound color, blistering, and pain response, supplemented by single-modal imaging to obtain limited surface information to complete the basic assessment of burn conditions.

[0003] Current experience-based burn assessments often rely on subjective judgments and two-dimensional proportional estimations based on visual observation of skin color, blister morphology, exudation, and pain response. This results in a limited depiction of the true spatial morphology of the wound, remaining largely from a planar perspective. When estimating the area of ​​complex body parts based on surface proportions, operators need to rely on memory and experience to delineate boundaries. For wounds with irregular edges, scattered distribution, or spanning multiple anatomical regions, manual drawing and estimation are prone to omissions, duplicates, or ambiguous area classifications, leading to significant fluctuations in area results among different operators. This makes it difficult to support refined treatment planning and comparative studies. Depth determination relies on a comprehensive understanding of surface color and texture. When interference factors such as changes in lighting conditions, differences in skin base color, or wound contamination exist, the boundary between superficial and deep wounds is easily confused. Inconsistencies in the assessment grade of the same site at different times or by different medical personnel frequently occur, failing to reflect the true trend of injury evolution over time. Such procedures lack the ability to continuously describe the undulations and depressions of wound surfaces, and often treat three-dimensional structures such as curved areas of the body surface as approximately planar. This can easily lead to underestimating or overestimating the extent of damage in areas with greater curvature, such as the shoulders, buttocks, and fingers and toes, resulting in biased assessments of the overall systemic injury load. In the stages of patient record-keeping and follow-up, the lack of stable and repeatable three-dimensional quantitative data makes it difficult to establish directly comparable continuous data between initial diagnoses and subsequent follow-ups. This leads to problems such as accumulated bias and limited statistical reliability in efficacy evaluation and multi-center collaborative studies. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for assessing burn conditions based on burn area and burn depth, comprising the following steps: S1: Acquire depth images of burn wounds, read depth data of the wound and surrounding normal skin, calculate the height difference, and form a sequence of wound height difference values ​​according to spatial location; S2: Based on the wound height difference sequence, sort the height differences by value and divide them into intervals according to a preset range. Extract parameters from the intervals and calculate the height change characteristics. Arrange the intervals from low to high according to the characteristic values ​​to form a height hierarchy sequence. S3: Based on the height hierarchy sequence, set three-dimensional nodes for the depth image position and assign height hierarchy, construct a wound triangular mesh based on the spatial connection of nodes, and record the height hierarchy information of all meshes to generate a wound height mesh dataset. S4: Calculate the three-dimensional surface area of ​​the triangular mesh based on the wound height grid dataset and form an area set. Accumulate the sets to obtain the three-dimensional area of ​​the wound. At the same time, read the height level and determine the depth level according to the clinical depth threshold to generate the area depth range. S5: Based on the area and depth range, organize the surface area of ​​the triangular grid according to the anatomical regions of the head, trunk, upper limbs, and lower limbs and calculate the area ratio. Read the area depth level and combine it with the clinical score to form assessment parameters and generate a combination of burn area and depth levels.

[0005] As a further aspect of the present invention, the wound height difference sequence includes a wound pixel location index sequence, a corresponding height difference parameter sequence, and a normal skin reference height difference sequence; the height hierarchy sequence includes a height interval number sequence, an interval height change feature parameter sequence, and an interval sorting hierarchy label; the wound height grid dataset includes a wound geometric node spatial coordinate set, a wound triangular grid cell set, and a triangular grid cell height hierarchy annotation set; the area depth interval includes a superficial second-degree burn area interval, a deep second-degree burn area interval, and a third-degree burn area interval; and the burn area depth level combination includes an anatomical region area proportion parameter set, an anatomical region depth level score set, and whole-body assessment parameter values.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Obtain depth image information of the burn wound area of ​​the burn patient, detect the depth data of the wound location in the depth image and detect the depth data of the corresponding location of the normal skin area at the edge of the wound, perform depth difference calculation based on the spatial correspondence between the wound pixels and the normal skin pixels, organize the calculation results according to the wound spatial index, and generate wound difference sequence values. S102: Based on the wound difference sequence value, call the difference set between indices in the sequence and perform segment detection based on the index interval, perform calculation on the data set within the segment and arrange the calculation results in segment order to generate segment mean sequence value; S103: Based on the mean sequence value of the segment, perform sequence splicing on the segment data according to the wound space order, and perform position correspondence processing on the spliced ​​sequence according to the original coordinate order to generate a wound height difference sequence.

[0007] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Arrange the height difference parameters in the sequence according to the numerical order based on the wound height difference sequence. For the arranged parameters, call the preset height difference range and perform interval division based on the range boundary. Parameters falling into the same range are grouped into the same interval and arranged according to the range order to generate a set of height interval values. S202: Based on the set of height interval values, for each interval in the set, call all height difference parameters within the interval and perform calculations based on the parameter set of each interval to obtain the height change of the corresponding interval and arrange all height changes in interval order to generate an interval change sequence. S203: Based on the interval change sequence, the changes within the sequence are arranged in order of magnitude and hierarchical order is established according to the arranged positions to generate a height hierarchy sequence.

[0008] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the height hierarchy sequence, set corresponding three-dimensional geometric nodes of the wound surface for the depth image position in space. For each geometric node, call the height hierarchy information corresponding to the node position in the height hierarchy sequence and perform association processing between the information and the node coordinates. Arrange all nodes in spatial order to generate a set of node hierarchy values. S302: Based on the set of node hierarchy values, call the spatial coordinates between adjacent geometric nodes in the set and perform connection judgment according to the spatial adjacency relationship of the nodes. Establish corresponding triangular structures for nodes that satisfy the adjacency relationship in a three-point combination manner and record the hierarchy value of the nodes in the triangular structure to generate the triangular structure combination quantity. S303: Based on the triangular structure combination quantity, call the node level value for the triangular structure and perform aggregation processing on the level value according to the triangular structure order. Arrange the aggregated data of all triangular structures according to spatial position to generate a wound height grid dataset.

[0009] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Obtain the wound height grid dataset and call the node coordinate information for each triangular grid region in the dataset to perform area calculation. Arrange the area values ​​of all triangular grid regions into an area sequence in spatial order and perform cumulative processing based on all area values ​​in the sequence to generate the area accumulation. S402: Based on the surface accumulation, for each triangular grid region in the wound height grid dataset, read the corresponding height level information of the region and perform combination processing on the height level information and the region area value, arrange the combination data of all regions in the region order, and generate a region combination quantity sequence; S403: Based on the region combination sequence, call the clinically used burn depth grading threshold for the height level information of each region in the sequence and perform grade determination based on the threshold. Then, perform corresponding processing on the determined depth level and the region area value to generate an area depth interval.

[0010] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Collect the corresponding head, torso, upper limb and lower limb anatomical information for the triangular mesh area according to the area and depth range, call all three-dimensional surface area parameters in the area to perform processing, calculate the proportion of the area parameters in the same anatomical area according to the total area of ​​the area and arrange them in the order of the areas to generate the area proportion. S502: Based on the region proportion, for each anatomical region, read the depth level in the area depth interval and call the clinically commonly used depth level score to perform score matching according to the depth level. Combine the matched score value with the corresponding region proportion and organize it according to the region order to generate a region combination parameter sequence. S503: Based on the region combination parameter sequence, the overall range of the anatomical region is called for all combination parameters in the sequence, and a summary processing is performed according to the value relationship between regions. The summarized values ​​are arranged in order to generate a combination of burn area and depth grades.

[0011] As a further aspect of the present invention, the burn wound refers to the skin area on the body surface of a burn patient where thermal damage exists, and the area is determined based on changes in skin color, the extent of epidermal damage, and texture differences in the depth image. The term "normal skin at the periphery" refers to the skin area located on the outer edge of the burn wound that has not been thermally damaged. The area is defined based on the portion of the image that retains normal skin color and structure. The height difference refers to the difference between the depth data of a single location in the burn wound area and the depth data of the corresponding location in the normal skin area at the edge of the wound. The wound height difference sequence refers to a sequence formed by arranging multiple wound height difference parameters in order of their corresponding image positions. The preset range refers to the height difference interval set according to the common height variation range of human skin depth mapping. The height variation characteristic refers to the parameter calculated from all height difference parameters within a single wound height range; The height hierarchy sequence refers to multiple levels arranged from low to high according to the height change characteristic parameters.

[0012] As a further aspect of the present invention, the three-dimensional node refers to a three-dimensional point generated from the spatial coordinates corresponding to the pixel positions in the depth image; The triangular mesh refers to a planar region formed by three three-dimensional geometric nodes of the wound surface; The wound height grid dataset refers to a data set consisting of all wound triangular grid regions and their corresponding height level information; The three-dimensional area refers to the area parameter calculated based on the spatial geometry of the triangular mesh region of the wound. The depth threshold refers to the depth grading boundary verified based on the clinical criteria for determining superficial second-degree, deep second-degree, and third-degree burns. The area-depth range refers to a data group composed of the three-dimensional area parameters of the wound triangular mesh region and the corresponding depth level; The anatomical regions refer to the head, trunk, upper limbs, and lower limbs regions as determined by the commonly used human body surface area division standards. The area ratio refers to the proportion of the wound area of ​​a single anatomical region relative to the total wound area; The depth grade refers to the clinical score value assigned based on the differential burn depth grade; The evaluation parameters refer to the combination of area proportion parameters and depth level scores. The burn area and depth grade combination refers to the quantitative result obtained by processing a set of assessment parameters based on multiple anatomical regions.

[0013] Systems for assessing burn conditions based on burn area and burn depth include: The wound height sequence module acquires depth images of the wound area and the normal skin area, detects the wound depth data and normal skin depth data at the locations in the images, calculates the height difference parameter based on the difference between the two types of depth data, and organizes the height difference parameter in spatial order to generate a wound height difference value sequence. The wound height interval module calls the wound height difference sequence, arranges it according to the size relationship of the sequence data, divides the height interval according to the preset height difference range, extracts the corresponding height difference parameters within the interval and calculates the height change feature parameters, establishes a hierarchical relationship according to the order of the height change feature parameters, and generates a height hierarchy sequence. The wound height grid module calls the height hierarchy sequence, sets three-dimensional geometric nodes for the image position according to the hierarchy information, organizes the spatial connection relationship of adjacent nodes and forms several triangular regions, and records the height hierarchy information of the regions to generate a wound height grid dataset. The three-dimensional area module of the wound calls the wound height grid dataset, calculates the three-dimensional surface area parameters for each triangular grid region and organizes them into a wound area data set. The area data set is superimposed to obtain the three-dimensional area parameters of the wound. Based on the height level information of the grid region and the clinical burn depth grading threshold, the depth level data is determined and the area depth range is generated. The burn grade combination module calls the area and depth range, organizes the three-dimensional surface area parameters of the corresponding triangular grid areas according to the anatomical regions of the head, trunk, upper limbs, and lower limbs, and calculates the area ratio parameters of the anatomical regions. The area ratio parameters are combined with the corresponding depth grade scores to form an assessment parameter set. The burn area and depth grade combination quantity is generated by summarizing the assessment parameter sets of all anatomical regions.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by extracting the depth difference between the wound and normal skin and forming a height sequence, the undulation state is continuously quantified. Based on the height characteristics, a hierarchy is constructed, making the deep and shallow structures clearer. The surface morphology is expressed with a three-dimensional grid, allowing the deformation details to be accurately presented. By calculating the three-dimensional area and combining it with depth partitioning, the damage distribution is shown. The area ratio is summarized according to the anatomical region and combined with the depth level, making the area calculation more accurate, the depth assessment more objective, and the overall load assessment more stable. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a method for assessing burn conditions based on burn area and burn depth, comprising the following steps: S1: Obtain depth image information of the burn wound area of ​​the burn patient, read the depth data of the location in the depth image, and at the same time read the depth data of the corresponding location of the normal skin area at the edge of the wound. Calculate the wound height difference parameter based on the difference between the two types of depth data, and organize the height difference parameter sequence according to the spatial position of the image to generate the wound height difference value sequence. The burn wound area refers to the area of ​​skin on the body surface of a burn patient where thermal damage has occurred. The area is determined based on changes in skin color, the extent of epidermal damage, and differences in texture in the depth image. The normal skin area at the edge of the wound refers to the skin area located on the periphery of the burn wound that has not been thermally damaged. The area is defined based on the part of the image that retains normal skin color and structure. The wound height difference parameter refers to the difference between the depth data of a single location in the burn wound area and the depth data of the corresponding location in the normal skin area at the edge of the wound. It is used to characterize the changes in the surface height of the local wound tissue. A wound height difference sequence is a sequence formed by arranging multiple wound height difference parameters in order of their corresponding image positions. It is used to describe the height variation distribution of the entire burn wound.

[0023] S2: Arrange the height difference parameters in numerical order according to the wound height difference sequence, divide the wound height into several intervals according to the preset height difference range, extract the corresponding height difference parameters in each height interval and calculate the height change feature parameters of the interval, arrange all intervals in order of height change feature parameters from low to high, and generate a height hierarchy sequence. The preset height difference range refers to the height difference interval set according to the common height variation range of human skin depth map imaging, which is used to classify and process the height difference parameter; The height change characteristic parameter refers to the parameter calculated from all height difference parameters within a single wound height range, which is used to reflect the overall height change trend of the range. A height hierarchy sequence refers to multiple levels arranged from low to high based on height variation characteristic parameters, used to describe the height distribution structure of wound tissue surface from shallow to deep.

[0024] S3: Based on the height hierarchy sequence, set the corresponding three-dimensional geometric nodes of the wound in space for the depth image position, assign height hierarchy information of position to each geometric node, generate a wound triangular mesh region composed of multiple small triangles according to the spatial connection relationship between adjacent geometric nodes, and record the corresponding height hierarchy information in all triangular mesh regions to generate a wound height mesh dataset. Three-dimensional geometric nodes of the wound refer to three-dimensional points generated from the spatial coordinates corresponding to the pixel positions in the depth image, which are used to construct the three-dimensional surface structure of the wound. The spatial connection relationship between adjacent geometric nodes refers to the node connection structure determined by the arrangement of adjacent pixels in the depth image, which is used to form a continuous three-dimensional surface; The wound triangular mesh region refers to a planar region formed by three three-dimensional geometric nodes of the wound, which is used to constitute the basic structural unit of the wound spatial surface; The wound height grid dataset refers to a data set consisting of all wound triangular grid regions and their corresponding height level information.

[0025] S4: Calculate the three-dimensional surface area parameters for each triangular grid region of the wound based on the wound height grid dataset and organize them into a wound area data set. Perform overlay processing on the wound area data set to obtain the three-dimensional area parameters of the wound. At the same time, read the height level information of each triangular grid region, determine the depth level of the region according to the clinically used burn depth grading threshold, and generate the area depth range. Three-dimensional surface area parameters refer to area parameters calculated based on the spatial geometry of the triangular mesh region of the wound, which are used to characterize the actual surface area of ​​the region. Burn depth grading thresholds refer to the depth grading boundaries verified based on clinical criteria for judging superficial second-degree, deep second-degree, and third-degree burns. They are used to determine the burn depth grade of a region based on its height level. Area-depth intervals refer to data groups composed of the three-dimensional area parameters of the wound triangular grid region and the corresponding depth level, used to describe the area-depth correspondence of different parts of the wound.

[0026] S5: Based on the area and depth range, organize the three-dimensional surface area parameters of the triangular grid area of ​​the wound according to the anatomical regions of the head, trunk, upper limbs and lower limbs, and calculate the area ratio parameters of the anatomical regions. Read the depth level of the corresponding anatomical region and refer to the clinically commonly used depth level score. Combine the area ratio parameter with the depth level score to form an assessment parameter set. Summarize the vital signs parameters in the entire anatomical region to generate the burn area and depth level combination quantity. Anatomical regions refer to the head, trunk, upper limbs, and lower limbs regions as determined by the commonly used human body surface area division standards; The area ratio parameter refers to the proportion of the wound area of ​​a single anatomical region to the total wound area, which is used to characterize the area distribution of burns throughout the body. The depth rating is a clinical score assigned based on the degree of burn difference, used to quantify areas of difference in burn depth. The assessment parameter set refers to the combination of area proportion parameters and depth grade scores, which is used to describe the burn characteristics of the anatomical region. The burn area and depth grade combination is a quantitative result obtained by processing a set of assessment parameters based on multiple anatomical regions, used to comprehensively describe the combination of burn area and depth in a patient.

[0027] The wound height difference sequence includes a wound pixel location index sequence, a corresponding height difference parameter sequence, and a normal skin reference height difference sequence. The height hierarchy sequence includes a height interval number sequence, an interval height change feature parameter sequence, and an interval sorting hierarchy label. The wound height grid dataset includes a wound geometric node spatial coordinate set, a wound triangular grid cell set, and a triangular grid cell height hierarchy annotation set. The area and depth intervals include shallow second-degree burn area intervals, deep second-degree burn area intervals, and third-degree burn area intervals. The burn area and depth grade combination includes an anatomical region area proportion parameter set, an anatomical region depth grade score set, and whole-body assessment parameter values.

[0028] Please see Figure 2 The specific steps of S1 are as follows: S101: Obtain depth image information of the burn wound area of ​​the burn patient, detect the depth data of the wound location in the depth image and detect the depth data of the corresponding location of the normal skin area at the edge of the wound, perform depth difference calculation based on the spatial correspondence between the wound pixels and the normal skin pixels, organize the calculation results according to the wound spatial index, and generate wound difference sequence values. First, a two-dimensional depth matrix is ​​established based on the depth camera output, and the wound location and normal skin edge location are determined using a clinical label map. Then, an index sequence is constructed using a point-by-point traversal method, and the wound depth and normal skin depth are extracted at each index. If some pixels do not have a corresponding normal skin location, the average depth of the neighborhood within a range of 1 to 2 pixels around that point is used as a substitute value. For example, if the wound depth at a certain index is 5.2 mm, and the surrounding normal skin depths are 3.0 mm, 3.2 mm, and 3.1 mm, then the average value of 3.1 mm is used as the normal skin depth for that index, and the difference is recorded as the negative difference obtained by subtracting 5.2 mm from 3.1 mm. The difference data set of all indices is obtained through point-by-point calculation. In the execution of the "calculation" action, only the subtraction is performed item by item according to the corresponding index position, and the result is written into the sequence, without calling external... The algorithm is based solely on depth subtraction. A clear standard must be established for judging the difference interval. For example, a difference greater than 2mm is defined as a higher interval, between -2mm and 2mm as a middle interval, and less than -2mm as a lower interval. This interval is determined based on the statistical range of skin depth differences in 200 forearm cases, making it a directly referable dataset. Interval classification allows for further screening of outliers. If a point's difference exceeds 8mm, it is considered an outlier and replaced in the difference sequence with the average difference of its five adjacent points through neighborhood interpolation. This 8mm threshold is selected from the combined range of approximately 1mm of the depth camera's standard accuracy and approximately 6mm of the most common indentation depth on the wound surface, with an additional 1mm redundancy coefficient to form the 8mm threshold. All indices are then rearranged in their initial order and output as a continuous difference sequence, which serves as the basis for subsequent segmentation.

[0029] S102: Based on the wound difference sequence value, call the difference set between indices in the sequence and perform segment detection based on the index interval. Perform calculation on the data set within the segment and arrange the calculation results in segment order to generate the segment mean sequence value. First, determine the index interval, for example, taking 20 pixels as a segment. This interval can be calculated based on the depth map resolution and the common wound area range to determine a suitable length, ensuring that a single segment covers a real area range of approximately 2 to 4 mm. Then, divide the sequence into multiple segments at equal intervals. Each segment contains a corresponding number of difference points. For example, a segment may contain 20 difference points, which may exhibit values ​​such as -2.1, -1.8, -2.2, and -1.9. The average difference for each segment needs to be calculated by summing all the differences within that segment and dividing by the number of points. If outliers exist within a segment, they are filtered out based on the fluctuation of the difference. For example, the fluctuation range of the difference in forearm skin wounds can be ±0.5 mm; points exceeding this range can be considered noise and removed from the segment. The average difference of the remaining points is then recalculated. For example, if 18 points remain after removing outliers, the accumulated value can be... If the difference is -36mm, then the average value is -2.0mm. The above calculation steps are automatically executed in the processing module of this invention according to preset calculation rules, program instructions or algorithms. The textual descriptions of accumulation, averaging and segment division are only used to illustrate the processing logic. The actual calculation process is completed by the system's software or hardware unit, including difference accumulation, outlier removal and segment result generation. That is, when the index accumulates to the segment length, it is automatically divided into the next segment. At the same time, in the "filtering" action, only the difference fluctuation range needs to be used as the benchmark, and no additional model support is required. All segment averages are arranged in the order of segment number to form a continuous average sequence. For example, the average difference of five segments may be -1.8mm, -2.0mm, -1.7mm, -2.3mm and -2.1mm, respectively. The sequence after sorting by segment order is the segment average result.

[0030] S103: Based on the average value of the segment sequence, perform sequence splicing on the segment data according to the wound space order, and perform position correspondence processing on the spliced ​​sequence according to the original coordinate order to generate a wound height difference sequence; First, expand the mean of each segment according to the segment order. For example, if the segment mean is -1.8mm and the corresponding segment length is 20 indices, then 20 consecutive -1.8mm values ​​need to be generated to fill the mapping interval of that segment. Then, for the second segment mean, such as -2.0mm, generate the corresponding number of repeated values ​​and place them immediately after the previous segment. By splicing the segments in this way according to the segment order, a new long sequence can be formed. At the same time, when dealing with the case where the length of the last segment is less than the length of the whole segment, it is necessary to repeat the corresponding mean only according to the remaining index number. For example, if the initial sequence length is 103 and the segment length is 20, then only 3 values ​​need to be generated for the last segment. For example, if the mean is -1.8mm... If 1.9mm is generated, three consecutive -1.9mm values ​​are filled into the end of the sequence. If a value jump occurs at the segment splicing point during the "adjustment" action, a simple linear smoothing method can be used for transition. For example, the average values ​​of two adjacent segments are mixed with a weight of 0.3 to form a transition point. This weight of 0.3 can be estimated based on the overall fluctuation range of the difference sequence of about 1mm. It is a medium-level smoothing coefficient that will not change the segment average value itself but can mitigate the sudden change of the segment connection point. Then, the spliced ​​long sequence is bound to the corresponding coordinates according to the original pixel coordinate order and the position correspondence processing is completed, so that all difference points return to the original coordinate system, forming a continuous height difference sequence that can be further processed.

[0031] Please see Figure 3 The specific steps of S2 are as follows: S201: Arrange the height difference parameters in the sequence according to the numerical order based on the wound height difference sequence. For the arranged parameters, call the preset height difference range and perform interval division based on the range boundary. Parameters falling into the same range are grouped into the same interval and sorted according to the range order to generate a set of height interval values. The height difference parameters are read one by one from the sequence and their original index positions are marked. For example, the height difference sequence may contain several point values ​​such as -3.2mm, -2.5mm, -1.8mm, -0.9mm, 0.2mm, 1.1mm, 1.9mm, 2.4mm, etc. These values ​​are all derived from the point-by-point comparison of the height difference between the wound area and the normal skin area, and are calculated by depth map difference method. Then, in the sorting action, all parameters need to be arranged in order from negative value to positive value and their mapping indexes need to be traceable. If two identical parameters appear during sorting, they are compared according to their index order as secondary rules. The above comparison operation is automatically executed by the processing module of the present invention through preset program instructions or comparison algorithms. The numerical comparison logic described in the instruction manual is only for explaining the sorting rules. In actual operation, the corresponding sorting and comparison operations are completed by computer software or hardware processing units. After sorting, the sorted parameters need to be divided into intervals according to the preset height difference range. The height difference range needs to be defined into multiple intervals based on the common depressions and bulges on the burn surface. For example, the low difference interval can be defined as -6mm to -3mm, the medium difference interval as -3mm to 0mm, the high difference interval as 0mm to 3mm, and the extremely high difference interval as greater than 3mm. These ranges can be set based on clinical wound height difference statistics so that the height difference can fall into the corresponding range under different lesion states. Subsequently, when dividing the intervals, it is necessary to check whether each height difference value falls into any of the four preset intervals and classify it accordingly. For example, - 3.2mm falls into the low difference range, -1.8mm into the medium difference range, 1.1mm into the high difference range, and 3.4mm into the extreme high difference range. Then, all parameters are reorganized according to the range order to form a data set within the range. In the "judgment" action, it is necessary to check whether each parameter exceeds the limit, that is, whether it exceeds all range boundaries. If values ​​such as -7.5mm or 5.1mm appear, they need to be marked as parameters outside the range and new extreme segments need to be determined according to the aforementioned statistical characteristics, or truncation processing is carried out according to the maximum and minimum boundaries. For example, -7.5mm is adjusted to -6mm as the lower limit replacement table, or 5.1mm is adjusted to 3mm as the upper limit replacement table. This adjustment needs to be based on the physical upper limit of height difference and the upper limit of depth camera measurement. All the parameters that have been divided are arranged in the range order to form a set of height range values.

[0032] S202: Based on the set of height interval values, all height difference parameters within the interval are called for the interval in the set, and the calculation is performed based on the parameter set of each interval to obtain the height change of the corresponding interval. All height changes are arranged in interval order to generate an interval change sequence. Each interval is processed individually. For example, the low-difference interval might contain values ​​such as -6.0mm, -5.4mm, -4.9mm, and -4.1mm; the medium-difference interval might contain values ​​such as -2.8mm, -1.9mm, and -0.7mm; the high-difference interval might contain values ​​such as 0.5mm, 1.1mm, and 1.8mm; and the extremely high-difference interval might contain values ​​such as 3.2mm and 3.7mm. The "calculation" action needs to derive the change in all parameters within each interval using a uniform method. The range of height difference variation can be represented by "the difference between the maximum and minimum values ​​within the interval," or by "the difference between the average value of the parameters within the interval and the theoretical center value of the interval." This should be explained in easily understandable and manually verifiable language. For example, if the maximum value in the low-difference interval is -4.1mm and the minimum is -6.0mm, the change is 1.9mm; if the maximum value in the medium-difference interval is -0.7mm and the minimum is -2.8mm, the change is 2.1mm; and if the maximum value in the high-difference interval is 1.8mm... The minimum value is 0.5mm, so the change is 1.3mm. The maximum value of the extreme height difference range is 3.7mm, and the minimum value is 3.2mm, so the change is 0.5mm. At the same time, clear judgment criteria need to be set for vague terms such as "relatively large" and "relatively small" in the interval change. For example, the change is defined as a high change range, the change is between 1mm and 2mm, the change is a medium change range, and the change is less than 1mm, the change is a low change range. This range is calculated based on the aforementioned height difference statistics. When performing interval processing, it is necessary to check whether there are abnormal data points in each interval. If an interval has extreme points such as -10mm or 9mm that are obviously inconsistent with the common depth difference, it is adjusted back to a reasonable range by averaging the neighborhood. For example, -10mm is adjusted to the average value of the parameters in the interval, such as -5.0mm, and 9mm is adjusted to the average value of the parameters in the interval, such as 3.4mm. This "adjustment" action is performed based on the statistical mean within the interval. Then, all interval changes are arranged in interval order and a change sequence is constructed to form a complete interval change sequence.

[0033] S203: Based on the interval change sequence, sort the changes within the sequence according to the order of the change magnitude, and establish a hierarchical order according to the sorted positions to generate a height hierarchy sequence; First, extract the values ​​from the change sequence item by item and sort them from smallest to largest. For example, the sequence may contain changes such as 0.5mm, 1.3mm, 1.9mm, and 2.1mm. In the "comparison" action, only the size of any two changes needs to be judged to complete the sorting description. If there are the same changes, such as two intervals with the same change of 1.3mm, retain their interval numbers as the order basis and sort them according to the rule of the earlier number first. Then, label the levels according to the sorted positions. For example, the smallest change is marked as the first level, the second smallest as the second level, and so on. If a change is 0.5mm, its level is 1; if it is 1.3mm, its level is 2 or 3. 3. Determined by the sorting position. In addition, during the hierarchical establishment process, a clear distinction must be made between terms such as "high-level" and "low-level". For example, a change greater than 2mm is defined as a high-level region, a change between 1mm and 2mm is defined as a mid-level region, and a change less than 1mm is defined as a low-level region. When performing hierarchical mapping, it is necessary to check whether there are boundary points. For example, when the change is exactly 1mm or 2mm, it can be included in the corresponding segment according to rounding or a fixed lower limit. For example, 1.0mm is included in the mid-level and 2.0mm is included in the high-level. This approach needs to be defined in advance and kept consistent. Then, all changes are formed into a stable corresponding hierarchical sequence according to the sorting position, and a continuous height hierarchical sequence is output.

[0034] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the height hierarchy sequence, set corresponding three-dimensional geometric nodes of the wound surface for the depth image position in space. For each geometric node, call the height hierarchy information corresponding to the node position in the height hierarchy sequence and perform association processing between the information and the node coordinates. Arrange all nodes in spatial order to generate a set of node hierarchy values. First, a 3D node framework consistent with the pixels is established based on the 2D coordinate grid of the depth image. Each node contains planar coordinates and extended coordinates in the depth direction. For example, when the depth map resolution is 240×240, 57600 basic nodes can be generated. The coordinates of each node are represented by (x, y, z), where x and y are derived from the pixel position index, and z can be initially set to 0 for subsequent height level mapping. Then, the level value consistent with the node position is extracted point by point from the height level sequence and written to the corresponding node. For example, if the coordinates of a node are (120, 85, 0), the k-th value in the height level sequence is consistent with the node position. For point index matching, if the level value is 3, then level 3 is recorded in the node attributes. If the node level is 1 or 2, it is written in the same way. The level here comes from the sorting steps of the preceding interval change and does not need to be recalculated; it only needs to be read by index. This "call" action is executed based on the one-to-one mapping relationship between node index and sequence index. Subsequently, in the "association processing" action, the node coordinates and level information need to be bundled to form a node data structure. For example, in textual terms, the node coordinates and level information together form a structure block, and the node number, spatial coordinates, and level are identified inside the structure block. The basic fields such as level are entered, and then the same processing is performed on all nodes. The entire operation is carried out in a spatial traversal manner, that is, each node is scanned sequentially from x=1 to x=240 and y=1 to y=240, and the level information is written one by one. During this process, if a node index is found to exceed the length range of the height level sequence, a value is selected from the level value of the neighboring node according to the nearest position replacement strategy. For example, when the sequence length is 56000 and the number of nodes is 57600, the last 1600 nodes can select the nearest level value based on their geometric distance from the neighboring nodes. For example, the level of the nearest node is 16000. For level 2, level 2 is selected for writing. Here, the distance comparison only needs to be compared numerically using the absolute value of the coordinate difference, without the need for complex calculations. In the "judgment" action, it is confirmed whether the distance is within an acceptable range. For example, if the distance difference is greater than 5 pixels, it can be set to an unacceptable range and switch to the average strategy, that is, take the average level of the five most recent available levels and round it to the nearest integer level. For example, if the five most recent levels are 3, 3, 2, 2, 3, then the average level is 2.6, and level 3 is taken as the node level. Then, all nodes are rearranged according to the spatial sorting method in a row-first-column manner to construct a complete node set and form a node level value set.

[0035] S302: Based on the set of node hierarchy values, call the spatial coordinates between adjacent geometric nodes in the set and perform connection judgment according to the spatial adjacency relationship of the nodes. Establish corresponding triangular structures for nodes that satisfy the adjacency relationship in a three-point combination method and record the hierarchy value of the nodes in the triangular structure to generate the triangular structure combination quantity. First, determine how to define the neighboring nodes of any given node. For example, the commonly used "four-neighbor" and "eight-neighbor" methods in depth maps can be used for screening. A four-neighbor node refers to a node that differs from the target node by one pixel in the up, down, left, or right directions. An eight-neighbor node adds four nodes diagonally opposite to the target node. To avoid an excessive number of nodes, the four-neighbor method can be prioritized for initial judgment. For example, the four-neighbor nodes of node (100, 100) are (99, 100), (101, 100), and (100, 99). For nodes (100, 101), check their existence one by one and read their level values. Then, when constructing the triangular structure, use a three-point combination mode. For example, combine the target node with any two of its adjacent nodes to form a triangular structure. For instance, nodes A (100, 100), B (101, 100), and C (100, 101) can be combined into a triangular unit. Then, record the level values ​​corresponding to the three nodes in the description of this triangular unit. For example, level A is level 3, level B is level 2, and level C is level 3. For level 3, the node hierarchy set of the triangular structure is [3, 2, 3]. In the "judgment" action, it is necessary to check whether the triangular combination satisfies geometric validity. For example, it is required that the three nodes cannot be collinear. The judgment method can be described in words as follows: if the three nodes have a completely consistent linear relationship in the x or y direction, it is considered an invalid triangle. This kind of judgment can be completed by comparing whether the x difference or y difference of the three nodes is zero at the same time. No geometric formula is required. When it is found that the spatial distance between a node and its neighboring nodes exceeds two pixels, the adjacent threshold can be set to 2 pixels. If the distance exceeds this, it is considered a non-adjacent point and does not participate in the combination. This threshold is set according to the sampling density of the depth map and the minimum distance between nodes is used as the reference value. For example, if the reference distance is 1 pixel, the maximum allowable distance can be set to 2 pixels to avoid excessively sparse data from destroying the consistency of the triangular combination. Then, a corresponding three-point structure is established for each valid triangular unit and all triangular structure information is recorded according to the node scanning order. These triangular structures are sorted by node number and spatial position to form the triangular structure combination quantity.

[0036] S303: Based on the combination quantity of the triangular structure, call the node level value for the triangular structure and perform the aggregation process according to the triangular structure order. Arrange the aggregated data of all triangular structures according to spatial position to generate the wound height grid dataset. The process involves sequentially reading the node hierarchy information within each triangular structure set and aggregating it into new data units according to the triangular structure order. For example, if a triangular structure contains nodes A, B, and C with hierarchies of 3, 2, and 3 respectively, it is aggregated into a triangular hierarchical unit denoted as [3, 2, 3]. For the next adjacent triangular structure with a node hierarchy of [2, 2, 3], it is followed to form a continuous data string. The aggregation process is completed in the order of the structure records throughout the traversal. This "aggregation" action can be performed by sequential reading and writing, without complex calculations. In the "arrangement" action, the aggregated data of all triangular structures must be sorted according to their spatial positions. Spatial positions can be obtained by comparing the geometric center points of the three nodes in the triangular structure. The geometric center point can be obtained by averaging the coordinates of the three nodes. For example, when the node coordinates are (100, 100), (101, 100), and (100, 101), the center point can be... The center point is set to (100.3, 100.3). When comparing the center points, only the size relationship between their x and y coordinates needs to be compared to complete the sorting. For example, the sorting is based on x first and then y, that is, the smaller x is first. If the x is the same, the y coordinate is used as the criterion. At the same time, for the phenomenon of consecutive high or consecutive low values ​​in the layer, the layer interval needs to be set. For example, the layer 1 to 2 is the low-level segment, the layer 3 to 4 is the medium-level segment, and the layer 5 and above is the high-level segment. This interval is set according to the commonly used division of wound depth gradient. If the layer combination of a certain triangular structure is all in the high level, such as [5, 6, 6], it is identified as a high-level triangular structure and arranged in the high-level grouping order in the data collection. If the combination is [1, 1, 2], it is classified into the low-level group and arranged according to the low-level rules. Then, all the triangular structure data collection is constructed into a new height grid data unit in the sorted order. The wound height grid dataset is formed by the sequential arrangement of all triangular structures.

[0037] Please see Figure 5 The specific steps of S4 are as follows: S401: Obtain the wound height grid dataset and call the node coordinate information for each triangular grid region in the dataset to perform area calculation. Arrange the area values ​​of all triangular grid regions into an area sequence in spatial order and perform cumulative processing based on all area values ​​in the sequence to generate the area accumulation. First, it is proven that the coordinates of the triangular nodes all originate from the previous spatial reconstruction process. For example, a triangular mesh cell consists of nodes A (105, 85, 3), B (106, 84, 2), and C (105, 84, 3). The area can be estimated based on the projection points of these three nodes in the XY plane. By geometrically comparing the coordinates of these three planes, it is determined whether they can form a valid triangle and whether they are collinear. For example, if the coordinate difference between A and B is (1, -1) and the coordinate difference between A and C is (0, -1), the directions of these two are inconsistent, which indicates that the three nodes are not collinear and meet the requirements of a valid triangular structure. Subsequently, during the "calculation" process, geometric area calculation methods described in words can be used. For example, the cross product length of two edge vectors can be used as the area reference, and the description can indicate that the area can be quantified using the horizontal and vertical differences between the two edges. For example, the horizontal difference of edge AB is 1 and the vertical difference is -1, and the horizontal difference of edge AC is 0 and the vertical difference is -1. When performing area calculation, two three-dimensional edge vectors are constructed based on the three-dimensional coordinates of the three nodes of the triangular mesh unit. For example, the node coordinates are denoted as A(x1, y1, z1), B(x2, y2, z2), and C(x3, y3, z3), thus obtaining vectors AB and AC. The 3D surface area of ​​the triangular mesh is calculated based on half the magnitude of the cross product of the two 3D vectors, using |AB×AC| / 2 as the area value. This ensures that the undulations along the Z-axis are fully accounted for, thus obtaining the true 3D wound surface area. This process is repeated for all triangular meshes. Furthermore, interval judgment rules need to be established for area size determination to facilitate subsequent processing. For example, areas less than 0.3 square units are classified as low-area intervals, those between 0.3 and 1.2 square units as medium-area intervals, and those greater than 1.2 square units as high-area intervals. This division is based on a depth map resolution of 240×240. Based on the sampling density of typical wound areas, it was calculated that the area of ​​common triangular mesh units falls between 0.2 and 1.5 square units. Subsequently, all obtained area values ​​were arranged in spatial order, which can be determined by sorting the coordinates of the center point of the triangular structure. For example, the center point can be the average of the X and Y coordinates of the three nodes and arranged in the order of X first and then Y, constructing the area sequence into a continuous sequence data. Finally, in the overlay process, all area values ​​in the sequence need to be numerically accumulated. For example, if the first five areas are 0.5, 0.6, 0.4, 0.8, and 0.7 square units respectively, the overlay amount can be recorded as 3.0 square units.

[0038] S402: Based on surface accumulation, the height level information corresponding to each triangular grid region in the wound height grid dataset is read and combined with the region area value. The combined data of all regions are arranged in the order of regions to generate a region combination quantity sequence. First, the level value of each triangular region is read. This level comes from the previous node level mapping step. For example, if a triangular mesh consists of three node levels, such as [3, 2, 3], the level of the triangular region can be determined by the majority level principle. For example, if level 3 appears twice in this example, the level of the triangular region is recorded as 3. When there are completely different three node levels, such as [1, 3, 5], the level of the triangular region can be determined by averaging the levels and rounding. For example, if the average value of this group of three nodes is 3, the level of the triangular region can be set to 3. This kind of "judgment" action is performed by numerical comparison and statistics, without the need for other tools. Subsequently, in the combination processing, the area value of the triangular region and its level need to be combined into a binary data unit. For example, if the area of ​​a region is 0.6 square units and the region level is 3, then the combination unit can be represented as (0.6, 3) Next, all regions are arranged in spatial order to form a sequence. If the area of ​​a continuous region is, for example, 0.4, 0.5, or 0.9 square units and the corresponding levels are 2, 3, and 3 respectively, then the combined sequence is written in order as (0.4, 2), (0.5, 3), and (0.9, 3). At the same time, during the combination process, specific intervals need to be set for vague terms such as "higher area", "lower area", and "higher level". For example, an area less than 0.3 square units is a low area segment, 0.3 to 1.2 square units is a medium area segment, and an area greater than 1.2 square units is a high area segment. Levels 1 to 2 are low levels, 3 to 4 are medium levels, and 5 and above are high levels. This partitioning is determined based on the commonly used classification standards for the degree of surface depression and depth changes of burn wounds. Then, all records in the sequence are arranged in order to form a region combination quantity sequence.

[0039] S403: Based on the region combination sequence, call the clinically used burn depth grading threshold for the height level information of each region in the sequence and perform grade determination based on the threshold. Then, perform corresponding processing on the determined depth grade and the region area value to generate the area depth interval. First, a clinical grading threshold range needs to be set. This range can be constructed based on the standard burn grading system. For example, basic grades (such as high grades 1 to 2) correspond to superficial burns, intermediate grades (high grades 3 to 4) correspond to moderate burns, and high grades (high grades 5 and above) correspond to deep burns. This threshold setting is based on commonly used grading standards for clinical depth changes and is consistent with the calculation system of the preceding height grades. During the judgment process, the grade of each region is read item by item. For example, if a region's grade is 3, it can be classified as a moderate burn according to the threshold; if the grade is 1, it is classified as a superficial burn; and if the grade is 6, it is classified as a deep burn. This "judgment" action is performed through a simple interval matching method and all recorded items are processed one by one. At the same time, it is necessary to note that if the grade value of a region is not... Within the standard range of 1 to 6, if a level of 0 or 8 appears, an "adjustment" action needs to be performed. This means adjusting level 0 to 1 as the lowest level and level 8 to 6 as the highest level. This adjustment can be based on the maximum and minimum boundaries of the standard clinical burn classification and requires no other additional processing. After the level determination is completed, the area of ​​each region needs to be correlated with its depth level. For example, if the area of ​​a region is 0.6 square units and the depth level is moderate burn, its area-depth unit is recorded as (0.6, moderate). If the area of ​​a region is 1.3 square units and the depth level is deep burn, the corresponding unit is (1.3, depth). By sequentially arranging the area-depth units of all regions according to the regional spatial arrangement, the area-depth interval dataset can be obtained.

[0040] Please see Figure 6 The specific steps of S5 are as follows: S501: Collect the corresponding anatomical information of the head, torso, upper limbs and lower limbs for the triangular mesh area according to the area and depth range, call all three-dimensional surface area parameters in the area to perform processing, calculate the proportion of the area parameters in the same anatomical area according to the total area of ​​the area and arrange them in the order of the area to generate the area proportion. First, the triangular mesh is mapped to a standard human anatomical region segmentation model based on its coordinates in the depth image. In two-dimensional plane coordinates, the human body region can be completely divided according to the vertical height ratio of the image, ensuring that the entire height range from 0% to 100% is covered. For example, the head region corresponds to the height range of 0% to 20% of the top of the image, the torso region corresponds to the height range of 20% to 65%, the lower limb region corresponds to the height range of 65% to 100%, and the upper limb region is divided horizontally according to the width range of 15% on the left and right sides of the image. This method eliminates undefined intervals in the vertical height direction, ensuring that all triangular meshes can be mapped to their corresponding anatomical regions. This type of region division is derived from commonly used human surface area rules, such as the "9% method," and extends the model, providing a basic index for subsequent region area combinations. Then, the geometric center position of each triangular mesh is read to determine if it falls within a specific anatomical region. Region matching is achieved by comparing the coordinates of the center point. For example, if the center point of a triangle is at 0.15 of the vertical height of the image, it is classified as a head region; if it is at 0.65, it is classified as a lower limb region. This "judgment" is performed by directly comparing numerical intervals, without relying on model calculations. After anatomical region matching is completed, the area and depth intervals of all triangular units within each region need to be read. The area values ​​are aggregated. For example, the head region contains areas of 0.3, 0.4, and 0.5 square units; the upper limb region contains areas of 0.8 and 1.1 square units; the torso region contains areas of 2.4, 2.8, and 3.1 square units; and the lower limb region contains areas of 1.9 and 2.2 square units. The total area of ​​each region is then summed. For example, the total area of ​​the torso region is 8.3 square units. If an area value significantly deviates from the normal range, such as a single triangular mesh exceeding 5 square units, an area threshold of 3 square units can be set, and areas exceeding the threshold are adjusted using a neighborhood average. For example, if an area is 4.6 square units, it is replaced by the average area of ​​the five surrounding adjacent meshes, such as 2.7 square units. This threshold is based on depth. Figure 3 The average area of ​​the corner grid is set within a range of 0.2 to 3.0 square units. This "adjustment" is done through substitution to maintain data rationality. Then, the area within the region is calculated as a percentage based on the total area of ​​the region. For example, if the total area of ​​the head is 1.2 square units, then the area of ​​the head is 0.3 square units, which is about 0.3 / 1.2, or 25%. The percentage can be expressed in words. All percentages are rearranged according to the conventional order of anatomical regions, such as head, trunk, upper limbs, and lower limbs, to form the region percentage.

[0041] S502: Based on the regional proportion, for each anatomical region, the depth level in the area depth interval is read and the commonly used clinical depth level score is called to perform score matching based on the depth level. The matched score value is combined with the corresponding regional proportion and sorted according to the region order to generate a regional combination parameter sequence. First, the depth grade of each region is read. For example, the head region may correspond to a moderate burn grade, the trunk region to a deep burn grade, the upper limb region to a superficial burn grade, and the lower limb region to a moderate burn grade. This depth grade is obtained from the preceding S403 classification. Next, a commonly used clinical scoring system needs to be set and scores mapped according to the depth grade. For example, a superficial burn is scored as 1 point, a moderate burn as 2 points, and a deep burn as 3 points. This scoring system is derived from representative depth quantification rules in multiple burn grading literatures. When more detailed depth grades exist, such as superficial I, superficial II, moderate II, deep II, and deep III, a more refined score range, such as 1 to 5 points, can be set. However, for example, a standard score of 1 to 3 is sufficient. Then, during the score matching process, mapping is performed item by item for each region. For example, if the trunk region corresponds to a deep burn, the score is 3; if the head region corresponds to a moderate burn, the score is 2; if the upper limb region corresponds to a deep burn, the score is 3. Superficial burns are scored as 1, while moderate burns in the lower limbs are scored as 2. This "judgment" is performed using an interval lookup table, without requiring additional complex calculations. Subsequently, the score value of each region needs to be combined with the region's proportion. For example, if the trunk's proportion is 40% and the score is 3, the combined unit can be written as (40%, 3). If the upper limb's proportion is 25% and the score is 1, the combined unit is (25%, 1). This combination can be stored in a textual two-parameter format. At the same time, if the proportion of a certain region is too low, such as less than 2%, a proportion threshold of 2% can be set, and that region can be recorded as an extremely low contribution interval. For example, if the threshold is set to 0.02, when the proportion is less than 0.02, its proportion can be fixed at 0.02 to avoid the distortion of subsequent data due to excessively low contribution. This threshold is obtained based on the clinical statistics that regions with an area of ​​less than 2% are classified as auxiliary fragments and can be used for illustrative purposes. Then, all combined units are arranged in anatomical region order to form a continuous sequence of region combined parameters.

[0042] S503: Based on the region combination parameter sequence, the overall range of the anatomical region is called for all combination parameters in the sequence, and the summary processing is performed according to the value relationship between regions. The summarized values ​​are sorted in order to generate the burn area depth grade combination quantity. First, the overall anatomical region needs to be defined, using four main regions: head, trunk, upper limbs, and lower limbs. Under each main category, all combined units are then grouped into a single station order. For example, if the head group is (25%, 2), the trunk group is (40%, 3), the upper limb group is (25%, 1), and the lower limb group is (30%, 2), then the percentages and scores of all regions need to be comprehensively analyzed during the summarization process. The summarization method can follow the clinically common "area-depth coupling model." For instance, the region percentages and scores are entered side-by-side in region order to form a depth level map, achieving value organization without averaging or weighting. If weighting coefficients need to be set, such as for the trunk... The system assigns a weight of 1.2 to different regions, 1.0 to the head region, 0.8 to the upper limbs, and 1.0 to the lower limbs. These weights are derived from the functional importance of different regions as defined in the BurnIndex literature and simplified using integer ratios in the examples. For instance, the trunk's weight of 1.2 is based on the statistical basis that trunk organs have high density and contribute the most area, while the upper limbs' weight of 0.8 is based on their relatively small average area of ​​involvement. When applying weights, for example, if the trunk region's original score is 3 and its area accounts for 40%, and the trunk's corresponding weight is 1.2, the system will calculate a weighted score of 3 × 1.2 = 3.6 according to a preset algorithm. This weighted score, as a quantified assessment parameter, can be directly used as a component of the combined score and for subsequent comprehensive assessment or diagnostic reference. The weighted results are automatically generated by the system and can be further stored, normalized, or included in the overall combination quantity summary as needed. Then, the combination quantities of all regions are rearranged in spatial order and merged into a single list, so that the list is presented in the form of [Head:(25%,2), Trunk:(40%,3.6), Upper Limb:(25%,0.8), Lower Limb:(30%,2)]. This list is used as the data structure for the combination quantity of burn area depth level.

[0043] Please see Figure 7 Systems for assessing burn conditions based on burn area and burn depth include: The wound height sequence module acquires depth images of the wound area and the normal skin area, detects the wound depth data and normal skin depth data at the locations in the images, calculates the height difference parameter based on the difference between the two types of depth data, and organizes the height difference parameter in spatial order to generate a wound height difference value sequence. The wound height interval module calls the wound height difference sequence, arranges it according to the size relationship of the sequence data, divides the height interval according to the preset height difference range, extracts the corresponding height difference parameters within the interval and calculates the height change feature parameters, establishes a hierarchical relationship according to the order of the height change feature parameters, and generates a height hierarchy sequence. The wound height grid module calls the height hierarchy sequence, sets three-dimensional geometric nodes for the image position according to the hierarchy information, organizes the spatial connection relationship of adjacent nodes and forms several triangular regions, and records the height hierarchy information of the regions to generate a wound height grid dataset. The three-dimensional area module of the wound calls the wound height grid dataset, calculates the three-dimensional surface area parameters for each triangular grid region and organizes them into a wound area data set. The area data set is superimposed to obtain the three-dimensional area parameters of the wound. Based on the height level information of the grid region and the clinical burn depth grading threshold, the depth level data is determined and the area depth range is generated. The burn grade combination module calls the area and depth range, organizes the three-dimensional surface area parameters of the corresponding triangular grid areas according to the anatomical regions of the head, trunk, upper limbs, and lower limbs, and calculates the area ratio parameters of the anatomical regions. The area ratio parameters are combined with the corresponding depth grade scores to form an assessment parameter set. The burn area and depth grade combination quantity is generated by summarizing the assessment parameter sets of all anatomical regions.

[0044] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for assessing burn conditions based on burn area and burn depth, characterized in that, Includes the following steps: S1: Acquire a depth image of the burn wound, read the depth data of the wound and the normal skin at the edge, calculate the height difference, perform anatomical region recognition on the corresponding human body parts in the depth image, obtain anatomical region label data corresponding to the pixel position, and form a wound height difference sequence according to the spatial position; S2: Based on the wound height difference sequence, sort the height differences by value and divide them into intervals according to a preset range. Extract parameters from the intervals and calculate the height change characteristics. Arrange the intervals from low to high according to the characteristic values ​​to form a height hierarchy sequence. S3: Based on the height hierarchy sequence, set three-dimensional nodes for the depth image position and assign height hierarchy, construct a wound triangular mesh based on the spatial connection of nodes, and record the height hierarchy information of all meshes to generate a wound height mesh dataset. S4: Calculate the three-dimensional surface area of ​​the triangular mesh based on the wound height grid dataset and form an area set. Accumulate the sets to obtain the three-dimensional area of ​​the wound. At the same time, read the height level and determine the depth level according to the clinical depth threshold to generate the area depth range. S5: Based on the area and depth range and anatomical region label data, organize the surface area of ​​the triangular grid according to the anatomical regions of the head, trunk, upper limb, and lower limb and calculate the area ratio. Read the region depth level and combine it with the clinical score to form assessment parameters and generate a combination of burn area and depth levels.

2. The method for assessing burn conditions based on burn area and burn depth according to claim 1, characterized in that, The wound height difference sequence includes a wound pixel location index sequence, a corresponding height difference parameter sequence, and a normal skin reference height difference sequence. The height hierarchy sequence includes a height interval number sequence, an interval height change feature parameter sequence, and an interval sorting hierarchy label. The wound height grid dataset includes a wound geometric node spatial coordinate set, a wound triangular grid cell set, and a triangular grid cell height hierarchy annotation set. The area depth interval includes a superficial second-degree burn area interval, a deep second-degree burn area interval, and a third-degree burn area interval. The burn area depth level combination includes an anatomical region area proportion parameter set, an anatomical region depth level score set, and whole-body assessment parameter values.

3. The method for assessing burn conditions based on burn area and burn depth according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain depth image information of the burn wound area of ​​the burn patient, detect the depth data of the wound location in the depth image and detect the depth data of the corresponding location of the normal skin area at the edge of the wound, perform depth difference calculation based on the spatial correspondence between the wound pixels and the normal skin pixels, organize the calculation results according to the wound spatial index, and generate wound difference sequence values. S102: Based on the wound difference sequence value, call the difference set between indices in the sequence and perform segment detection based on the index interval, perform calculation on the data set within the segment and arrange the calculation results in segment order to generate segment mean sequence value; S103: Based on the mean sequence value of the segment, perform sequence splicing on the segment data according to the wound space order, and perform position correspondence processing on the spliced ​​sequence according to the original coordinate order to generate a wound height difference sequence.

4. The method for assessing burn conditions based on burn area and burn depth according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Arrange the height difference parameters in the sequence according to the numerical order based on the wound height difference sequence. For the arranged parameters, call the preset height difference range and perform interval division based on the range boundary. Parameters falling into the same range are grouped into the same interval and arranged according to the range order to generate a set of height interval values. S202: Based on the set of height interval values, all height difference parameters within the interval are called for each interval and calculation is performed based on the parameter set of each interval to obtain the height change of the corresponding interval. All height changes are arranged in interval order to generate an interval change sequence. S203: Based on the interval change sequence, the changes within the sequence are arranged in order of magnitude and hierarchical order is established according to the arranged positions to generate a height hierarchy sequence.

5. The method for assessing burn conditions based on burn area and burn depth according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Based on the height hierarchy sequence, set corresponding three-dimensional geometric nodes of the wound surface for the depth image position in space. For each geometric node, call the height hierarchy information corresponding to the node position in the height hierarchy sequence and perform association processing between the information and the node coordinates. Arrange all nodes in spatial order to generate a set of node hierarchy values. S302: Based on the set of node hierarchy values, call the spatial coordinates between adjacent geometric nodes in the set and perform connection judgment according to the spatial adjacency relationship of the nodes. Establish corresponding triangular structures for nodes that satisfy the adjacency relationship in a three-point combination manner and record the hierarchy value of the nodes in the triangular structure to generate the triangular structure combination quantity. S303: Based on the triangular structure combination quantity, call the node level value for the triangular structure and perform aggregation processing on the level value according to the triangular structure order. Arrange the aggregated data of all triangular structures according to spatial position to generate a wound height grid dataset.

6. The method for assessing burn conditions based on burn area and burn depth according to claim 5, characterized in that, The specific steps of S4 are as follows: S401: Obtain the wound height grid dataset and call the node coordinate information for each triangular grid region in the dataset to perform area calculation. Arrange the area values ​​of all triangular grid regions into an area sequence in spatial order and perform cumulative processing based on all area values ​​in the sequence to generate surface accumulation. S402: Based on the surface accumulation, for each triangular grid region in the wound height grid dataset, read the corresponding height level information of the region and perform combination processing with the height level information and the region area value, arrange the combination data of all regions in the region order, and generate a region combination quantity sequence; S403: Based on the region combination sequence, call the clinically used burn depth grading threshold for the height level information of each region in the sequence and perform grade determination based on the threshold. Then, perform corresponding processing on the determined depth level and the region area value to generate an area depth interval.

7. The method for assessing burn conditions based on burn area and burn depth according to claim 6, characterized in that, The specific steps of S5 are as follows: S501: Collect the corresponding head, torso, upper limb and lower limb anatomical information for the triangular mesh area according to the area and depth range, call all three-dimensional surface area parameters in the area to perform processing, calculate the proportion of the area parameters in the same anatomical area according to the total area of ​​the area and arrange them in the order of the areas to generate the area proportion. S502: Based on the region proportion, for each anatomical region, read the depth level in the area depth interval and call the clinically commonly used depth level score to perform score matching according to the depth level. Combine the matched score value with the corresponding region proportion and organize it according to the region order to generate a region combination parameter sequence. S503: Based on the region combination parameter sequence, the overall range of the anatomical region is called for all combination parameters in the sequence, and a summary processing is performed according to the value relationship between regions. The summarized values ​​are arranged in order to generate a combination of burn area and depth grades.

8. The method for assessing burn conditions based on burn area and burn depth according to claim 1, wherein the burn wound refers to the skin area on the body surface of the burn patient with thermal damage, and the area is determined based on the skin color changes, epidermal damage range and texture differences in the depth image; The term "normal skin at the periphery" refers to the skin area located on the outer edge of the burn wound that has not been thermally damaged. The area is defined based on the portion of the image that retains normal skin color and structure. The height difference refers to the difference between the depth data of a single location in the burn wound area and the depth data of the corresponding location in the normal skin area at the edge of the wound. The wound height difference sequence refers to a sequence formed by arranging multiple wound height difference parameters in order of their corresponding image positions. The preset range refers to the height difference interval set according to the common height variation range of human skin depth mapping. The height variation characteristic refers to the parameter calculated from all height difference parameters within a single wound height range; The height hierarchy sequence refers to multiple levels arranged from low to high according to the height change characteristic parameters.

9. The method for assessing burn conditions based on burn area and burn depth according to claim 1, characterized in that, The three-dimensional node refers to a three-dimensional point generated from the spatial coordinates corresponding to the pixel position in the depth image; The triangular mesh refers to a planar region formed by three three-dimensional geometric nodes of the wound surface; The wound height grid dataset refers to a data set consisting of all wound triangular grid regions and their corresponding height level information; The three-dimensional area refers to the area parameter calculated based on the spatial geometry of the triangular mesh region of the wound. The depth threshold refers to the depth grading boundary verified based on the clinical criteria for determining superficial second-degree, deep second-degree, and third-degree burns. The area-depth range refers to a data group composed of the three-dimensional area parameters of the wound triangular mesh region and the corresponding depth level; The anatomical regions refer to the head, trunk, upper limbs, and lower limbs regions as determined by the commonly used human body surface area division standards. The area ratio refers to the proportion of the wound area of ​​a single anatomical region relative to the total wound area; The depth grade refers to the clinical score value assigned based on the differential burn depth grade; The evaluation parameters refer to the combination of area proportion parameters and depth level scores. The burn area and depth grade combination refers to the quantitative result obtained by processing a set of assessment parameters based on multiple anatomical regions.

10. A system for assessing burn conditions based on burn area and burn depth, characterized in that, The system is used to implement the method for assessing burn conditions based on burn area and burn depth as described in any one of claims 1-9, the system comprising: The wound height sequence module acquires depth images of the wound area and the normal skin area, detects the wound depth data and normal skin depth data at the locations in the images, calculates the height difference parameter based on the difference between the two types of depth data, and organizes the height difference parameter in spatial order to generate a wound height difference value sequence. The wound height interval module calls the wound height difference sequence, arranges it according to the size relationship of the sequence data, divides the height interval according to the preset height difference range, extracts the corresponding height difference parameters within the interval and calculates the height change feature parameters, establishes a hierarchical relationship according to the order of the height change feature parameters, and generates a height hierarchy sequence. The wound height grid module calls the height hierarchy sequence, sets three-dimensional geometric nodes for the image position according to the hierarchy information, organizes the spatial connection relationship of adjacent nodes and forms several triangular regions, and records the height hierarchy information of the regions to generate a wound height grid dataset. The three-dimensional area module of the wound calls the wound height grid dataset, calculates the three-dimensional surface area parameters for each triangular grid region and organizes them into a wound area data set. The area data set is superimposed to obtain the three-dimensional area parameters of the wound. Based on the height level information of the grid region and the clinical burn depth grading threshold, the depth level data is determined and the area depth range is generated. The burn grade combination module calls the area and depth range, organizes the three-dimensional surface area parameters of the corresponding triangular grid areas according to the anatomical regions of the head, trunk, upper limbs, and lower limbs, and calculates the area ratio parameters of the anatomical regions. The area ratio parameters are combined with the corresponding depth grade scores to form an assessment parameter set. The burn area and depth grade combination quantity is generated by summarizing the assessment parameter sets of all anatomical regions.

Images