Serum inflammatory factor detection method for perianal abscess postoperative infection condition analysis

Abstract

The present application relates to the technical field of protein detection, in particular to a serum inflammatory factor detection method for perianal abscess postoperative infection condition analysis, through obtaining serum detection data and wound exudate detection data of a target patient after operation, determining metabolic consumption index and comprehensive immune score of the target patient and mapping to a pre-constructed two-dimensional coordinate system, determining the target attribution cluster of the current mapping point in each preset cluster; when the target attribution cluster is the mixed cluster, the permeability correction is performed on the transmembrane concentration ratio of the common inflammatory marker; when the corrected concentration ratio meets the local active synthesis condition, the contribution concentration in the non-infection background is removed from the concentration of the common inflammatory marker in the serum detection data, and the net concentration of the common inflammatory marker contributed by the infection factor is obtained. The present application can separate the net concentration dominated by real infection in the serum inflammatory signal, and improve the reliability of serum inflammation detection.

Description

Technical Field

[0001] This invention relates to the field of protein detection technology, specifically to a method for detecting serum inflammatory factors for analyzing postoperative infection status in perianal abscesses. Background Technology

[0002] Postoperative open wounds typically result from perianal abscesses and anal fistulas. Due to their unique anatomical location, these wounds are highly exposed to the intestinal microecological environment. Under the continuous stimulation of intestinal contents and endotoxins, these wounds face a high risk of bacterial proliferation and infection. Currently, clinical assessment of postoperative wound infection status primarily relies on detecting serum inflammatory factors (such as IL-6, TNF-α, and CRP) concentrations. However, existing assessment systems based on serum inflammatory factors suffer from insufficient specificity and are mainly affected by interference from non-infectious signals in the following two aspects: One factor is non-infectious endotoxin stimulation. Postoperative wound contact with remnants of intestinal flora and their endotoxin components can non-specifically activate the body's innate immune system, triggering a compensatory increase in serum inflammatory factors. This non-proliferative, aseptic inflammatory response is easily misdiagnosed as pathogen infection by existing detection methods.

[0003] On the other hand, there is drug-induced therapeutic inflammation. When specific topical medications (such as traditional Chinese medicine preparations with the effect of promoting tissue regeneration) are applied to the wound, their pharmacological mechanism usually actively induces a moderate local inflammatory response to promote the shedding of necrotic tissue and tissue regeneration. This iatrogenic targeted treatment process also leads to a significant increase in the level of inflammatory factors in the patient's peripheral blood.

[0004] Under the interference of the aforementioned dual factors, the postoperative serum inflammatory factor signals obtained by conventional testing are essentially a superposition of multiple inflammatory signals. Current techniques typically attribute the elevation of total inflammatory markers such as IL-6 and CRP in serum solely to potential infection indicators, failing to effectively isolate and identify the differentiated components originating from drug treatment effects, non-infectious endotoxin stimulation, and bacterial proliferative infection within the mixed signal. Although these three biological processes differ fundamentally in terms of temporal dynamics, molecular proportions, local metabolic characteristics, and tissue damage mechanisms, the high degree of overlap in apparent inflammatory signals leads to a lack of specificity in traditional single-dimensional serological testing methods when differentiating between true infection and non-infectious inflammation, ultimately resulting in inaccurate serum inflammatory factor detection results. Summary of the Invention

[0005] To address the aforementioned technical problem of inaccurate serum inflammatory factor detection results due to interference factors, the present invention aims to provide a method for detecting serum inflammatory factors for analyzing postoperative infection status in perianal abscesses. The specific technical solution adopted is as follows: This invention provides a method for detecting serum inflammatory factors for analyzing postoperative infection status in perianal abscess surgery, comprising the following steps: Obtain postoperative serum and wound exudate data of the target patient at the current testing time point; The metabolic consumption index of the target patient is calculated based on the numerical differences between the energy metabolism characteristic parameters in the wound exudate detection data and the serum detection data. The comprehensive immune score is calculated based on the concentrations of each inflammatory marker and common inflammatory marker in the wound exudate detection data. The metabolic consumption index and the comprehensive immune score are mapped to a pre-constructed two-dimensional coordinate system. The distance between the current mapping point and each preset cluster is calculated, and the cluster with the smallest distance to the current mapping point is determined as the target cluster. The clusters include prototype clusters and hybrid clusters. When the target cluster is the mixed cluster, the transmembrane concentration ratio of the common inflammatory markers is permeability corrected based on the concentrations of vascular permeability reference markers in the wound exudate detection data and serum detection data. When the corrected concentration ratio meets the local active synthesis conditions, the contribution concentration of the common inflammatory markers in the non-infectious background is estimated, and the contribution concentration is removed from the concentration of common inflammatory markers in the serum detection data to obtain the net concentration of the common inflammatory markers contributed by the infectious factors.

[0006] Furthermore, the serum detection data includes the concentrations of glucose, lactate, albumin, C-reactive protein, and interleukin-6 in the serum; the wound exudate detection data includes the concentrations of glucose, lactate, albumin, interleukin-6, interleukin-1β, procalcitonin, and high-mobility group box B1 in the wound exudate, as well as the measured partial pressure of oxygen in the wound exudate.

[0007] Further, the metabolic consumption index of the target patient is calculated, including: Lactic acid, glucose, and oxygen partial pressure were used as the energy metabolism characteristic parameters. The ratio of lactic acid concentration in the wound exudate detection data to the serum detection data is calculated to obtain a first ratio. The ratio of glucose concentration in the wound exudate detection data to that in the serum detection data is calculated to obtain a second ratio. The ratio of the measured oxygen partial pressure in the wound exudate detection data to the preset oxygen partial pressure in normal tissue is calculated to obtain the third ratio. The first ratio, the second ratio, and the third ratio are combined to obtain the metabolic consumption index.

[0008] Further, the comprehensive immunity score is calculated, including: Interleukin-1β, procalcitonin, and high-mobility group box 1 were used as the respective inflammatory markers, and interleukin-6 was used as the common inflammatory marker. The ratio of the concentration of each of the inflammatory markers to the concentration of the common inflammatory markers in the wound exudate detection data was calculated and standardized to obtain the standardized ratio. The standardized ratios are weighted and fused using the corresponding feature coefficients of the pre-defined linear discriminant analysis model to obtain the comprehensive immune score.

[0009] Furthermore, the process of obtaining the corresponding feature coefficients of the pre-defined linear discriminant analysis model includes: Obtain the immune activation feature vector and category label for each typical sample in a set of historical typical samples that conform to the characteristics of different inflammatory stimulus prototypes; wherein, the immune activation feature vector is composed of the standardized ratio of the concentration of each of the inflammatory concomitant markers to the concentration of the common inflammatory marker in the wound exudate detection data of the typical sample; the category label includes infection prototype, endotoxin prototype and drug prototype; Based on the immune activation feature spectral vectors of typical samples of each category label, the intra-class scatter matrix within each category and the inter-class scatter matrix between different prototype categories are calculated. Based on the intra-class scatter matrix and the inter-class scatter matrix, obtain the optimal projection vector direction that maximizes the ratio of inter-class scatter to intra-class scatter. Based on the coefficient vector corresponding to the optimal projection vector direction, the corresponding feature coefficients of the preset linear discriminant analysis model are obtained.

[0010] Furthermore, the process of obtaining the prototype cluster and the hybrid cluster includes: Obtain the metabolic consumption index, comprehensive immune score, and category label for each typical sample in a historical typical sample set that conforms to the characteristics of different inflammatory stimulus prototypes; wherein, the category label includes infection prototype, endotoxin prototype, and drug prototype; Obtain the metabolic consumption index and comprehensive immune score for each mixture sample in the historical mixed sample set without class labels; A two-dimensional coordinate system is constructed with metabolic consumption index and comprehensive immune score as coordinate dimensions. Based on the metabolic consumption index and comprehensive immune score of typical samples in the historical typical sample set and mixed samples in the historical mixed sample set, the typical samples and the mixed samples are mapped to the two-dimensional coordinate system to obtain the mapping points of the typical samples and the mixed samples. Calculate the center coordinates of the mapping points of typical samples of each category label in the two-dimensional coordinate system, and use them as the reference coordinates for the infection prototype, endotoxin prototype and drug prototype. Cluster the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system to obtain several clusters and their cluster centers. Calculate the Euclidean distance from the center of each cluster to each reference coordinate. Label the three different clusters that are closest to the reference coordinates of the infection prototype point, endotoxin prototype point, and drug prototype point as the infection prototype cluster, endotoxin prototype cluster, and drug prototype cluster, respectively, and label the remaining clusters as mixed clusters.

[0011] Furthermore, clustering is performed on the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system, including: Determine the optimal number of clusters, wherein the optimal number of clusters is greater than the number of types of the prototype cluster; Based on the optimal number of clusters, k-means clustering is used to cluster the mapping points of all typical samples in the historical typical sample set and all mixed samples in the historical mixed sample set in the two-dimensional coordinate system.

[0012] Furthermore, permeability correction was performed on the transmembrane concentration ratios of common inflammatory markers, including: Interleukin-6 was used as the common inflammatory marker, and albumin was used as the reference marker for vascular permeability. The concentration ratio of common inflammatory markers in the wound exudate detection data and the serum detection data is calculated to obtain the transmembrane concentration ratio. The albumin concentration ratio in the wound exudate detection data and the serum detection data is calculated to obtain the albumin leakage index, which is used to characterize capillary permeability. The albumin leakage index is used to correct the transmembrane concentration ratio to obtain the corrected concentration ratio.

[0013] Furthermore, the contribution concentrations of common inflammatory markers in a non-infectious background were estimated, including: The concentration of the target inflammatory marker in the wound exudate detection data is multiplied by the albumin leakage index to obtain the local drug signal intensity; The local drug signal intensity and the concentration of C-reactive protein in the serum detection data are weighted and fused to obtain the contribution concentration of common inflammatory markers in the non-infectious background.

[0014] Furthermore, the method also includes: The infection contribution ratio is obtained by calculating the ratio of the net concentration of the common inflammatory marker contributed by the infectious factor to the concentration of the common inflammatory marker in the serum detection data; When the infection contribution ratio is greater than or equal to the preset contribution ratio, the infection dominance determination result is output; otherwise, the systemic synthesis dominance determination result is output.

[0015] This invention also provides a serum inflammatory factor detection system for analyzing postoperative infection status in perianal abscess surgery, the system comprising: Data acquisition module 1 is used to acquire serum test data and wound exudate test data of the target patient at the current test time point after surgery; The metabolic consumption analysis module 2 is used to calculate the metabolic consumption index of the target patient based on the numerical difference between the energy metabolism characteristic parameters in the wound exudate detection data and the serum detection data. Immunoanalysis module 3 is used to calculate the comprehensive immune score based on the concentrations of each inflammatory marker and common inflammatory marker in the wound exudate detection data; The cluster affiliation analysis module 4 is used to map the metabolic consumption index and the comprehensive immune score to a pre-constructed two-dimensional coordinate system, calculate the distance between the current mapping point and each preset cluster, and determine the cluster with the smallest distance to the current mapping point as the target affiliation cluster; wherein, the clusters include prototype clusters and hybrid clusters; The net concentration identification module 5 is used to perform permeability correction on the transmembrane concentration ratio of the common inflammatory markers based on the concentrations of vascular permeability reference markers in the wound exudate detection data and serum detection data when the target cluster belongs to the mixed cluster; when the corrected concentration ratio meets the local active synthesis conditions, it estimates the contribution concentration of the common inflammatory markers in the non-infectious background, and removes the contribution concentration from the concentration of the common inflammatory markers in the serum detection data to obtain the net concentration of the common inflammatory markers contributed by the infectious factors.

[0016] The present invention also provides a computer-readable storage medium storing computer program code that, when executed on a computer, causes the computer to perform the steps of the serum inflammatory factor detection method described above for analyzing postoperative infection status of perianal abscess.

[0017] This invention has the following beneficial effects: By simultaneously acquiring serum and wound exudate detection data of target patients after surgery, the metabolic consumption index is calculated using the numerical differences of energy metabolism characteristic parameters, and a comprehensive immune score is calculated by combining the concentrations of inflammatory concomitant markers and common inflammatory markers. This achieves objective quantification of the inflammatory state of the microenvironment from two dimensions: bacterial metabolic activity and immune pathway activation characteristics. Furthermore, by mapping the above two-dimensional parameters to a pre-constructed two-dimensional coordinate system for cluster distance comparison, preliminary qualitative screening of single-origin inflammation and complex mixed inflammation is achieved. For mixed cluster samples with mixed signals, a vascular permeability reference marker is introduced to correct the transmembrane concentration ratio, effectively eliminating the interference of physical passive leakage caused by capillary barrier damage. Furthermore, when the local active synthesis conditions are met, by estimating and eliminating the contribution concentration of non-infectious background such as drugs and endotoxins, the net concentration driven by bacterial infection is successfully separated from the multi-source superimposed serum signals. This effectively solves the technical problem of false positives and assessment confusion of inflammatory markers caused by endotoxin contamination and local drug treatment after perianal abscess surgery, and significantly improves the specificity and accuracy of postoperative true infection assessment. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the embodiments of the present invention include the following figures: Figure 1 This is a flowchart of the steps of the serum inflammatory factor detection method for analyzing postoperative infection status of perianal abscess according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining the metabolic consumption index according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating the process of obtaining the comprehensive immune score according to an embodiment of the present invention. Figure 4 This is a flowchart illustrating the steps of permeability correction for the transmembrane concentration ratio of common inflammatory markers according to an embodiment of the present invention. Figure 5 This is a flowchart illustrating the steps for estimating the contribution concentration of common inflammatory markers in a non-infectious background, according to an embodiment of the present invention. Figure 6 This is a schematic diagram of the serum inflammatory factor detection system for analyzing postoperative infection status of perianal abscess according to an embodiment of the present invention. Detailed Implementation

[0019] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings.

[0020] To address the issue of inaccurate serum inflammatory factor detection results due to interference factors, this invention provides a method for analyzing postoperative infection status in perianal abscesses. This method analyzes serum and wound exudate data from postoperative target patients to quantitatively separate the net concentration dominated by actual infection from mixed serum inflammatory signals, thus providing objective and accurate auxiliary interpretation support for easily confused inflammatory detection data in postoperative patients.

[0021] The following will describe in detail, with reference to the accompanying drawings, a method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscesses provided by an embodiment of the present invention.

[0022] Figure 1 This invention provides a schematic flowchart of a method for detecting serum inflammatory factors to analyze postoperative infection status in perianal abscesses, according to an embodiment of the present invention. The method specifically includes the following steps: Step S1: Obtain serum test data and wound exudate test data of the target patient at the current test time point after surgery.

[0023] For target patients after perianal abscess surgery, serum and wound exudate data are obtained at specific postoperative time points (e.g., days 1, 3, 5, or 7, selected based on clinical needs). Serum data refers to the concentrations of various biochemical substances and inflammatory immune factors in serum separated from the patient's peripheral venous blood. Wound exudate data refers to the concentrations and physical parameters of various biochemical substances, metabolites, and inflammatory immune factors in the supernatant obtained from inflammatory exudate collected from open surgical wounds after centrifugation to remove cell debris.

[0024] Furthermore, the serum detection data includes the concentrations of glucose, lactate, albumin, C-reactive protein, and interleukin-6 in the serum; the wound exudate detection data includes the concentrations of glucose, lactate, albumin, interleukin-6, interleukin-1β, procalcitonin, and high-mobility group box B1 in the wound exudate, as well as the measured partial pressure of oxygen in the wound exudate.

[0025] The serum test data mentioned above were obtained from peripheral venous blood samples of the target patient at the current testing time point after centrifugation and biochemical analysis. For example, peripheral venous blood samples of the target patient collected at the current testing time point were separated into serum samples after being allowed to stand at room temperature and centrifuged (e.g., centrifuged at 3000 rpm for 10 min). Biochemical and immunological tests were performed on the serum samples to obtain the concentrations of indicators such as glucose (Glu), lactate (Lac), albumin, C-reactive protein (CRP), and interleukin-6 (IL-6), thereby obtaining the serum test data.

[0026] The above-mentioned wound exudate detection data are obtained from deep wound exudate samples of the target patient at the current detection time point after elution, centrifugation, and biochemical analysis. For example, for exudate collected from the deep wound of the target patient at the current detection time point using a sterile sampling tool (such as a sterile cotton swab), the exudate sample is placed in a sterile elution buffer and eluted by shaking. It is then subjected to low-temperature centrifugation (e.g., centrifugation at 12000 rpm for 10 min at 4°C) to remove cell debris and bacterial remains, and the supernatant is used as the test sample. Biochemical and immunological tests are performed on the supernatant to obtain the concentrations of glucose (Glu), lactate (Lac), albumin, interleukin-6 (IL-6), interleukin-1β (IL-1β), procalcitonin (PCT), and high-mobility group box 1 (HMGB1) in the exudate, thus obtaining the wound exudate detection data. In addition, after obtaining the collected exudate samples, blood gas analysis technology was immediately used to measure and record the measured oxygen partial pressure (pO2) of the wound exudate.

[0027] Step S2: Calculate the metabolic consumption index of the target patient based on the numerical difference between the energy metabolism characteristic parameters in the wound exudate detection data and the serum detection data.

[0028] Bacterial proliferation is not only a pathological process but also an energy-consuming process that consumes substrates (glucose and oxygen) and excretes metabolic waste (lactic acid). When bacterial proliferation occurs at the wound site, active bacterial metabolism consumes large amounts of glucose and oxygen in the local microenvironment, while simultaneously generating lactic acid. This abnormal local consumption and accumulation disrupts the transmembrane diffusion balance, resulting in an abnormally high lactic acid concentration in the exudate compared to serum, while glucose concentration and oxygen partial pressure are significantly lower than peripheral circulation levels. Therefore, this embodiment constructs a metabolic consumption index for the target patient by comparing the numerical differences between the aforementioned metabolic substrates and products in the exudate and serum, thereby objectively quantifying the degree of bacterial metabolic activity at the lesion site.

[0029] Furthermore, such as Figure 2 As shown, step S2 calculates the metabolic consumption index of the target patient, including: using lactate, glucose, and oxygen partial pressure as the energy metabolism characteristic parameters; calculating the ratio of lactate concentration in the wound exudate detection data and the serum detection data to obtain a first ratio; calculating the ratio of glucose concentration in the wound exudate detection data and the serum detection data to obtain a second ratio; calculating the ratio of the measured oxygen partial pressure in the wound exudate detection data to the preset oxygen partial pressure of normal tissue to obtain a third ratio; and fusing the first ratio, the second ratio, and the third ratio to obtain the metabolic consumption index.

[0030] When exudate glucose decreases, lactic acid increases, and oxygen partial pressure decreases, it indicates high bacterial metabolic activity. The corresponding first ratio increases, while the second and third ratios decrease. The metabolic consumption index of the target patient will increase, suggesting active bacterial metabolic activity. The formula for calculating the metabolic consumption index of the target patient is:

[0031] in: The metabolic consumption index of the target patient; This represents the concentration of lactic acid in the wound exudate test data; This refers to the concentration of lactate in serum test data; The concentration of glucose in the wound exudate test data; This refers to the concentration of glucose in serum test data; This represents the measured value of oxygen partial pressure in the wound exudate detection data; This is the preset normal tissue oxygen partial pressure.

[0032] It should be understood that the ratios used in the above calculation of the metabolic consumption index are all theoretical expressions. In order to avoid abnormal calculations when the denominator is 0, when the computer actually executes instructions involving division or ratios, a preset non-zero minimum threshold (such as the minimum detection limit value of the instrument corresponding to this parameter, or 0.01) will be automatically added to the corresponding denominator as a smoothing factor.

[0033] Step S3: Calculate the comprehensive immune score based on the concentrations of each inflammatory marker and common inflammatory marker in the wound exudate detection data.

[0034] Common inflammatory markers (such as IL-6) are public pathway indicators that are generally expressed under various pro-inflammatory factors such as pathogen infection, drug stimulation, or endotoxin exposure, and mainly reflect the overall baseline intensity of the local inflammatory response. Associated inflammatory markers (such as IL-1β, PCT, and HMGB1) are specific indicators that are selectively released depending on the activation of specific immune pattern receptors or specific cytopathological events (such as bacterial-specific proliferation, aseptic tissue decay, or late-stage cell necrosis). The comprehensive immune score reflects the specific immune activation pattern (i.e., the fingerprint characteristics of the inflammatory nature) of the patient's wound.

[0035] Although various inflammatory stimuli (pathogen infection, drug stimulation, or endotoxin exposure) induce a general increase in common inflammatory markers such as interleukin-6, the composition ratio of specific inflammatory concomitant markers (such as PCT reflecting infection and IL-1β reflecting tissue damage) varies significantly due to differences in the immune pattern recognition receptors they activate. By analyzing the concentration distribution of various inflammatory concomitant markers and common inflammatory markers, the absolute differences in basic inflammation intensity among individuals can be effectively eliminated, the characteristics that truly reflect the immune activation pattern can be purified, and the comprehensive immune score can be determined, thereby achieving quantitative decoding of complex inflammatory signals.

[0036] Furthermore, such as Figure 3 As shown, step S3 calculates the comprehensive immune score, including: using interleukin-1β, procalcitonin, and high-mobility group box 1 as the respective inflammatory markers, and interleukin-6 as the common inflammatory marker; calculating the ratio of the concentration of each inflammatory marker to the concentration of the common inflammatory marker in the wound exudate detection data, and performing standardization processing to obtain standardized ratios; and using the corresponding characteristic coefficients of a preset linear discriminant analysis model to perform weighted fusion of each standardized ratio to obtain the comprehensive immune score.

[0037] Inflammatory stimuli from different sources activate different immune pathways, leading to differences in the inflammatory characteristics of the local microenvironment. Interleukin-6 (IL-6) is a common signaling pathway, and almost all inflammatory stimuli (infection, trauma, endotoxins, drug stimulation) can induce the production of IL-6. Interleukin-1β (IL-1β) is a signal of aseptic inflammation or tissue damage, procalcitonin (PCT) is a specific marker of bacterial infection, and high-mobility group box 1 (HMGB1) is a signal of late-stage inflammation and necrosis. Therefore, IL-6 in exudate is considered a common inflammatory marker induced by various inflammatory stimuli, and IL-1β, PCT, and HMGB1 in exudate are considered concomitant markers of various inflammations.

[0038] The ratios of the concentrations of three inflammatory markers (IL-1β, PCT, and HMGB1) to the concentration of the common inflammatory marker IL-6 in wound exudate data were calculated to quantify the contribution of each indicator to overall inflammation. It should be understood that, to avoid aberrant calculations with a denominator of 0, the aforementioned data preprocessing rules were applied to automatically add a preset non-zero minimum threshold to the corresponding denominator when calculating these ratios. The ratios were then Z-score standardized to eliminate the influence of dimensions, resulting in three standardized ratios. The mean and standard deviation used in Z-score standardization were pre-fixed constant parameters, determined by statistically analyzing the overall mean and standard deviation of the three concentration ratios on a historical typical sample set selected by experts. The three standardized ratios were combined to generate the immune activation feature vector of the current target patient. ,in, These are the standardized ratios of the concentrations of IL-1β, PCT, HMGB1, and IL-6 in the wound exudate detection data.

[0039] A pre-defined linear discriminant analysis (LDA) model is obtained. The dimension of each feature coefficient in this model is equal to the number of standardized ratios, representing the weighted contribution of each standardized ratio in distinguishing different sources of inflammation (infection, endotoxin, drug). This model is used to linearly combine and dimensionality-reduced map the three-dimensional immune activation feature vectors into a one-dimensional comprehensive immune score. Using the corresponding feature coefficients of this LDA model, the immune activation feature vectors are... The three standardized ratios are weighted and summed to obtain the comprehensive immunity score. .in, These are the standardized ratios of the concentrations of IL-1β, PCT, HMGB1, and IL-6 in the wound exudate detection data, respectively, and their corresponding characteristic coefficients in the linear discriminant analysis model.

[0040] Furthermore, the process of obtaining the corresponding feature coefficients of the preset linear discriminant analysis model includes: obtaining the immune activation feature vector and category label of each typical sample in the historical typical sample set that conforms to the characteristics of different inflammatory stimulus prototypes; wherein, the immune activation feature vector is composed of the standardized ratio of the concentration of each inflammatory concomitant marker to the concentration of the common inflammatory marker in the wound exudate detection data of the typical sample; the category label includes infection prototype, endotoxin prototype, and drug prototype; based on the immune activation feature vector of the typical sample of each category label, the intra-class divergence matrix within each category and the inter-class divergence matrix between different prototype categories are calculated; based on the intra-class divergence matrix and the inter-class divergence matrix, the optimal projection vector direction that maximizes the ratio of inter-class divergence to intra-class divergence is obtained; based on the coefficient vector corresponding to the optimal projection vector direction, the corresponding feature coefficients of the preset linear discriminant analysis model are obtained.

[0041] Based on the extracted metabolic consumption index and immune activation feature vector, the features of the three prototypes are as follows: 1. Infection prototype: high LMI + immune spectrum features [medium, high, high]; 2. Endotoxin prototype: low LMI + immune spectrum features [low, low, low]; 3. Drug prototype: medium LMI + immune spectrum features [high, very low, medium].

[0042] Based on expert experience, three sets of historical typical samples were pre-selected that conform to the characteristics of the three types of inflammatory stimuli mentioned above. Each typical sample in the three sets corresponds to a patient who has undergone perianal abscess surgery, and the sample collection time of the typical sample is at the same postoperative time point as the current detection time of the target patient. The category label of each typical sample in the three sets of historical typical samples was obtained (label 1 for infection prototype, label 2 for endotoxin prototype, and label 3 for drug prototype). Following the same method used to obtain the immune activation feature vector of the target patient, the immune activation feature vector of each typical sample was obtained. Using the category label of the typical samples in the three sets of historical typical samples as the dependent variable and the three standardized ratios in the immune activation feature vector of the typical samples as independent variables, linear discriminant analysis (LDA) was performed to obtain the discriminant function coefficient vector. The discriminant function coefficient vector maximizes the inter-class distance and minimizes the intra-class distance of the three types of historical typical sample sets in the projection direction. In The three vector elements serve as the corresponding feature coefficients of this linear discriminant analysis model.

[0043] Suppose that 90 typical samples are pre-selected through expert experience, with 30 samples of each of the three types, labeled Y=1, 2, and 3 respectively, resulting in a matrix of 90 rows × 4 columns (3 features X + 1 label Y):

[0044] When this matrix is ​​input into the LDA algorithm, the algorithm calculates the intra-class and inter-class scatter matrices of the sample points within each category (infection, endotoxin, or drug). Then, by solving the generalized eigenvalue problem, it extracts the optimal projection vector direction corresponding to the largest eigenvalue, which maximizes the ratio of inter-class scatter to intra-class scatter. The coefficient vector corresponding to this optimal projection vector direction (with dimensions consistent with the three ratio features) can distinguish the three categories of samples as a whole and capture the typical high-low relationship of each ratio feature in different categories, preserving detailed information of the immune feature spectrum. Therefore, this coefficient vector is used as the coefficient vector of the aforementioned discriminant function. Since the LDA operation process is existing technology, it will not be elaborated on further here.

[0045] Step S4: Map the metabolic consumption index and the comprehensive immune score to a pre-constructed two-dimensional coordinate system, calculate the distance between the current mapping point and the preset centers of each cluster, and determine the cluster with the smallest distance to the current mapping point as the target cluster; wherein, the clusters include prototype clusters and hybrid clusters.

[0046] The metabolic consumption index and the comprehensive immune score quantify the inflammatory state of the wound microenvironment from two independent dimensions: the local metabolic activity of bacteria and the activation characteristics of host immune pathways. In the two-dimensional feature space constituted by the metabolic consumption index and the comprehensive immune score, spatial geometric distance directly characterizes the similarity of pathophysiological states. Typical inflammation from a single source clusters to form specific prototypical clusters, while complex inflammation from multiple sources is distributed in transitional mixed clusters. Using the nearest neighbor distance metric, the current mapping point of the target patient in the two-dimensional feature space is matched to the prototypical and mixed clusters of pre-acquired known state patterns, thereby achieving qualitative classification of the current mapping point according to the principle of maximizing similarity.

[0047] Further, the process of obtaining the prototype cluster and the hybrid cluster in step S4 includes: obtaining the metabolic consumption index, comprehensive immune score, and category label of each typical sample in the historical typical sample set that conforms to the prototype characteristics of different inflammatory stimuli; wherein, the category label includes infection prototype, endotoxin prototype, and drug prototype; obtaining the metabolic consumption index and comprehensive immune score of each hybrid sample in the historical hybrid sample set without category labels; constructing a two-dimensional coordinate system with the metabolic consumption index and comprehensive immune score as the coordinate dimension, and mapping the typical samples and the hybrid samples to the two-dimensional coordinate system according to the metabolic consumption index and comprehensive immune score of the typical samples in the historical typical sample set and the hybrid samples in the historical hybrid sample set. The mapping points of the typical samples and the mixed samples are obtained; the center coordinates of the mapping points of the typical samples of each category label in the two-dimensional coordinate system are calculated as the reference coordinates of the infection prototype, endotoxin prototype and drug prototype; the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system are clustered to obtain several clusters and their cluster centers; the Euclidean distance from each cluster center to each reference coordinate is calculated, and the three different clusters closest to the reference coordinates of the infection prototype point, endotoxin prototype point and drug prototype point are marked as the infection prototype cluster, endotoxin prototype cluster and drug prototype cluster, respectively, and the remaining other clusters are marked as mixed clusters.

[0048] For the three sets of historical typical samples pre-screened by expert experience in step S3, which conform to the characteristics of different inflammatory stimulus prototypes, based on the category labels of each typical sample in the three sets of historical typical samples, the metabolic consumption index and comprehensive immune score of each typical sample in the three sets of historical typical samples are obtained in the same way as the metabolic consumption index and comprehensive immune score of the current target patient. At the same time, the metabolic consumption index and comprehensive immune score of each mixed sample in the historical mixed sample set without category labels are obtained. The historical mixed sample set consists of real clinical case samples that were in a state of multiple inflammatory stimuli superimposed after similar surgeries (such as local topical drug intervention and mild intestinal flora colonization at the same time), or past clinical case samples that cannot be clearly classified into a single infection prototype, endotoxin prototype, or drug prototype after expert evaluation.

[0049] Because the metabolic consumption index and the comprehensive immune score differ significantly in magnitude and fluctuation range, to avoid a single dimension having absolute weight in distance calculation, the metabolic consumption index and comprehensive immune score of all typical and mixed samples are normalized (e.g., Min-Max normalization). The maximum and minimum values ​​used in Min-Max normalization are the historical global maximum and minimum values ​​obtained by statistically analyzing the metabolic consumption index and comprehensive immune score of all historical samples (including typical and mixed samples). A two-dimensional coordinate system is constructed with the normalized metabolic consumption index as the first dimension and the normalized comprehensive immune score S as the second dimension. All typical and mixed samples are then projected onto this two-dimensional coordinate system. The projected points of all typical and mixed samples in this two-dimensional coordinate system are then clustered into k clusters. When performing clustering, the elbow rule or silhouette coefficient can be used to determine the optimal number of clusters k. To ensure that there are enough clusters to cover the typical prototype and multi-source mixed state, the optimal number of clusters k is set to be greater than the number of types of prototype clusters, that is, the optimal number of clusters k must be greater than 3, such as setting the value range of k to [5,7]. Then, the k-means clustering algorithm is used to cluster all the projection points in the two-dimensional coordinate system to obtain k clusters.

[0050] The center coordinates of all projection points of the three types of historical typical sample sets in a two-dimensional coordinate system are determined respectively, serving as the reference coordinates for the infection prototype, endotoxin prototype, and drug prototype, and are denoted as follows: For the k clusters obtained from the above clustering, the Euclidean distance from the cluster center of each cluster to the three reference coordinates is calculated. Then, a greedy strategy is adopted to assign the nearest cluster to each prototype among the infection prototype, endotoxin prototype, and drug prototype, ensuring that the three prototypes correspond to different clusters. For example, find the clusters closest to the reference coordinates. The nearest cluster is marked as the infected prototype cluster, and then the clusters closest to the reference coordinate are found sequentially from the remaining clusters. The most recent clusters are labeled as the endotoxin prototype cluster and the drug prototype cluster, respectively. All remaining clusters are labeled as mixed clusters, representing transitions where multiple signals are superimposed or features are shifted.

[0051] To achieve a preliminary qualitative assessment of the source of inflammation, the metabolic consumption index and comprehensive immune score of the target patient were subjected to the same normalization (e.g., Min-Max normalization) and projected onto a two-dimensional coordinate system. The Euclidean distance between the target patient's current mapping point in the two-dimensional coordinate system and the cluster centers of all pre-clustered clusters (i.e., k clusters) was calculated. The cluster with the smallest distance to the current mapping point was determined, and the category label (infection prototype cluster, endotoxin prototype cluster, drug prototype cluster, or mixed cluster) assigned to this cluster during the clustering stage was used as the target cluster.

[0052] Step S5: When the target cluster is the mixed cluster, based on the concentration of vascular permeability reference markers in the wound exudate detection data and serum detection data, the transmembrane concentration ratio of the common inflammatory markers is permeability corrected; when the corrected concentration ratio meets the local active synthesis conditions, the contribution concentration of the common inflammatory markers in the non-infectious background is estimated, and the contribution concentration is removed from the concentration of common inflammatory markers in the serum detection data to obtain the net concentration of the common inflammatory markers contributed by the infectious factors.

[0053] When the target cluster is the infection prototype cluster, the current inflammatory state is determined to be dominated by a single bacterial infection, and the concentration of common inflammatory markers in the serum test data is directly used as the net concentration contributed by the infectious factor. When the target cluster is the endotoxin prototype cluster or the drug prototype cluster, the current inflammatory state is determined to be dominated by non-infectious factors, and the net concentration of common inflammatory markers contributed by the infectious factor is directly assigned to zero, and the determination result of non-infectious factor dominance is output.

[0054] When the target cluster is a mixed cluster, the patient's capillary permeability undergoes drastic and unstable changes in a complex mixed inflammatory state, leading to distortion of passive diffusion of macromolecules between the interstitial space and the blood. In this case, simply relying on the absolute concentration ratio of common inflammatory markers (such as IL-6) in exudate and serum cannot distinguish whether the factor is due to passive leakage caused by barrier disruption or active, excessive synthesis in local tissues. By introducing a vascular permeability reference marker (such as albumin, which is mainly present in the blood and enters the exudate purely through passive physical leakage) for standardization correction, interference from permeability changes can be eliminated, thereby accurately determining whether there is excessive active synthesis and release of common inflammatory markers (such as IL-6) at the wound site.

[0055] When the corrected concentration ratio meets the condition for local active synthesis, it indicates the presence of excessive synthesis of inflammatory factors at the wound site. This local active synthesis may originate from genuine bacterial proliferative infection or from therapeutic inflammation induced by topical drugs, and the serum itself also contains a systemic stimulating background from endotoxins. By estimating the contribution of non-infectious factors (drug-induced inflammation and enterogenic endotoxin stimulation) and removing them from the total measured concentrations of common inflammatory markers, the concentration of inflammatory factors directly driven by bacterial proliferative infection can be obtained, achieving quantitative decoupling of the true infectious component in the mixed phenotype.

[0056] Furthermore, such as Figure 4As shown, step S5 involves permeability correction of the transmembrane concentration ratio of common inflammatory markers, including: using interleukin-6 as the common inflammatory marker and albumin as the vascular permeability reference marker; calculating the concentration ratio of the common inflammatory markers in the wound exudate detection data and the serum detection data to obtain the transmembrane concentration ratio; calculating the concentration ratio of albumin in the wound exudate detection data and the serum detection data to obtain the albumin leakage index used to characterize capillary permeability; and using the albumin leakage index to correct the transmembrane concentration ratio to obtain the corrected concentration ratio.

[0057] Inflammatory stimuli from different sources can cause a systematic shift in the concentration gradient of common inflammatory markers between the local microenvironment (wound exudate) and systemic circulation (serum). The concentration of these common inflammatory markers in serum depends on the rate at which they enter the bloodstream via capillaries after local tissue synthesis, and the dynamic balance between systemic dilution and clearance rates. By quantifying the degree of concentration shift during transmembrane diffusion, the primary source of release of common inflammatory markers can be traced. Specifically, if the common inflammatory marker mainly originates from local tissue activation caused by wound irritation (such as infection or drug stimulation), its concentration in wound exudate will be significantly higher than its serum concentration, forming a positive concentration gradient. Conversely, if it mainly originates from systemic immune activation (such as endotoxin entry into the bloodstream), or severe impairment of local vascular barrier function leading to passive osmotic exchange, the serum concentration will approach or even exceed the exudate concentration, forming no gradient or a negative concentration gradient.

[0058] Based on this, when the target cluster is a mixed cluster, the ratio of the concentration of the common inflammatory marker IL-6 in the wound exudate detection data to the concentration of the common inflammatory marker IL-6 in the serum detection data is calculated to obtain the transmembrane concentration ratio. When the transmembrane concentration ratio is greater than 1, it indicates that the common inflammatory markers are mainly synthesized by local tissues; when the transmembrane concentration ratio is less than or equal to 1, it indicates that the common inflammatory markers are mainly synthesized by the whole body, or indicates that the local capillary barrier function is severely impaired.

[0059] The above concentration ratio analysis is subject to physical interference from changes in capillary permeability. The process by which molecules enter the bloodstream from the interstitial space is influenced by molecular size, the degree of tissue edema, and capillary permeability. In an infected state, bacterial toxins and inflammatory mediators disrupt the tight junctions of capillary endothelial cells, leading to a significant increase in vascular permeability. In cases of simple endotoxin stimulation or drug-induced inflammation, although permeability increases, the degree is usually lower than in viable bacterial infections. Higher permeability means less resistance for large molecules to enter the bloodstream, directly altering the expected concentration gradient; therefore, quantitative assessment and correction of individual real-time vascular permeability are essential.

[0060] Because albumin is present in extremely low concentrations in the interstitial space under normal circumstances, it only passively leaks from blood vessels into the interstitial space when vascular permeability increases, leading to an increase in albumin concentration in the exudate. Therefore, the albumin leakage index is obtained by calculating the ratio of albumin concentration in wound exudate to that in serum. This index eliminates interference from active secretion and directly and quantitatively reflects the degree of physical permeability of capillaries under the current state.

[0061] Using the albumin leakage index The transmembrane concentration ratio of common inflammatory markers Perform division standardization correction to obtain the corrected concentration ratio. A preset transparency threshold (e.g., set to 1.5) is set when... When the permeability exceeds the preset first threshold, it indicates that the local enrichment of common inflammatory markers is much higher than that of passively leaked albumin, thus satisfying the conditions for local active synthesis. This synthesis can be caused by bacterial infection or drug-induced therapeutic inflammation. When the concentration is less than or equal to the preset first permeability threshold, it indicates that the increase in common inflammatory markers is mainly due to systemic synthesis (such as an immune response triggered by endotoxin entry into the bloodstream), or that the local and systemic concentrations have reached physical equilibrium due to extremely high vascular permeability. In this case, the possibility of infection as the primary factor is low, and non-infectious factors such as systemic stimulation are mainly considered. Therefore, the condition for local active synthesis is not met. It should be understood that the calculation of the transmembrane concentration ratio R and the albumin leakage index... In order to avoid abnormal calculations when the denominator is 0, the aforementioned data preprocessing rules will automatically add a preset non-zero minimum threshold to the corresponding denominator.

[0062] If the target cluster is a mixed cluster and it is determined that it does not meet the local active synthesis condition, the net concentration of the common inflammatory marker contributed by the infectious factor is directly assigned to zero. Conversely, if the target cluster is a mixed cluster and it is determined that it meets the local active synthesis condition, it is necessary to further estimate the contribution concentration of the common inflammatory marker in the non-infectious background, and remove the contribution concentration from the concentration of the common inflammatory marker in the serum detection data to obtain the concentration of the common inflammatory marker of the source of infection used to characterize the infection status.

[0063] Furthermore, such as Figure 5As shown, step S5, estimating the contribution concentration of common inflammatory markers in a non-infectious background, includes: multiplying the concentration of the target inflammatory marker in the wound exudate detection data by the albumin leakage index to obtain the local drug signal intensity; and weighting and fusing the local drug signal intensity and the concentration of C-reactive protein in the serum detection data to obtain the contribution concentration of common inflammatory markers in a non-infectious background.

[0064] In addition to causing an increase in common inflammatory markers, drug-induced inflammation can also specifically trigger a large release of interleukin-1β, which is related to tissue repair / damage. Almost all of this release originates from local synthesis. Therefore, the concentration of IL-1β in exudate is selected as a quantitative indicator of drug effect. Endotoxin stimulation is characterized by systemic immune activation, which is mainly reflected in a significant increase in serum C-reactive protein (CRP) levels. Therefore, serum CRP concentration is selected as an indicator representing the background of systemic inflammation.

[0065] Based on the above characteristics, the contribution concentration of the common inflammatory marker (IL-6) in a non-infectious background can be estimated using the following formula:

[0066] in, The contributing concentration of the common inflammatory marker (IL-6) in a non-infectious background; The concentration of the target inflammatory marker (i.e., interleukin-1β) in the wound exudate detection data; The permeability-corrected local drug signal intensity is used to quantify the amount of IL-6 entering the bloodstream after drug-induced inflammation. The concentration of C-reactive protein in serum test data is used to quantify the serum IL-6 concentration contributed by systemic factors such as endotoxin stimulation. , These are the preset weighting coefficients.

[0067] The above weighting coefficients , All samples were pre-obtained through linear regression fitting of a historically representative sample set selected by experts. Specifically, It is the corrected signal intensity of all typical samples in the drug prototype cluster of the historical typical sample set. The results were obtained by performing a linear regression through the origin with IL-6 concentration in serum test data as the dependent variable and IL-6 concentration as the independent variable. It was obtained by performing a linear regression through the origin using serum CRP concentration as the independent variable and serum IL-6 concentration as the dependent variable in all typical samples of the endotoxin prototype cluster in the historical typical sample set.

[0068] After removing the contributing concentrations from the concentrations of the common inflammatory marker (IL-6) in the serum test data, the net concentration of the common inflammatory marker (IL-6) contributed by the infectious agent is obtained:

[0069] in: The net concentration of the common inflammatory marker (IL-6) contributed by infectious agents; The concentration of the common inflammatory marker (IL-6) in serum test data; The contributing concentration of the common inflammatory marker (IL-6) in the non-infectious background.

[0070] Furthermore, the method also includes: calculating the ratio of the net concentration of the shared inflammatory marker contributed by the infectious factor to the concentration of the shared inflammatory marker in the serum detection data to obtain the infection contribution ratio; when the infection contribution ratio is greater than or equal to a preset contribution ratio, outputting the infection-dominant determination result; otherwise, outputting the systemic synthesis-dominant determination result.

[0071] Calculate the net concentration of the common inflammatory marker (IL-6) contributed by the infectious agent. Concentration of the inflammatory marker (IL-6) shared with serum test data The ratio of the two values ​​is used to obtain the infection contribution ratio. When the infection contribution ratio When the infection contribution rate is greater than or equal to a preset contribution ratio (e.g., 50%), the inflammation status of the current target patient is determined to be infection-dominated, and the infection-dominated determination result and the calculated net concentration value are output; otherwise, when the infection contribution rate is less than or equal to a preset contribution ratio, the infection-dominated status of the target patient is determined to be infection-dominated. When the percentage is less than 50%, the current target patient's inflammatory state is determined to be systemic synthetic-dominated, meaning that non-infectious factors dominate or the patient is still in a mixed inflammatory state, and a clinically required continuous observation message is generated.

[0072] Based on the same inventive concept, embodiments of the present invention also provide a serum inflammatory factor detection system for analyzing postoperative infection status in perianal abscesses, such as... Figure 6 As shown, the system includes a data acquisition module 1, a metabolic consumption analysis module 2, an immune analysis module 3, a clustering analysis module 4, and a net concentration identification module 5, wherein: Data acquisition module 1 is used to acquire serum test data and wound exudate test data of the target patient at the current test time point after surgery; The metabolic consumption analysis module 2 is used to calculate the metabolic consumption index of the target patient based on the numerical difference between the energy metabolism characteristic parameters in the wound exudate detection data and the serum detection data. Immunoanalysis module 3 is used to calculate the comprehensive immune score based on the concentrations of each inflammatory marker and common inflammatory marker in the wound exudate detection data; The cluster affiliation analysis module 4 is used to map the metabolic consumption index and the comprehensive immune score to a pre-constructed two-dimensional coordinate system, calculate the distance between the current mapping point and each preset cluster, and determine the cluster with the smallest distance to the current mapping point as the target affiliation cluster; wherein, the clusters include prototype clusters and hybrid clusters; The net concentration identification module 5 is used to perform permeability correction on the transmembrane concentration ratio of the common inflammatory markers based on the concentrations of vascular permeability reference markers in the wound exudate detection data and serum detection data when the target cluster belongs to the mixed cluster; when the corrected concentration ratio meets the local active synthesis conditions, it estimates the contribution concentration of the common inflammatory markers in the non-infectious background, and removes the contribution concentration from the concentration of the common inflammatory markers in the serum detection data to obtain the net concentration of the common inflammatory markers contributed by the infectious factors.

[0073] Based on the same inventive concept, embodiments of the present invention also provide a computer-readable storage medium storing computer program code, which, when executed on a computer, causes the computer to perform the various steps in the serum inflammatory factor detection method for analyzing postoperative infection status of perianal abscess.

[0074] It should be noted that the above embodiments are only used to illustrate the technical concept and technical solution of the present invention, and are not intended to limit the scope of protection of the present invention. Any modifications or partial improvements made to the technical solution of the present invention by those skilled in the art without departing from the inventive concept of the present invention, or equivalent substitutions or transformations of some technical features, shall fall within the scope of protection defined by the present invention.

Claims

1. A method for detecting serum inflammatory factors for analyzing postoperative infection status in perianal abscess surgery, characterized in that, Includes the following steps: Obtain postoperative serum and wound exudate data of the target patient at the current testing time point; The metabolic consumption index of the target patient is calculated based on the numerical differences between the energy metabolism characteristic parameters in the wound exudate detection data and the serum detection data. The comprehensive immune score is calculated based on the concentrations of each inflammatory marker and common inflammatory marker in the wound exudate detection data. The metabolic consumption index and the comprehensive immune score are mapped to a pre-constructed two-dimensional coordinate system. The distance between the current mapping point and each preset cluster is calculated, and the cluster with the smallest distance to the current mapping point is determined as the target cluster. The clusters include prototype clusters and hybrid clusters. When the target cluster is the mixed cluster, the transmembrane concentration ratio of the common inflammatory markers is permeability corrected based on the concentrations of vascular permeability reference markers in the wound exudate detection data and serum detection data. When the corrected concentration ratio meets the local active synthesis conditions, the contribution concentration of the common inflammatory markers in the non-infectious background is estimated, and the contribution concentration is removed from the concentration of common inflammatory markers in the serum detection data to obtain the net concentration of the common inflammatory markers contributed by the infectious factors.

2. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 1, characterized in that, The serum detection data includes the concentrations of glucose, lactate, albumin, C-reactive protein, and interleukin-6 in the serum; the wound exudate detection data includes the concentrations of glucose, lactate, albumin, interleukin-6, interleukin-1β, procalcitonin, and high-mobility group box B1 in the wound exudate, as well as the measured partial pressure of oxygen in the wound exudate.

3. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 2, characterized in that, Calculating the metabolic consumption index of the target patient includes: Lactic acid, glucose, and oxygen partial pressure were used as the energy metabolism characteristic parameters. The ratio of lactic acid concentration in the wound exudate detection data to the serum detection data is calculated to obtain a first ratio. The ratio of glucose concentration in the wound exudate detection data to that in the serum detection data is calculated to obtain a second ratio. The ratio of the measured oxygen partial pressure in the wound exudate detection data to the preset oxygen partial pressure in normal tissue is calculated to obtain the third ratio. The first ratio, the second ratio, and the third ratio are combined to obtain the metabolic consumption index.

4. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 2, characterized in that, Calculate the overall immunity score, including: Interleukin-1β, procalcitonin, and high-mobility group box 1 were used as the respective inflammatory markers, and interleukin-6 was used as the common inflammatory marker. The ratio of the concentration of each of the inflammatory markers to the concentration of the common inflammatory markers in the wound exudate detection data was calculated and standardized to obtain the standardized ratio. The standardized ratios are weighted and fused using the corresponding feature coefficients of the pre-defined linear discriminant analysis model to obtain the comprehensive immune score.

5. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 4, characterized in that, The process of obtaining the corresponding feature coefficients of the pre-defined linear discriminant analysis model includes: Obtain the immune activation feature vector and category label for each typical sample in a set of historical typical samples that conform to the characteristics of different inflammatory stimulus prototypes; wherein, the immune activation feature vector is composed of the standardized ratio of the concentration of each of the inflammatory concomitant markers to the concentration of the common inflammatory marker in the wound exudate detection data of the typical sample; the category label includes infection prototype, endotoxin prototype and drug prototype; Based on the immune activation feature spectral vectors of typical samples of each category label, the intra-class scatter matrix within each category and the inter-class scatter matrix between different prototype categories are calculated. Based on the intra-class scatter matrix and the inter-class scatter matrix, obtain the optimal projection vector direction that maximizes the ratio of inter-class scatter to intra-class scatter. Based on the coefficient vector corresponding to the optimal projection vector direction, the corresponding feature coefficients of the preset linear discriminant analysis model are obtained.

6. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 1, characterized in that, The process of obtaining the prototype cluster and the hybrid cluster includes: Obtain the metabolic consumption index, comprehensive immune score, and category label for each typical sample in a historical typical sample set that conforms to the characteristics of different inflammatory stimulus prototypes; wherein, the category label includes infection prototype, endotoxin prototype, and drug prototype; Obtain the metabolic consumption index and comprehensive immune score for each mixture sample in the historical mixed sample set without class labels; A two-dimensional coordinate system is constructed with metabolic consumption index and comprehensive immune score as coordinate dimensions. Based on the metabolic consumption index and comprehensive immune score of typical samples in the historical typical sample set and mixed samples in the historical mixed sample set, the typical samples and the mixed samples are mapped to the two-dimensional coordinate system to obtain the mapping points of the typical samples and the mixed samples. Calculate the center coordinates of the mapping points of typical samples of each category label in the two-dimensional coordinate system, and use them as the reference coordinates for the infection prototype, endotoxin prototype and drug prototype. Cluster the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system to obtain several clusters and their cluster centers. Calculate the Euclidean distance from the center of each cluster to each reference coordinate. Label the three different clusters that are closest to the reference coordinates of the infection prototype point, endotoxin prototype point, and drug prototype point as the infection prototype cluster, endotoxin prototype cluster, and drug prototype cluster, respectively, and label the remaining clusters as mixed clusters.

7. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 2, characterized in that, Clustering is performed on the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system, including: Determine the optimal number of clusters, wherein the optimal number of clusters is greater than the number of types of the prototype cluster; Based on the optimal number of clusters, k-means clustering is used to cluster the mapping points of all typical samples in all historical typical sample sets and all mixed samples in all historical mixed sample sets in the two-dimensional coordinate system.

8. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 2, characterized in that, Permeability correction was performed on the transmembrane concentration ratios of common inflammatory markers, including: Interleukin-6 was used as the common inflammatory marker, and albumin was used as the reference marker for vascular permeability. The concentration ratio of common inflammatory markers in the wound exudate detection data and the serum detection data is calculated to obtain the transmembrane concentration ratio; The albumin concentration ratio in the wound exudate detection data and the serum detection data is calculated to obtain the albumin leakage index, which is used to characterize capillary permeability. The albumin leakage index is used to correct the transmembrane concentration ratio to obtain the corrected concentration ratio.

9. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 8, characterized in that, Estimating the contribution concentrations of common inflammatory markers in a non-infectious background, including: The concentration of the target inflammatory marker in the wound exudate detection data is multiplied by the albumin leakage index to obtain the local drug signal intensity; The local drug signal intensity and the concentration of C-reactive protein in the serum detection data are weighted and fused to obtain the contribution concentration of common inflammatory markers in the non-infectious background.

10. The method for detecting serum inflammatory factors for analyzing postoperative infection status of perianal abscess according to claim 1, characterized in that, The method further includes: The infection contribution ratio is obtained by calculating the ratio of the net concentration of the common inflammatory marker contributed by the infectious factor to the concentration of the common inflammatory marker in the serum detection data; When the infection contribution ratio is greater than or equal to the preset contribution ratio, the infection dominance determination result is output; otherwise, the systemic synthesis dominance determination result is output.

Images