Wound infection safety assessment system based on phage cocktail

Abstract

The invention relates to the technical field of medical data processing, in particular to a bacteriophage cocktail-based wound infection safety assessment system, which comprises a wound adaptation information construction module, an immune intervention window identification module, a bacteriophage inactivation risk assessment module, a cocktail adaptation consistency analysis module and an application safety conclusion generation module. According to the method, multiple biochemical and immune related indexes in wound exudate and tissue fluid are continuously collected, a time sequence is constructed, quantitative description of dynamic changes of a wound microenvironment is carried out, so that a host immune state is subjected to sequential expression, and an immune active section is identified based on the change trend of the immune indexes; the immune enzyme and defensive factor intensity are combined for collaborative analysis, so that the inactivation risk of the bacteriophage in the immune environment is assessed, and a safety judgment basis is formed through proportional statistics, thereby improving the risk identification capability and application reliability of a treatment scheme under complex individual difference conditions.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, and in particular to a wound infection safety assessment system based on phage cocktails. Background Technology

[0002] The field of medical data processing technology involves the collection, analysis, and evaluation of medical data related to clinical, pathological, and microbiological fields. Its core aspects include medical information modeling, biomarker extraction, disease risk assessment, treatment response analysis, and personalized medical support. It often uses multi-source heterogeneous data fusion systems combined with quantitative assessment methods to systematically understand patient status and assist in decision-making for intervention plans. Among these, the traditional wound infection safety assessment system refers to a method for analyzing the potential risks of using phage cocktail preparations before clinical treatment for infections caused by multidrug-resistant Pseudomonas aeruginosa. It lacks an assessment mechanism that simultaneously considers the pathogen's drug resistance spectrum, phage-pathogen matching characteristics, and the host wound microenvironment. It usually uses in vitro pathogen susceptibility testing or single-index screening as the basis for risk assessment. The methods used include bacterial culture and phage lysis tests, single-item analysis of microecological factors, and manual experience judgment.

[0003] Current technologies for assessing wound infection risk often rely on single bacterial in vitro susceptibility tests or individual indicators of microecological factors, lacking a systematic and dynamic quantitative description of the host wound microenvironment. This approach depends on static data at a single time point, making it difficult to reflect the temporal evolution of the host's immune status and prone to errors in risk assessment due to timing biases that may overlook periods of immune activity. For example, static detection may fail to capture the peak expression of certain immune factors during transient periods, leading to an underestimation of the risk of phage inactivation. Furthermore, current technologies typically fail to assess the compatibility between the biological characteristics of phage components and the host microenvironment, resulting in the unidentified presence of incompatible components in treatment strategies, leading to significant differences in treatment efficacy or increased risk of side effects. Traditional methods also fail to integrate the synergistic relationships between multiple indicators through quantitative means, resulting in a lack of logical closure in assessment conclusions and hindering the safety and accuracy of personalized medical decisions. Methods based on human experience are more susceptible to subjective interference, lack stable reproducibility, and suffer from high error rates and low universality in actual clinical practice. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a wound infection safety assessment system based on phage cocktails. The technical solution is as follows:

[0005] On the one hand, a wound infection safety assessment system based on phage cocktails is provided, which includes: The wound adaptation information construction module is used to receive exudate and tissue fluid sample data from the infected wound area collected by the sampling device, extract neutrophil count, lactate dehydrogenase concentration and total protein concentration data, and integrate them in chronological order to form a wound adaptation information sequence group. The immune intervention window identification module extracts the neutrophil number change sequence based on the wound adaptation information sequence group, combines the rising trend of lactate dehydrogenase concentration within the same time period, and filters the time period that meets the preset threshold condition based on the synchronicity and change amplitude of the two in the time dimension, and generates an immune active time segment identifier. Combined with the lactate dehydrogenase concentration change trend, the module performs joint screening to identify immune active time periods and generate immune active time segment identifiers. The phage inactivation risk assessment module obtains the concentration change trends of myeloperoxidase and defensin based on the immune active time segment identifier, extracts the intensity and time point, calculates the immune enzyme synergistic level index, and compares it with the phage inactivation reference standard to generate inactivation risk judgment result. The cocktail compatibility analysis module calls the inactivation risk assessment results, combines the lysis cycle information of the phage cocktail components, analyzes the relationship between the duration and the coverage of the immune active time period, judges the compatibility of the components with the wound environment, and generates a wound compatibility score list. The application safety conclusion generation module calculates the proportion of components that do not meet the adaptation conditions based on the wound adaptation score list, compares it with the preset evaluation threshold, and generates an application safety evaluation conclusion. The preset evaluation thresholds include three graded thresholds: 0.10, 0.30, and 0.50. These thresholds are set based on the statistical relationship between the proportion of incompatible components and the incidence of adverse reactions in historical clinical cases, and are used to determine the application safety level of phage cocktails.

[0006] As a further aspect of the present invention, the wound adaptation information sequence set includes continuous recordings of neutrophil count, lactate dehydrogenase concentration, and total protein concentration; the immune active time segment identifier includes the immune active start time point, immune active end time point, and duration of immune active; the inactivation risk assessment result includes the immune enzyme synergy level index value, phage inactivation reference standard control conclusion, and inactivation risk level; the wound adaptation score list includes the component adaptation score value, component coverage matching score, and component adaptation status marker; the application safety assessment conclusion includes the incompatible component proportion judgment value, threshold control judgment conclusion, and safety assessment level conclusion.

[0007] As a further aspect of the present invention, the wound adaptation information construction module includes: The exudate extraction submodule collects exudate and tissue fluid from the infected wound area, records the collection time and labels the sample number, organizes the samples in chronological order, extracts the liquid volume information and constructs a structured data list to generate a continuous liquid volume sequence. The neutrophil analysis submodule obtains samples with corresponding numbers based on the continuous liquid volume sequence, extracts neutrophil count information and lactate dehydrogenase concentration information, performs correlation analysis on the correlation parameters according to the time sequence, calculates the relationship between changes in neutrophils and lactate dehydrogenase, and generates dynamic parameters of neutrophils. The adaptation information sequence generation submodule extracts lactate dehydrogenase concentration information and total protein concentration information at corresponding time points based on the granulocyte dynamic parameters, integrates the three types of data to construct sequence data units, connects the units in chronological order to form a complete sequence, and generates a wound adaptation information sequence group.

[0008] As a further aspect of the present invention, the immune intervention window recognition module includes: The sequence extraction submodule extracts the neutrophil count value and sampling time corresponding to the time point of the wound adaptation information sequence group, constructs the count change sequence in chronological order, filters continuous fragments according to the change direction and duration, and generates neutrophil change intervals. The trend screening submodule extracts the lactate dehydrogenase concentration value at the corresponding time point based on the granulocyte change interval, determines whether the direction of change and time span of the concentration value in a continuous period meet the screening requirements, establishes a continuous segment that meets the conditions, and generates a lactate dehydrogenase trend segment. The time recognition submodule calls the neutrophil number change sequence corresponding to the lactate dehydrogenase trend segment, and determines whether the immune intervention threshold standard is met based on the synchronous change relationship between the two within the same time period. It locates the time period that continuously meets the conditions and generates an immune active time segment identifier. The immune intervention threshold standard includes: within the time period, the change in the number of neutrophils is not less than a preset neutrophil threshold, the change in lactate dehydrogenase concentration is not less than a preset lactate dehydrogenase threshold, and the two change in the same direction.

[0009] As a further aspect of the present invention, the process of determining whether the direction of change of lactate dehydrogenase concentration value and the time span within a continuous period meet the screening requirements is as follows: the direction of change is determined by the sign of the difference between the lactate dehydrogenase concentration values ​​corresponding to each two adjacent time points. If the direction of change of lactate dehydrogenase concentration value is consistent in no less than three consecutive time points, and the total duration of the continuous time period is not less than 48 hours, then the direction of change is deemed to meet the screening requirements. The process of locating the time period that continuously meets the conditions is as follows: under the premise of meeting the screening requirements of the direction of change of lactate dehydrogenase concentration and time span, extract the difference in the number of time points in the sequence of changes in the number of neutrophils within the time period, and determine whether the difference in the number of time points is not less than the set neutrophil threshold. If the conditions are met, the corresponding time period is identified as the immune active time segment. The neutrophil threshold is set based on experimental repeatability data, and is preferably not less than 0.09 × 10⁻⁶. 6 The difference between the number of samples per milliliter, or the adjustable parameters automatically set by the testing platform.

[0010] As a further aspect of the present invention, the phage inactivation risk assessment module includes: The indicator extraction submodule obtains the trends of myeloperoxidase concentration and defensin concentration within the time period based on the immune active time segment identifier, extracts the intensity of change and corresponding time value at the time point, and organizes them into an indicator data sequence in chronological order to generate an immune-related indicator sequence. The collaborative evaluation submodule calls the data of myeloperoxidase concentration and defensin concentration at the corresponding time points in the immune-related index sequence, performs parallel evaluation on the direction and magnitude of change of the two indicators, determines the degree of synergy in the time series, and generates an immune enzyme synergy level index. The risk assessment submodule extracts the entire range of values ​​based on the immune enzyme synergy level index, performs interval matching with the phage inactivation reference standard, classifies and identifies the index values ​​within the time period, determines whether they are within the inactivation assessment range, and generates an inactivation risk assessment result.

[0011] As a further aspect of the present invention, the start and end range of the time period for which the immune active time segment identifier is obtained is not less than 24 hours, and the interval between consecutive sampling frequencies is not higher than 2 hours. In the trends of myeloperoxidase concentration change and defensin concentration change, the intensity of the change is extracted based on the criterion that the concentration difference between two adjacent time points is not less than the set concentration fluctuation threshold of 0.1 μg / mL. Time point data that do not meet the condition are not included in the immune-related index sequence. The immune enzyme synergistic level index is calculated based on the concentration change trend of myeloperoxidase and defensin during the period of immune activity, taking into account the consistency of their change direction and the difference in their magnitude. The amplitude difference is calculated by normalizing the change intensity of the two indicators and then generating the collaborative level value in reverse according to the preset mapping function. The range of the synergistic level index corresponding to the inactivation determination range is from -0.3 to 0.3. Data outside the range are processed as non-inactivation markers. In the process of generating the synergistic level index of the immune enzyme, the consistency requirement for the change direction of the myeloperoxidase concentration and the defensin concentration is that the proportion of consistent direction in three or more adjacent time points is not less than 70%. In the process of generating the inactivation risk assessment result, the value range of the immune enzyme synergistic level index corresponding to the inactivation assessment range is -0.3 to 0.3, and data outside the range are treated as non-inactivation data.

[0012] As a further aspect of the present invention, the cocktail compatibility analysis module includes: The periodic extraction submodule calls the inactivation risk judgment result to obtain all component information in the phage cocktail, extracts the lysis cycle and duration data of the components, classifies and sorts them in combination with the corresponding component numbers, establishes a correspondence table between component numbers and cycle duration, and generates a component cycle mapping list. The segment matching submodule extracts the duration segments of components according to the component period mapping list, compares the overlap ratio of the time period with the immune active time segment identifier on the time axis, calculates the coverage degree and marks the difference range, and generates the time segment coverage relationship. The adaptation assessment submodule calls the time segment coverage relationship and component cycle duration data, assigns weights to the duration coverage and degree of overlap of each component with the immunization time period, calculates the matching degree index, scores all components, and generates a wound adaptation score list. The weighting is a preset ratio allocation, with the duration coverage score and the overlap score of the immunization period weighted at 65% and 35% respectively to calculate the matching degree index.

[0013] As a further aspect of the present invention, the application security conclusion generation module includes: The scoring and filtering submodule obtains the score value corresponding to the component based on the wound adaptation score list, judges the score value item by item according to the adaptation judgment criteria, marks the components that do not meet the adaptation conditions and records the component number, and sorts the filtering results in order of number to obtain the set of unsuitable component tags. The proportion statistics submodule obtains the number of marked components based on the set of mismatched component markers, and at the same time obtains the number of all components in the wound compatibility scoring list. It calculates the proportion of the two types of quantities to form proportion data reflecting the proportion of mismatched components and generates the proportion of mismatched components. The security determination submodule compares the proportion of the incompatible components with a preset evaluation threshold, classifies the application status according to the threshold range in which the proportion falls, organizes the corresponding determination tags, and generates an application security assessment conclusion.

[0014] As a further aspect of the present invention, the process of calculating the proportion is as follows: the number of components included in the set of unsuitable component markers and the total number of components in the wound compatibility scoring list are used as the numerator and denominator respectively to perform a ratio calculation, and the result of the ratio calculation is retained to two decimal places. The interval comparison process involves comparing the proportion of the unsuitable component with three thresholds of 0.10, 0.30, and 0.50 in the preset evaluation threshold set. If the proportion of the unsuitable component is less than 0.10, it is marked as "applicable". If the proportion of the incompatible component is between 0.10 and 0.30, which is greater than or equal to 0.10 and less than 0.30, it is marked as "limited application"; If the proportion of the incompatible component is between 0.30 and 0.50, which is greater than or equal to 0.30 and less than 0.50, it is marked as "use with caution"; If the proportion of the incompatible component is greater than or equal to 0.50, it is marked as "unapplicable".

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, multiple biochemical and immune-related indicators in wound exudate and tissue fluid are continuously collected and time-series data are constructed to quantitatively characterize the dynamic changes in the wound microenvironment. This allows for the temporal expression of the host's immune status, identification of immune-active regions based on the trends in immune indicator changes, and synergistic analysis of the strength of immune enzymes and defense factors. This enables precise assessment of the risk of phage inactivation in the immune environment, and the formation of a safety judgment basis through proportional statistics. This improves the risk identification capability and application reliability of treatment plans under complex individual differences. Attached Figure Description

[0016] 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.

[0017] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 This is a flowchart of the wound adaptation information construction module in this invention; Figure 4 This is a flowchart of the immune intervention window recognition module in this invention; Figure 5This is a flowchart of the phage inactivation risk assessment module in this invention; Figure 6 This is a flowchart of the cocktail adaptation consistency analysis module in this invention; Figure 7 This is a flowchart of the safety conclusion generation module used in this invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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.

[0023] This invention provides a wound infection safety assessment system based on phage cocktails, such as... Figure 1-2 The diagram shown illustrates a wound infection safety assessment system based on phage cocktails. The system includes: The wound adaptation information construction module is used to receive exudate and tissue fluid sample data from the infected wound area collected by the sampling device, obtain the exudate and tissue fluid from the infected wound area to be treated with phage cocktail, extract neutrophil count information, lactate dehydrogenase concentration information and total protein concentration information, integrate the collected data into continuous records in chronological order, and generate wound adaptation information sequence groups. Wound adaptation information sequence set refers to a multivariate time series set composed of biochemical detection data of neutrophil count, lactate dehydrogenase concentration and total protein concentration collected from the wound area to which phage cocktail therapy is to be applied. The sequence is used to describe the dynamic changes of the wound microenvironment over continuous time and is derived from standardized clinical sample collection and biochemical indicator detection procedures. The immune intervention window identification module extracts the neutrophil number change sequence based on the wound adaptation information sequence group, combines the rising trend of lactate dehydrogenase concentration within the same time period, and selects time periods that meet the preset threshold conditions based on the synchronicity and change magnitude of the two in the time dimension. It generates immune active time segment identifiers and performs joint screening in combination with the lactate dehydrogenase concentration change trend to identify the corresponding immune active time periods and generate immune active time segment identifiers. Immune active time segment identifiers refer to continuous time periods used to represent the active state of immune response in the wound microenvironment. The time periods are determined by combining the rapid changes in neutrophil count and the synchronous upward trend of lactate dehydrogenase concentration. The identifier information consists of the start and end times of the time segment and is used to determine the correlation between physiological interference behaviors. The phage inactivation risk assessment module identifies the immune active time segment, obtains the changes in myeloperoxidase concentration and defensin concentration within the time segment, extracts the intensity of the indicators and the corresponding time points, calculates the immune enzyme synergistic level indicators within the time segment, and compares them with the phage inactivation reference standard to generate the inactivation risk judgment result. The immune enzyme synergy level index refers to the level of combined action of myeloperoxidase and defensins, two immune-related enzymes, in a target time period. The index is obtained by detecting the concentration change trend and time overlap relationship, and is used to infer the intensity of the risk of inactivation of phage structure exposed in the immune active segment. Phage inactivation reference standards refer to the concentration thresholds used to determine whether there is a risk of structural inactivation of phages in the wound environment. The standards are set based on in vitro phage stability experimental data, enzyme concentration range, and functional inactivation limits reported in the literature, and are used for comparison with synergistic indicators. The cocktail compatibility analysis module calls the inactivation risk assessment results, combines the lysis cycle information of the components in the phage cocktail, compares the duration with the coverage of the immune active time period, judges the compatibility of the components with the wound environment, and generates a wound compatibility score list. The phage component wound adaptation score list is a set of data used to represent the adaptation relationship between the components in the phage cocktail and the current wound immune status. Each component corresponds to an adaptation status evaluation label, which is judged based on whether the lysis cycle avoids the immune active segment. The application safety conclusion generation module calculates the percentage of components that do not meet the adaptation criteria based on the wound adaptation score list, compares the percentage with the preset evaluation threshold, and generates an application safety evaluation conclusion based on the comparison results. The assessment threshold is a proportional standard used as a classification boundary condition when determining the safety level of phage cocktail applications. The threshold is used to determine whether the proportion of phage components that do not have compatibility in all components has reached the critical range. The source can be set based on clinical risk tolerance standards or system preset strategies. The preset assessment thresholds include three grading thresholds: 0.10, 0.30, and 0.50. These thresholds are set based on the statistical relationship between the proportion of incompatible components and the incidence of adverse reactions in historical clinical cases, and are used to determine the application safety level of phage cocktails.

[0024] The wound adaptation information sequence set includes continuous recordings of neutrophil count, lactate dehydrogenase concentration, and total protein concentration; the immune activity time segment markers include the immune activity start time point, immune activity end time point, and duration of immune activity; the inactivation risk assessment results include immune enzyme synergy index values, phage inactivation reference standard control conclusions, and inactivation risk levels; the wound adaptation score list includes component adaptation score values, component coverage matching scores, and component adaptation status markers; the application safety assessment conclusions include incompatible component proportion judgment values, threshold control judgment conclusions, and safety assessment level conclusions.

[0025] Specifically, such as Figure 2 , 3 As shown, the wound adaptation information construction module includes: The exudate extraction submodule collects exudate and tissue fluid from the infected wound area, records the collection time and labels the sample number, organizes the samples in chronological order, extracts the liquid volume information and constructs a structured data list to generate a continuous liquid volume sequence. A fluid collection path was set up in the infected wound area. The collector used a disposable sterile negative pressure drainage tube connected to a graduated collection bottle. The inlet of the collector was fixed at the lowest point of the wound, and the outlet was sealed to the collection bottle. The scale resolution on the outer wall of the collection bottle was set to 1.0 ml. At the beginning of each collection, a collection timestamp and sample number were written. The sample number used a combination of a fixed prefix and an incremental sequence number. The prefix was the wound number, and the sequence number started from 1 and increased. After collection, the difference between the scale reading and the net weight was read. The scale reading was used for quick recording, and the net weight difference was used for volume verification. The net weight difference was measured using an electronic balance with a resolution of 0.01 g. The volume was approximately converted according to the water sample density. When the deviation between the verification volume and the scale reading exceeded 2.0 ml, a retest was triggered, and both readings were retained. Each collection record was written into a structured data list. The fields were in a fixed order: sample number, collection time, fluid volume, verification volume, operator number, and anomaly marker. The data was sorted in ascending order by collection time, and a continuous fluid volume sequence was generated after sorting. To ensure sequence continuity, a constraint is applied to the time difference between two adjacent data collections. If the time difference exceeds 24.0 hours, a missing marker is inserted into the sequence and the reason for the missing record is recorded. Missing records do not participate in subsequent association operations; only the index position is retained.

[0026] Table 1 Sampling Records and Liquid Volume Sequence Table Sample number Collection time Liquid volume, milliliters Verification volume, milliliters Exception marking W1-1 2022-12-20 08:00 12.0 11.8 0 W1-2 2022-12-20 20:00 18.0 18.3 0 W1-3 2022-12-21 08:00 22.0 21.6 0 W1-4 2022-12-21 20:00 16.0 16.2 0 W1-5 2022-12-22 08:00 10.0 9.9 0 W1-6 2022-12-22 20:00 8.0 8.1 0 As shown in Table 1, the collection time and liquid volume are organized in chronological order to form a continuous liquid volume sequence, and the corresponding sample is retrieved according to the sample number.

[0027] The neutrophil analysis submodule acquires samples with corresponding numbers based on continuous liquid volume sequences, extracts neutrophil count information and lactate dehydrogenase concentration information, performs correlation analysis on the associated parameters according to the time sequence, calculates the relationship between changes in neutrophils and lactate dehydrogenase, and generates dynamic parameters of neutrophils. Centrifuge tubes were retrieved from the sample library based on the sample number in the continuous liquid volume sequence. Samples were centrifuged within 30 minutes of collection, with centrifugation conditions set at 3000 rpm for 10 minutes. The supernatant was used for lactate dehydrogenase concentration detection, and the precipitated cells were used for neutrophil counting. Neutrophil counting was performed using a flow cytometry scheme. The input data was a cell suspension event stream. The event stream was first gated by forward and side scattering to remove debris, and then gated by fluorescent antibody labeling to identify the neutrophil population. The output was the number of cells per milliliter of cell suspension, which was then converted to the total neutrophil count for the sample, using the corresponding total cell suspension volume. Lactate dehydrogenase (LDH) concentration was measured using the standard procedure for body fluid LDH detection, with output units in International Units (IU). Body fluid LDH is a routine body fluid test for assessing fluid-related damage. The numerical range can be set with reference to the interpretation of body fluid tests and the criteria for identifying exudates. In exudate identification, body fluid LDH is often correlated with the upper limit range of serum levels. For each pair of adjacent time points, a correlation operation is performed: first, the difference in neutrophil count between two adjacent time points is taken, then the difference in LDH concentration between two adjacent time points is taken. These two types of differences are then paired according to the same time period to form a change relationship entry. A direction marker field is added to the change relationship entry. The direction marker rule is: a positive difference indicates an increase, a negative difference indicates a decrease, and an absolute difference less than the repeatability threshold of the indicator is considered stable. The repeatability threshold was obtained through repeated measures experiments. The experiment involved testing the same mixed sample 10 times consecutively, and calculating the difference between the maximum and minimum values ​​of the 10 results. The neutrophil count threshold was set to 0.30 times this difference, and the lactate dehydrogenase concentration threshold was also set to 0.30 times this difference. In the example, the maximum value of the repeated measurements of neutrophil count was... The minimum value is Its range is:

[0028] ; The neutrophil threshold is set at 30% of the range, i.e.: ; In the example, the maximum value of neutrophil count measured 10 times was 4.90 x 10^6, the minimum value was 4.60 x 10^6, and the difference was 0.30 x 10^6. Therefore, the threshold was set to 0.09 x 10^6. The maximum value of lactate dehydrogenase concentration measured 10 times was 520.0 IU / L, the minimum value was 500.0 IU / L, and the difference was 20.0 IU / L. Therefore, the threshold was set to 6.0 IU / L. The direction marker and change intensity of each time period were written into the neutrophil dynamic parameter list. The change intensity was recorded according to the absolute difference and also written into the time span field. The time span was obtained by the difference between adjacent collection times, and in the example, it was 12.0 hours.

[0029] The adaptation information sequence generation submodule extracts lactate dehydrogenase concentration information and total protein concentration information at corresponding time points based on granulocyte dynamic parameters, integrates the three types of data to construct sequence data units, connects the units in chronological order to form a complete sequence, and generates a wound adaptation information sequence group. Based on the dynamic parameters of granulocytes, time points were located one by one, and the lactate dehydrogenase concentration and total protein concentration information at the same time point were retrieved. The total protein concentration was detected using the biuret method, and the output unit was grams per liter. The total protein concentration of wound exudate could be set with reference to the clinical exudate protein concentration level. There are reports in the literature that 2.9 grams per deciliter is used as the exudate protein concentration for loss estimation, which can be converted to 29.0 grams per liter as a reference level. The fluid volume, lactate dehydrogenase concentration, and neutrophil count were simultaneously aligned to the same time index, and the total protein concentration was added to form a sequence data unit. The unit field order was fixed as sample number, collection time, fluid volume, lactate dehydrogenase concentration, total protein concentration, neutrophil count, time span, and orientation marker. The units were connected in chronological order to form a complete sequence, and a wound adaptation information sequence group was generated. In the example, at time point W1-1, the liquid volume was 12.0 mL, the lactate dehydrogenase concentration was 480.0 U / L, the total protein concentration was 28.0 g / L, and the neutrophil count was 4.70 × 10⁻⁶. 6 The volume of the fluid at time points W1-2 was 18.0 mL, the concentration of lactate dehydrogenase was 520.0 U / L, the concentration of total protein was 30.0 g / L, and the number of neutrophils was 5.10 × 10⁻⁶. 6 In the example, at time point W1-1, the liquid volume was 12.0 mL, the lactate dehydrogenase concentration was 480.0 IU / L, the total protein concentration was 28.0 g / L, and the neutrophil count was 4.70 x 10⁶. At time point W1-2, the liquid volume was 18.0 mL, the lactate dehydrogenase concentration was 520.0 IU / L, the total protein concentration was 30.0 g / L, and the neutrophil count was 5.10 x 10⁶. This sequence set was used as the sole input sequence in subsequent modules, and all time point localization was based on the acquisition time index of this sequence.

[0030] Specifically, such as Figure 2 , 4 As shown, the immune intervention window recognition module includes: The sequence extraction submodule extracts the neutrophil count value and sampling time corresponding to the time point of the wound adaptation information sequence group, constructs the number change sequence in chronological order, filters continuous fragments according to the change direction and duration, and generates neutrophil change intervals. Neutrophil counts and sampling times are read from the wound adaptation information sequence set. First, an integrity check is performed, removing records with missing time markers and records marked as 1 (abnormal). The temporal index continuity of the records is preserved, and a neutrophil count change sequence is constructed chronologically. The neutrophil count change sequence uses the difference in neutrophil counts between adjacent time points as the core entry, and the direction and duration of change are written into each entry. The duration is obtained by subtracting adjacent sampling times; in the example, it is 12.0 hours. Continuous segments are filtered based on the direction and duration of change. The filtering rules use two conditions: consistent direction and minimum duration. Consistent direction requires three consecutive entries to have the same direction marker, and the minimum duration requires the segment to cover a time span of at least 24.0 hours. The threshold for this rule was determined through labeled data experiments. The experimental data came from 30 cases of infected wounds, with each case sampled at least 6 times consecutively. The period of active immune response was manually labeled as a control label. The minimum duration threshold was traversed through the set of values ​​from 12.0 hours to 48.0 hours, and the threshold that resulted in the highest consistency evaluation index was selected. In the example, 24.0 hours corresponds to a consistency evaluation index of 0.86, which is higher than 0.73 for 12.0 hours and 0.79 for 48.0 hours. Therefore, 24.0 hours was fixed. The granulocyte change interval was output, and the interval start time, interval end time, the first and last values ​​of the number of neutrophils in the interval, and the direction label were recorded.

[0031] The trend screening submodule extracts the lactate dehydrogenase concentration value at the corresponding time point based on the granulocyte change interval, determines whether the direction of change and time span of the concentration value in a continuous period meet the screening requirements, establishes a continuous segment that meets the conditions, and generates a lactate dehydrogenase trend segment. Lactate dehydrogenase (LDH) concentration values ​​were retrieved at corresponding time points within each granulocyte variation interval to construct a LDH concentration variation sequence. The direction and time span of these changes were then screened. Screening requirements included directional consistency and a minimum variation amplitude. Directional consistency required that all adjacent differences within the interval had the same direction. The minimum variation amplitude threshold was determined through a repeatability experiment of body fluid LDH detection. The experiment was repeated 10 times on the same sample using the same method, yielding a threshold of 6.0 IU / L. A safety margin was further introduced, setting the threshold to 12.0 IU / L. The threshold setting process was validated on 30 labeled data points. The false alarm rates were calculated for thresholds of 6.0, 12.0, and 18.0 IU / L, with example false alarm rates of 0.22, 0.11, and 0.08, and false negative rates of 0.09, 0.12, and 0.20, respectively. The average false alarm and false negative rates for a threshold of 12.0 IU / L were 0.115, lower than the 0.155 for a threshold of 6.0 and 0.14 for a threshold of 18.0. Therefore, 12.0 IU / L was fixed. Continuous segments meeting the screening requirements were written into the lactate dehydrogenase trend segment. The segment fields included start time, end time, first and last lactate dehydrogenase concentration values, total change range, and direction marker. In the example, the granulocyte change range covered the period from 08:00 on December 20, 2022 to 08:00 on December 21, 2022, with the neutrophil count increasing from 4.70 × 10⁻⁶. 6 The number rose to 5.80×10 6 The lactate dehydrogenase concentration increased from 480.0 U / L to 610.0 U / L, with a total change of 130.0 U / L, satisfying the minimum change range of 12.0 U / L and the direction of change being consistent. The number of neutrophils increased from 4.70 x 10^6 to 5.80 x 10^6, and the lactate dehydrogenase concentration increased from 480.0 IU / L to 610.0 IU / L, with a total change of 130.0 IU / L, satisfying the minimum change range of 12.0 IU / L and the direction of change being consistent.

[0032] The time recognition submodule calls the neutrophil number change sequence corresponding to the lactate dehydrogenase trend segment, and determines whether the immune intervention threshold standard is met based on the synchronous change relationship between the two within the same time period. It then locates the time period that continuously meets the conditions and generates an immune active time segment identifier. The threshold criteria for immune intervention include: within the time period, the change in the number of neutrophils is not less than a preset neutrophil threshold, the change in lactate dehydrogenase concentration is not less than a preset lactate dehydrogenase threshold, and the two changes are in the same direction. For each lactate dehydrogenase trend segment, neutrophil count change sequences within the same time period were retrieved, and a synchronous change relationship determination was performed. The determination used two conditions: first, the two sequences were aligned in direction; second, the change intensity of both sequences exceeded their respective thresholds. The neutrophil count change intensity threshold was determined through neutrophil count repeatability experiments, with an example of 0.09 x 10^6, and a safety margin was introduced by setting the threshold to 0.18 x 10^6. When the synchronous change relationship is met, the immune intervention threshold standard is further implemented. The threshold standard adopts a combination of intervals, rather than using a single threshold. Interval 1 is defined as a difference of 0.18 x 10^6 to 0.60 x 10^6 between the first and last neutrophils and a total change in lactate dehydrogenase between 12.0 IU / L and 80.0 IU / L. Interval 2 is defined as a difference of more than 0.60 x 10^6 between the first and last neutrophils and a total change in lactate dehydrogenase exceeding 80.0 IU / L. The interval boundaries are derived from the quantile statistics of 30 labeled data. The 25th and 75th percentiles of the neutrophil count difference in the immune-active labeled samples are taken as the boundaries of interval 1, and the values ​​above the 75th percentile are taken as the boundaries of interval 2. The change in lactate dehydrogenase is determined similarly. Time periods that continuously meet the conditions are identified and marked as immune-active time intervals. In the example, the difference in neutrophil count between 08:00 on December 20, 2022 and 08:00 on December 21, 2022 was 1.10 x 10^6, and the total change in lactate dehydrogenase was 130.0 IU / L, falling into interval 2. Therefore, an immune-active time segment identifier was generated, and this segment was used as input for subsequent risk assessment. The advantage of this operational logic is that by simultaneously constraining the changes in neutrophil count and lactate dehydrogenase, the false alarm rate of the time segment identifier in the validation of labeled data decreased from 0.22 when using only a single indicator for screening to 0.11, a reduction of 50.0%.

[0033] Specifically, such as Figure 2 , 5 As shown, the phage inactivation risk assessment module includes: The indicator extraction submodule obtains the trends of myeloperoxidase concentration and defensin concentration within the time period based on the immune active time segment identifier, extracts the intensity of change and corresponding time value at the time point, and organizes them into an indicator data sequence in chronological order to generate an immune-related indicator sequence. Based on the immune-active time segment markers, the myeloperoxidase (MOP) and defensin concentration records for that time period were retrieved. MOP concentration was output using an immunoturbidimetric or enzyme-linked immunosorbent assay (ELISA) procedure. The assay procedure needed to cover a quantitative range down to nanograms per milliliter and have an upper limit of 200.0 nanograms per milliliter. The quantitative range could be set with reference to the upper and lower limits of quantitation for the MOP assay method. Defensin concentration was output using a defensin ELISA procedure. The input sample was wound supernatant. The sample underwent pretreatment before testing, including centrifugation to remove residue, filtration to remove particulates, and setting the dilution factor according to the kit requirements. The dilution factor was determined in a preliminary experiment. The preliminary experiment involved diluting samples at the same time point by 5, 10, and 20 times and testing them. The dilution factor falling in the middle of the quantitative curve was selected as the fixed factor. In the example, the absorbance of the 10-fold dilution fell in the middle of the curve, therefore a 10-fold dilution was used. Defensives can be detected using either alpha or beta defensin detection channels. The detection platform needs to output nanograms per milliliter and provide stable positive and negative control lines. For each time point, the intensity of change in two indicators and their corresponding time values ​​are extracted. The intensity of change is taken as the absolute value of the difference between adjacent time points. This data is then organized chronologically into an indicator data sequence with fields including time, myeloperoxidase concentration, myeloperoxidase intensity of change, defensin concentration, and defensin intensity of change. To ensure reproducibility of intensity of change, the original concentrations are denoised using a moving window median strategy, covering three time points. The output is the median within the window. If the difference before and after denoising exceeds 20.0% of the original value, the original value is retained and a fluctuation marker is set.

[0034] Table 2. Immune-related index sequence data. time Myeloperoxidase concentration, nanograms per milliliter defensin concentration, nanograms per milliliter 2022-12-20 08:00 45.0 120.0 2022-12-20 20:00 70.0 150.0 2022-12-21 08:00 95.0 210.0 As shown in Table 2, two indicator sequences are formed in time order during the period of active immunity, and subsequent collaborative evaluation directly calls the data of the same time point.

[0035] The collaborative assessment submodule calls the data of myeloperoxidase concentration and defensin concentration at corresponding time points in the immune-related indicator sequence, performs parallel assessment of the direction and magnitude of change of the two indicators, determines the degree of synergy in the time series, and generates an immune enzyme synergy level index. Parallel evaluation was performed on the immune-related indicator sequences. First, the directional markers of the two indicators within each adjacent time period were calculated, following the same rules as described above. Then, an amplitude consistency score was calculated, obtained by the ratio of the intensity changes of the two indicators. Specifically, the intensity changes of myeloperoxidase and defensin were taken, and each was divided by its respective upper limit of measurement range to obtain a normalized intensity change. The upper limit of measurement range was the upper limit of detection quantitation, with myeloperoxidase set at 200.0 ng / mL and defensin at 500.0 ng / mL. After normalization, the absolute value of the difference between the two was calculated. The smaller the absolute value of the difference, the higher the synergy. Furthermore, directional consistency and amplitude consistency were combined to form an immune enzyme synergy level index, with the index ranging from 0 to 1. When the directions were inconsistent, it was directly recorded as 0. When the directions were consistent, the absolute value of the amplitude difference was mapped to the synergy level. The mapping rules were obtained through experimental calibration. In the experiment, 20 groups of myeloperoxidase and defensin concentration gradients were constructed in an in vitro mixed system, and the phage titer decay was recorded as a control. The set of mapping thresholds that maximized the distinguishability of titer decay groups was selected. The example mapping rule is that the absolute value of the amplitude difference is not more than 0.05 and is recorded as 0.90, not more than 0.10 and is recorded as 0.75, not more than 0.20 and is recorded as 0.60, and more than 0.20 and is recorded as 0.40. In the example, from 08:00 to 20:00 on December 20, 2022, the intensity of myeloperoxidase change was 25.0 ng / mL, and the intensity of defensin change was 30.0 ng / mL. After normalization, these values ​​were 0.125 and 0.060, respectively, with an absolute difference of 0.065, falling within the range of no more than 0.10. The synergy level was recorded as 0.75. From 20:00 on December 20, 2022 to 08:00 on December 21, 2022, the intensity of myeloperoxidase change was 25.0 ng / mL, and the intensity of defensin change was 60.0 ng / mL. After normalization, these values ​​were 0.125 and 0.120, respectively, with an absolute difference of 0.005. The synergy level was recorded as 0.90. The synergy levels of the two time periods were averaged to obtain the immune enzyme synergy level index for this period of immune activity, with an average result of 0.825.

[0036] The risk assessment submodule extracts the entire range of values ​​based on the immune enzyme synergy level index, performs interval matching with the phage inactivation reference standard, classifies and identifies the index values ​​within the time period, determines whether they are within the inactivation assessment range, and generates an inactivation risk assessment result. Read the full range of values ​​for the immune enzyme synergy level index and match the intervals with the phage inactivation reference standard. The inactivation reference standard was established through an in vitro phage inactivation experiment. The experiment included three combinations of myeloperoxidase concentration and three combinations of defensin concentration, forming nine conditions. An equal volume of phage cocktail was added to each condition and incubated at 37.0°C for 60.0 minutes. Phage titers were measured before and after incubation. A titer decrease exceeding 90.0% was considered inactivation; a decrease between 50.0% and 90.0% was considered high-risk; and a decrease below 50.0% was considered low-risk. A correspondence table was established between the synergy level index and the above labeling results. In the example, a synergy level of ≥0.85 corresponded to an inactivation rate of 77.8%; a synergy level between 0.70 and 0.85 corresponded to a high-risk rate of 66.7%; and a synergy level below 0.70 corresponded to a low-risk rate of 72.2%. Based on this, an interval matching rule was set: a synergy level of ≥0.85 was considered inactivation, 0.70 to 0.85 was considered high-risk, and below 0.70 was considered low-risk. In the example, the synergy level is 0.825, falling within the 0.70 to 0.85 range. The output inactivation risk assessment result is marked as high risk, and this marker, along with the time period index, is passed to the subsequent consistency analysis module. The advantage of this operational logic is that by incorporating the difference in normalized change intensity between the two immune-related indicators into the synergy assessment, the accuracy of inactivation group identification in in vitro validation is increased from 68.0% (based solely on single indicator concentration thresholds) to 82.0%, an improvement of 20.6%.

[0037] Specifically, such as Figure 2 , 6 As shown, the cocktail compatibility consistency analysis module includes: The periodic extraction submodule calls the inactivation risk assessment result to obtain all component information in the phage cocktail, extracts the lysis period and duration data of the components, classifies and sorts them according to the corresponding component numbers, establishes a correspondence table between component numbers and period duration, and generates a component periodic mapping list. All component information of the phage cocktail is retrieved from the inactivation risk assessment results. Component information fields include component number, phage name, target strain label, lysis period, duration, formulation concentration, and dosing start time. Lysis period and duration data are obtained from component factory quality control and in vitro lysis curve experiments. The lysis curve experiments were conducted on standard target strains. The lysis period is the time point at which stable lysis first appears, and the duration is the total duration of the lysis signs. A correspondence table is established between component numbers and cycle durations to generate a component cycle mapping list, sorted in ascending order by component number. The unit for lysis period and duration is uniformly hours. The lysis period typically falls between 0.5 hours and 6.0 hours in in vitro lysis experiments, and the duration typically falls between 6.0 hours and 72.0 hours. Records exceeding these ranges are marked with quality control flags and require review when writing to the mapping list.

[0038] The segment matching submodule extracts the duration segments of components based on the component period mapping list, compares the overlap ratio of the time period with the immune active time segment identifier on the time axis, calculates the coverage degree and marks the difference range, and generates the time segment coverage relationship. Each component duration segment in the component cycle mapping list is read and aligned with the immune-active time segment identifier on the same time axis. The alignment operation uses the drug administration start time as the starting point of the component duration segment, adds the duration to the starting point to obtain the ending point, and calculates the overlap ratio. The overlap ratio is obtained by dividing the overlap duration by the duration of the immune-active time period. The overlap duration is obtained by subtracting the intersection start point and intersection end point of the two time periods. The intersection start point is the later of the two start points, and the intersection end point is the earlier of the two end points. If the intersection end point is earlier than the intersection start point, the overlap duration is recorded as 0. In the example, the immune-active time period is from 08:00 on December 20, 2022 to 08:00 on December 21, 2022, with a duration of 24.0 hours. The administration of component P1 began at 06:00 on December 20, 2022, and lasted for 36.0 hours. Therefore, the duration of component P1 was from 06:00 on December 20, 2022 to 18:00 on December 21, 2022. The overlap with the period of active immunity was from 08:00 on December 20, 2022 to 08:00 on December 21, 2022, with an overlap of 24.0 hours and an overlap ratio of 1.00. The administration of component P2 started at 18:00 on December 20, 2022, and lasted for 24.0 hours. The intersection was from 18:00 on December 20, 2022 to 08:00 on December 21, 2022, with an overlap duration of 14.0 hours and an overlap ratio of 0.58. The overlap ratio and difference range of each component were written into the time segment coverage relationship, and the difference range was recorded as the start and end segments not covered during the period of active immunity.

[0039] The adaptation assessment submodule calls the time segment coverage relationship and component cycle duration data, assigns weights to the duration coverage and degree of overlap of each component with the immunization time period, calculates the matching degree index, scores all components, and generates a wound adaptation score list. The weighting is a preset ratio allocation, with the duration coverage score and the immunization time period overlap score weighted at 65% and 35% respectively to calculate the matching degree index. The system reads the time segment coverage relationship and component cycle duration data, calculates the matching degree index for each component, and generates a wound adaptation score list. The matching degree index is a weighted summary of two parts: the first part is the duration coverage score, and the second part is the lysis cycle matching score. The duration coverage score is directly mapped to a score from 0 to 100 based on the overlap ratio. The mapping rule is as follows: an overlap ratio of not less than 0.80 is recorded as a score of 90 to 100; an overlap ratio between 0.50 and 0.80 is recorded as a score of 70 to 89; and an overlap ratio below 0.50 is recorded as a score of 0 to 69. A linear piecewise interpolation table is used within each interval. The interpolation table is fixed as discrete mapping points in the system configuration file and obtained by nearest neighbor lookup, avoiding the use of mathematical formulas at runtime. The lysis cycle matching score is obtained by matching the lysis cycle with the rate of neutrophil change during the immune-active period. The rate of neutrophil change is obtained by dividing the difference between the first and last neutrophil counts within the output segment of the immune intervention window recognition module by the duration of the time period. For example, the first and last difference is 1.10 x 10^6, the duration is 24.0 hours, and the rate of change is 0.0458 x 10^6 per hour. After the rate of change falls into the preset rate range, the corresponding lysis cycle target range is selected. The rate range is calibrated by comparing treatment records and bacterial load changes in historical cases. For example, a rate of not less than 0.040 x 10^6 per hour corresponds to a lysis cycle target range of 0.5 hours to 2.0 hours. A lysis cycle falling into the target range is scored out of 100 points. When deviating from the target range, points are deducted according to the degree of deviation. The deduction rules are implemented through a discrete deduction table. For example, a deviation of no more than 1.0 hour deducts 10 points, a deviation of no more than 2.0 hours deducts 25 points, and a deviation of more than 2.0 hours deducts 50 points. The weight settings were determined through experimental control. Combinations of duration coverage weights ranging from 0.40 to 0.80 and fragmentation cycle matching weights ranging from 0.20 to 0.60 were used, with the sum of the weights fixed at 1.00. The consistency between the scores and the human adaptation conclusions was calculated on 30 labeled data points, and the weight combination with the highest consistency was selected. In the example, when the duration coverage weight was 0.65 and the fragmentation cycle matching weight was 0.35, the consistency was 0.84, higher than the consistency of 0.78 obtained with the 0.50 and 0.50 combination, and the consistency of 0.80 obtained with the 0.75 and 0.25 combination. Therefore, the weight configuration of 0.65 and 0.35 was consistently used, as it was higher than the 0.78 of the 0.50 and 0.50 combination and the 0.80 of the 0.75 and 0.25 combination.In the example, P1 has an overlap ratio of 1.00, corresponding to a duration coverage score of 100 points. Its lysis period is 1.5 hours, falling within the target range of 0.5 to 2.0 hours, resulting in a lysis period matching score of 100 points. The weighted summation yields a matching degree index of 100 points. P2 has an overlap ratio of 0.58, corresponding to a duration coverage score of 78 points (obtained from a table). Its lysis period is 3.0 hours, deviating from the target range by 1.0 hour, resulting in a lysis period matching score of 90 points. The weighted summation yields a matching degree index of 82.2 points. A wound adaptation score list is output for all components and passed to the application safety conclusion generation module. The advantage of this calculation logic is that by simultaneously incorporating the overlap ratio of the immune active time period and the target lysis period range into the score, the consistency between the control experiment and the artificial adaptation conclusion is improved from 0.76 (based solely on the overlap ratio score) to 0.84, an increase of 10.5%.

[0040] Specifically, such as Figure 2 , 7 As shown, the application security conclusion generation module includes: The scoring and filtering submodule obtains the corresponding score values ​​of the components based on the wound adaptation score list, judges the score values ​​item by item according to the adaptation judgment criteria, marks the components that do not meet the adaptation conditions and records the component number, and sorts the filtering results in order of number to obtain the set of unsuitable component tags. The wound fit score list was read, and the scores were evaluated item by item in ascending order of component number to generate a set of labels for incompatible components. A dual-threshold structure was used for fit determination: the first threshold excluded low-scoring components, and the second threshold labeled boundary components. Threshold settings were determined through experimental control, using follow-up records of 40 wound cases. Recorded fields included component scores, changes in bacterial culture 24 hours after drug administration, changes in wound exudate volume, and local adverse reaction records. Cases with negative bacterial culture or a decrease in colony count of at least 90.0% were labeled as the pass group, and others as the fail group. The score distributions of the two groups were compared, and the 75th percentile of the fail group was used as the first threshold, and the 25th percentile of the pass group was used as the second threshold. Adjustments were made on the validation set to minimize the average of the missed labeling rate and the false labeling rate. In this example, the first threshold is 70.0 points, and the second threshold is 85.0 points. The rule is that scores below 70.0 points are directly marked as unsuitable, scores between 70.0 and 85.0 points are marked as pending review, and scores above 85.0 points are marked as suitable. When a label is written to the set, a trigger reason field is also written. This trigger reason field only records the threshold interval number and the score value, without any purpose description. In this example, P2's score is 82.2 points, falling within the 70.0 to 85.0 point range, and it enters the pending review label set. If P3 has a score of 66.0 points, it enters the unsuitable label set.

[0041] The proportion statistics submodule obtains the number of marked components based on the mismatch component marker set, and at the same time obtains the number of all components in the wound compatibility score list. It calculates the proportion of the two types of quantities to form proportion data reflecting the proportion of mismatch components, and generates the mismatch component proportion. The system reads the set of mismatched component markers and the total group score of the scoring list, performs percentage calculation, and generates the percentage of mismatched components. The percentage calculation process uses textual arithmetic logic: first, it counts the number of mismatched markers in the marker set; then, it counts the total number of components in the scoring list; finally, it divides the number of mismatched components by the total number to obtain the percentage value, rounded to two decimal places. In the example, the total group score is 5, and the number of mismatched markers is 1, so the percentage value is 0.20. Simultaneously, the percentage pending review is calculated in the same way. In the example, if the number of pending reviews is 2, the percentage pending review is 0.40; if the number of pending reviews is 1, the percentage pending review is 0.20. Both percentages are written to the proportional data record and output to the security judgment submodule.

[0042] The security determination submodule compares the proportion of incompatible components with the preset evaluation threshold, classifies the application status according to the threshold range in which the proportion falls, organizes the corresponding determination tags, and generates the application security assessment conclusion. The proportion of incompatible components was compared with preset assessment thresholds to generate an application safety assessment conclusion. The assessment thresholds used a three-segment interval, with interval boundaries determined through historical adverse event statistics. The statistical sample consisted of 100 cases, recording the component proportion and adverse event occurrence for each case. Adverse events were determined based on any one of the following three criteria: increased local redness and swelling, an increase in pain score of at least 2.0 points, or a fever of at least 38.0 degrees Celsius. The incidence of adverse events was analyzed using binning techniques to measure changes in the proportion of incompatible components, and the interval boundaries were selected based on significant jumps in the incidence rate. In the example, the incidence rate of adverse events is 8.0% when the proportion is not higher than 0.20, 18.0% when the proportion is between 0.20 and 0.40, and 35.0% when the proportion is higher than 0.40. Based on this, interval 1 is marked as available when the proportion is not higher than 0.20, interval 2 is marked as restricted when the proportion is between 0.20 and 0.40, and interval 3 is marked as unavailable when the proportion is higher than 0.40. When comparing, the inclusion boundary rule is used: interval 1 includes 0.20, and interval 2 does not include 0.20 but includes 0.40. In the example, the proportion of the incompatible component is 0.20, falling into interval 1. The output application safety assessment conclusion is "usable label." Simultaneously, the proportion pending review (0.420) is written to the additional prompt field for subsequent manual review. A summary of the control experiment data is also output, for example: incidence rate of 8.0% in interval 1, 18.0% in interval 2, and 35.0% in interval 3. A description of the differences from the control groups is also provided, for example: the current proportion corresponds to an incidence rate group of 8.0%, a decrease of 77.1% compared to 35.0% in interval 3. The advantage of this calculation logic is that by linking the component scoring threshold with the proportion interval comparison, the consistency between the conclusion label and the historical adverse event incidence rate grouping reaches 0.88 in the validation set, an improvement of 11.4% compared to the consistency of 0.79 when only a single scoring threshold is used to output the conclusion.

[0043] Table 3. Results of phage component periodicity and adaptation score Component number Pyrolysis cycle, hours Duration, hours Coincidence ratio Matching degree index, score mark P1 1.5 36.0 1.00 100.0 adaptation P2 3.0 24.0 0.58 82.2 pending review P3 4.5 18.0 0.33 66.0 Incompatible P4 1.0 12.0 0.50 79.5 pending review P5 2.0 48.0 0.80 93.0 adaptation As shown in Table 3, the overlap ratio between the periodic extraction and the segment matching output is weighted and summarized into a matching degree index by the adaptation evaluation, and forms a set of adapted, pending, and unsuitable labels in the scoring and screening. The ratio statistics and security judgment output the application security assessment conclusion based on this.

[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 wound infection safety assessment system based on phage cocktails, characterized in that, The system includes: The wound adaptation information construction module is used to receive exudate and tissue fluid sample data from the infected wound area collected by the sampling device, extract neutrophil count, lactate dehydrogenase concentration and total protein concentration data, and integrate them in chronological order to form a wound adaptation information sequence group. The immune intervention window identification module extracts the neutrophil number change sequence based on the wound adaptation information sequence group, combines the rising trend of lactate dehydrogenase concentration within the same time period, and filters the time period that meets the preset threshold condition based on the synchronicity and change amplitude of the two in the time dimension, and generates an immune active time segment identifier. Combined with the lactate dehydrogenase concentration change trend, the module performs joint screening to identify immune active time periods and generate immune active time segment identifiers. The phage inactivation risk assessment module obtains the concentration change trends of myeloperoxidase and defensin based on the immune active time segment identifier, extracts the intensity and time point, calculates the immune enzyme synergistic level index, and compares it with the phage inactivation reference standard to generate inactivation risk judgment result. The cocktail compatibility analysis module calls the inactivation risk assessment results, combines the lysis cycle information of the phage cocktail components, analyzes the relationship between the duration and the coverage of the immune active time period, judges the compatibility of the components with the wound environment, and generates a wound compatibility score list. The application safety conclusion generation module calculates the proportion of components that do not meet the adaptation conditions based on the wound adaptation score list, compares it with the preset evaluation threshold, and generates an application safety evaluation conclusion. The preset evaluation thresholds include three graded thresholds: 0.10, 0.30, and 0.

50. These thresholds are set based on the statistical relationship between the proportion of incompatible components and the incidence of adverse reactions in historical clinical cases, and are used to determine the application safety level of phage cocktails.

2. The wound infection safety assessment system based on phage cocktails according to claim 1, characterized in that: The wound adaptation information sequence group includes continuous recordings of neutrophil count, lactate dehydrogenase concentration, and total protein concentration; the immune active time segment identifier includes the immune active start time point, immune active end time point, and duration of immune active. The inactivation risk assessment results include the level of immune enzyme synergy index, the conclusion of the phage inactivation reference standard, and the inactivation risk level; the wound compatibility score list includes the component compatibility score, the component coverage matching score, and the component compatibility status marker; the application safety assessment conclusion includes the incompatible component proportion judgment value, the threshold control judgment conclusion, and the safety assessment level conclusion.

3. The wound infection safety assessment system based on phage cocktails according to claim 1, characterized in that: The wound adaptation information construction module includes: The exudate extraction submodule collects exudate and tissue fluid from the infected wound area, records the collection time and labels the sample number, organizes the samples in chronological order, extracts the liquid volume information and constructs a structured data list to generate a continuous liquid volume sequence. The neutrophil analysis submodule obtains samples with corresponding numbers based on the continuous liquid volume sequence, extracts neutrophil count information and lactate dehydrogenase concentration information, performs correlation analysis on the correlation parameters according to the time sequence, calculates the relationship between changes in neutrophils and lactate dehydrogenase, and generates dynamic parameters of neutrophils. The adaptation information sequence generation submodule extracts lactate dehydrogenase concentration information and total protein concentration information at corresponding time points based on the granulocyte dynamic parameters, integrates the three types of data to construct sequence data units, connects the units in chronological order to form a complete sequence, and generates a wound adaptation information sequence group.

4. The wound infection safety assessment system based on phage cocktails according to claim 3, characterized in that: The immune intervention window recognition module includes: The sequence extraction submodule extracts the neutrophil count value and sampling time corresponding to the time point of the wound adaptation information sequence group, constructs the count change sequence in chronological order, filters continuous fragments according to the change direction and duration, and generates neutrophil change intervals. The trend screening submodule extracts the lactate dehydrogenase concentration value at the corresponding time point based on the granulocyte change interval, determines whether the direction of change and time span of the concentration value in a continuous period meet the screening requirements, establishes a continuous segment that meets the conditions, and generates a lactate dehydrogenase trend segment. The time recognition submodule calls the neutrophil number change sequence corresponding to the lactate dehydrogenase trend segment, and determines whether the immune intervention threshold standard is met based on the synchronous change relationship between the two within the same time period. It locates the time period that continuously meets the conditions and generates an immune active time segment identifier. The immune intervention threshold standard includes: within the time period, the change in the number of neutrophils is not less than a preset neutrophil threshold, the change in lactate dehydrogenase concentration is not less than a preset lactate dehydrogenase threshold, and the two change in the same direction.

5. The wound infection safety assessment system based on phage cocktails according to claim 4, characterized in that: The process of determining whether the direction of change and time span of lactate dehydrogenase concentration value in a continuous period meet the screening requirements is as follows: the direction of change is determined by the sign of the difference between the lactate dehydrogenase concentration values ​​corresponding to each two adjacent time points. If the direction of change of lactate dehydrogenase concentration value is consistent in no less than three consecutive time points, and the total duration of the continuous time period is not less than 48 hours, then the direction of change is considered to meet the screening requirements. The process of locating the time period that continuously meets the conditions is as follows: under the premise of meeting the screening requirements of the direction of change of lactate dehydrogenase concentration and time span, extract the difference in the number of time points in the sequence of changes in the number of neutrophils within the time period, and determine whether the difference in the number of time points is not less than the set neutrophil threshold. If the conditions are met, the corresponding time period is identified as the immune active time segment. The neutrophil threshold is set based on experimental repeatability data, and is preferably not less than 0.09 × 10⁻⁶. 6 The difference between the number of samples per milliliter, or the adjustable parameters automatically set by the testing platform.

6. The wound infection safety assessment system based on phage cocktails according to claim 4, characterized in that: The phage inactivation risk assessment module includes: The indicator extraction submodule obtains the trends of myeloperoxidase concentration and defensin concentration within the time period based on the immune active time segment identifier, extracts the intensity of change and corresponding time value at the time point, and organizes them into an indicator data sequence in chronological order to generate an immune-related indicator sequence. The collaborative evaluation submodule calls the data of myeloperoxidase concentration and defensin concentration at the corresponding time points in the immune-related index sequence, performs parallel evaluation on the direction and magnitude of change of the two indicators, determines the degree of synergy in the time series, and generates an immune enzyme synergy level index. The risk assessment submodule extracts the entire range of values ​​based on the immune enzyme synergy level index, performs interval matching with the phage inactivation reference standard, classifies and identifies the index values ​​within the time period, determines whether they are within the inactivation assessment range, and generates an inactivation risk assessment result.

7. The wound infection safety assessment system based on phage cocktails according to claim 6, characterized in that: The start and end range of the time period for which the immune active time segment identifier is obtained shall not be less than 24 hours, and the interval between consecutive sampling frequencies shall not be higher than 2 hours. In the trends of myeloperoxidase concentration change and defensin concentration change, the intensity of the change is extracted based on the criterion that the concentration difference between two adjacent time points is not less than the set concentration fluctuation threshold of 0.1 μg / mL. Time point data that do not meet the condition are not included in the immune-related index sequence. The immune enzyme synergistic level index is calculated based on the concentration change trend of myeloperoxidase and defensin during the period of immune activity, taking into account the consistency of their change direction and the difference in their magnitude. The amplitude difference is calculated by normalizing the change intensity of the two indicators and then generating the collaborative level value in reverse according to the preset mapping function. The range of the synergistic level index corresponding to the inactivation determination range is from -0.3 to 0.

3. Data outside the range are processed as non-inactivation markers. In the process of generating the synergistic level index of the immune enzyme, the consistency requirement for the change direction of the myeloperoxidase concentration and the defensin concentration is that the proportion of consistent direction in three or more adjacent time points is not less than 70%. In the process of generating the inactivation risk assessment result, the value range of the immune enzyme synergistic level index corresponding to the inactivation assessment range is -0.3 to 0.3, and data outside the range are treated as non-inactivation data.

8. The wound infection safety assessment system based on phage cocktails according to claim 6, characterized in that: The cocktail compatibility analysis module includes: The periodic extraction submodule calls the inactivation risk judgment result to obtain all component information in the phage cocktail, extracts the lysis cycle and duration data of the components, classifies and sorts them in combination with the corresponding component numbers, establishes a correspondence table between component numbers and cycle duration, and generates a component cycle mapping list. The segment matching submodule extracts the duration segments of components according to the component period mapping list, compares the overlap ratio of the time period with the immune active time segment identifier on the time axis, calculates the coverage degree and marks the difference range, and generates the time segment coverage relationship. The adaptation assessment submodule calls the time segment coverage relationship and component cycle duration data, assigns weights to the duration coverage and degree of overlap of each component with the immunization time period, calculates the matching degree index, scores all components, and generates a wound adaptation score list. The weighting is a preset ratio allocation, with the duration coverage score and the overlap score of the immunization period weighted at 65% and 35% respectively to calculate the matching degree index.

9. The wound infection safety assessment system based on phage cocktails according to claim 8, characterized in that: The application security conclusion generation module includes: The scoring and filtering submodule obtains the score value corresponding to the component based on the wound adaptation score list, judges the score value item by item according to the adaptation judgment criteria, marks the components that do not meet the adaptation conditions and records the component number, and sorts the filtering results in order of number to obtain the set of unsuitable component tags. The proportion statistics submodule obtains the number of marked components based on the set of mismatched component markers, and at the same time obtains the number of all components in the wound compatibility scoring list. It calculates the proportion of the two types of quantities to form proportion data reflecting the proportion of mismatched components and generates the proportion of mismatched components. The security determination submodule compares the proportion of the incompatible components with a preset evaluation threshold, classifies the application status according to the threshold range in which the proportion falls, organizes the corresponding determination tags, and generates an application security assessment conclusion.

10. The wound infection safety assessment system based on phage cocktails according to claim 9, characterized in that: The process of calculating the ratio is as follows: the number of components included in the set of unsuitable component markers and the total number of components in the wound compatibility score list are used as the numerator and denominator respectively to perform a ratio calculation. The result of the ratio calculation is retained in decimal form with two decimal places. The interval comparison process involves comparing the proportion of the unsuitable component with three thresholds of 0.10, 0.30, and 0.50 in the preset evaluation threshold set. If the proportion of the unsuitable component is less than 0.10, it is marked as "applicable". If the proportion of the incompatible component is between 0.10 and 0.30, which is greater than or equal to 0.10 and less than 0.30, it is marked as "limited application"; If the proportion of the incompatible component is between 0.30 and 0.50, which is greater than or equal to 0.30 and less than 0.50, it is marked as "use with caution"; If the proportion of the incompatible component is greater than or equal to 0.50, it is marked as "unapplicable".

Images