A method and system for early warning of chronic obstructive pulmonary disease

Abstract

The application discloses a kind of early warning method and system of chronic obstructive pulmonary disease, by obtaining the current environmental data and physiological data of patient, specific disease causes and its current level are identified and evaluated, such as environmental temperature difference, haze concentration, environmental low temperature, respiratory rhythm anomaly and biomarker level etc., according to the type of cause and real-time level, from preset rule base, corresponding early warning threshold adjustment strategy is adaptively matched and applied, to dynamically generate individualized early warning threshold that fits current risk condition.Through comparing real-time physiological data with dynamic threshold, precise early warning is realized.The application realizes the comprehensive response to complex risk situation by intelligently integrating the adjustment strategies of multiple causes, significantly improves the accuracy and timeliness of early warning, and provides an effective technical means for early intervention of chronic obstructive pulmonary disease.

Description

Technical Field

[0001] This application relates to the field of computer-based intelligent medical technology, and in particular to a method and system for early warning of chronic obstructive pulmonary disease. Background Technology

[0002] Chronic obstructive pulmonary disease (COPD), a prevalent chronic respiratory disease, often experiences significant exacerbations due to acute exacerbations, severely impacting patients' quality of life and increasing the healthcare burden. Current clinical practice primarily relies on regular pulmonary function tests or static assessment mechanisms based on fixed physiological parameter thresholds for COPD monitoring. These methods struggle to capture dynamic changes in risk factors encountered by patients in their daily lives and fail to reflect the complex interactions between environmental stimuli and individual physiological states, thus limiting the timeliness and individual adaptability of early warning systems.

[0003] The triggers for acute exacerbations of COPD are highly diverse, including increased smog concentrations, drastic fluctuations in ambient temperature, prolonged exposure to low temperatures, elevated respiratory inflammatory markers, and emotional stress. While existing remote health monitoring systems can continuously collect patients' physiological data, their warning logic generally employs uniform and static threshold judgment rules, failing to differentiate responses based on different trigger types and their actual intensity. Especially in real-life scenarios, various triggers often coexist and overlap. Fixed threshold strategies cannot identify the contribution weight of specific triggers, nor can they dynamically adjust warning sensitivity based on their real-time levels. This leads to the system being prone to missed warnings in high-risk situations and frequent false alarms in low-risk situations.

[0004] Therefore, there is an urgent need for an intelligent early warning method that can integrate multi-source environmental and physiological data, accurately identify specific disease triggers, and adaptively adjust the early warning criteria based on the real-time trigger level, so as to achieve individualized, context-aware risk prediction and early intervention support for COPD patients. Summary of the Invention

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] A method for early warning of chronic obstructive pulmonary disease, applied to a computer system, the method comprising:

[0007] S1, acquire the patient's current environmental data and current physiological data, and determine multiple target triggers for the onset of the disease based on the current environmental data;

[0008] S2, based on the current environmental data and the current physiological data, assess the current precipitating factor level of each of the multiple target disease-inducing factors;

[0009] S3, based on the type and current level of each of the multiple target disease triggers, adaptively determine multiple warning threshold adjustment strategies from a preset warning strategy rule base. The warning strategy rule base defines the mapping relationship between different trigger types and different trigger level ranges and warning threshold adjustment strategies. The warning threshold adjustment strategies are used to dynamically adjust the warning judgment threshold.

[0010] S4, according to the adaptively determined multiple early warning threshold adjustment strategies, adjust the benchmark threshold used for early warning determination to obtain the dynamic early warning threshold;

[0011] S5, compare the current physiological data with the dynamic early warning threshold, and when the current physiological data meets the early warning triggering condition based on the dynamic early warning threshold, generate and output early warning information for the patient's chronic obstructive pulmonary disease.

[0012] Furthermore, the method also includes:

[0013] The target triggering factors include at least one of smog concentration, ambient temperature difference, and low ambient temperature.

[0014] The determination of multiple target triggering factors based on the current environmental data includes:

[0015] Extract the haze concentration monitoring value for a continuous first predetermined time period from the current environmental data;

[0016] Calculate the average haze concentration over the first predetermined time period;

[0017] If the average concentration of haze exceeds the first environmental reference value, then the haze concentration is determined as the target inducing factor for the disease.

[0018] Furthermore, the method also includes:

[0019] The target triggers for the disease include environmental temperature differences;

[0020] Based on the current environmental data, assessing the current precipitating factor level of each of the target disease-causing factors includes:

[0021] Obtain the ambient temperature value at the current time point from the current environmental data, and obtain historical ambient temperature data within the past second predetermined time period;

[0022] Based on the historical ambient temperature data, calculate the average ambient temperature value within the past second predetermined time period;

[0023] Calculate the absolute difference between the current ambient temperature value and the average ambient temperature value, and use it as the current temperature difference range;

[0024] The current temperature difference is divided into multiple predefined level levels, which include at least low level temperature difference, medium level temperature difference and high level temperature difference;

[0025] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0026] Based on the type of cause, namely environmental temperature difference, and the level of the current temperature difference, query the warning strategy rule base;

[0027] The early warning strategy rule base contains the following pre-stored rules:

[0028] When the trigger type is ambient temperature difference and the trigger level is low temperature difference, a first adjustment strategy is associated, which indicates that the baseline threshold is lowered by a first amount; when the trigger type is ambient temperature difference and the trigger level is medium temperature difference, a second adjustment strategy is associated, which indicates that the baseline threshold is lowered by a second amount, the second amount being greater than the first amount; when the trigger type is ambient temperature difference and the trigger level is high temperature difference, a third adjustment strategy is associated, which indicates that the baseline threshold is lowered by a third amount, the third amount being greater than the second amount.

[0029] Furthermore, the method also includes:

[0030] The target triggers include smog concentration; based on the current environmental data, assessing the current trigger levels of each of the target triggers includes:

[0031] Obtain the real-time haze concentration value at the current time point from the current environmental data, and obtain the haze concentration change sequence within the third predetermined time period before the current time point;

[0032] Analyze the haze concentration change sequence to determine whether the real-time haze concentration value is in a continuous upward trend;

[0033] If the real-time haze concentration value is in a continuous upward trend, then calculate the duration of the continuous upward trend and the cumulative increase in concentration.

[0034] Based on the duration and the cumulative increase in concentration, combined with the real-time haze concentration value itself, a comprehensive level index of haze causes is obtained by weighted calculation.

[0035] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0036] Based on the type of cause, namely haze concentration, and the comprehensive level index of haze causes, query the early warning strategy rule base;

[0037] The early warning strategy rule base pre-stores: dividing the numerical range of the comprehensive level index of haze causes into multiple continuous intervals, with each interval associated with a different early warning threshold adjustment strategy. The early warning threshold adjustment strategy includes a reduction ratio of the baseline threshold, and the reduction ratio increases as the interval to which the comprehensive level index of haze causes belongs increases.

[0038] Furthermore, the method also includes:

[0039] Based on the current physiological data, multiple physiological target inducing factors are identified, including abnormal respiratory rhythm indicators;

[0040] The step of determining multiple physiological target precipitating factors based on the current physiological data includes: assessing the current precipitating factor level of the abnormal respiratory rhythm indicators based on the current physiological data;

[0041] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: determining the corresponding early warning threshold adjustment strategy from the early warning strategy rule base based on the trigger type of the respiratory rhythm abnormality index and its current trigger level.

[0042] Furthermore, this method assesses the current precipitating factor level of respiratory rhythm abnormality indicators in the following ways:

[0043] Extract respiratory waveform data for a fourth predetermined time period from the current physiological data;

[0044] Identify the duration of each respiratory cycle in the respiratory waveform data;

[0045] Calculate the coefficient of variation between the durations of consecutive respiratory cycles;

[0046] If the coefficient of variation exceeds the first physiological reference value, then the causative level of the abnormal respiratory rhythm index is determined to be abnormal.

[0047] Furthermore, the method also includes:

[0048] The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategy to obtain the dynamic early warning threshold includes:

[0049] When there are multiple target triggers for disease onset, obtain the early warning threshold adjustment strategy determined for each target trigger.

[0050] The adjustment operation on the baseline threshold indicated by each early warning threshold adjustment strategy is analyzed, and the adjustment operation includes the adjustment direction and adjustment amount;

[0051] All adjustment operations are integrated according to preset conflict resolution rules to generate a composite adjustment command;

[0052] The baseline threshold is adjusted once according to the composite adjustment instruction to obtain a unified dynamic early warning threshold.

[0053] Furthermore, the method targets disease-causing factors including the levels of biomarkers characterizing the risk of respiratory inflammation, and based on the current physiological data, identifies multiple physiological target disease-causing factors, including:

[0054] Obtain the current biomarker concentration value measured by the detection device from the current physiological data;

[0055] Obtain the baseline biomarker concentration values ​​of the patients;

[0056] Calculate the ratio of the current biomarker concentration value to the baseline biomarker concentration value;

[0057] The current precipitating factor level of the biomarker is determined based on the ratio, and the level includes at least normal level, slightly elevated level, and significantly elevated level.

[0058] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0059] In response to determining that the current level of the trigger for the biomarker is slightly elevated, a first biomarker adjustment strategy is obtained from the warning strategy rule base. The first biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter used for warning determination is lowered by a first proportion, while keeping the baseline threshold of the second physiological parameter used for warning determination unchanged.

[0060] In response to determining that the current trigger level of the biomarker is significantly elevated, a second biomarker adjustment strategy is obtained from the warning strategy rule base. The second biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter is lowered by a second proportion and the baseline threshold of the second physiological parameter is lowered by a third proportion, wherein the second proportion is greater than the first proportion.

[0061] Target triggers include low environmental temperature. Based on the current environmental data, the current trigger level of each of the target triggers is assessed, including:

[0062] Obtain the current ambient temperature value from the current environmental data;

[0063] If the current ambient temperature value is lower than the second ambient reference value, then the duration for which the current ambient temperature value is lower than the second ambient reference value is further obtained;

[0064] Based on the extent to which the current ambient temperature is lower than the second environmental reference value and the duration of the difference, the low temperature exposure intensity index is calculated.

[0065] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0066] Based on the type of cause, namely low environmental temperature, and the low temperature exposure intensity index, the early warning strategy rule base is queried to obtain the corresponding early warning threshold adjustment strategy. The early warning threshold adjustment strategy indicates that the adjustment amount of the benchmark threshold is positively correlated with the low temperature exposure intensity index.

[0067] Furthermore, the method also includes:

[0068] Identify whether the current physiological data contains indicators that characterize the patient's emotional tension.

[0069] If included, the level of emotional stress triggers is assessed based on the data of the indicators representing the patient's emotional stress.

[0070] Based on the type of emotional stress trigger and the assessed level of emotional stress trigger, a corresponding emotional-related early warning threshold adjustment strategy is determined from the early warning strategy rule base;

[0071] The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategies to obtain the dynamic early warning threshold includes: combining the emotion-related early warning threshold adjustment strategies determined as emotional tension triggers to comprehensively adjust the baseline threshold.

[0072] According to a second aspect of the present invention, the present invention claims protection for an early warning system for chronic obstructive pulmonary disease, comprising:

[0073] One or more processors;

[0074] A memory having stored one or more programs that, when executed by one or more processors, enable the one or more processors to implement the aforementioned method for early warning of chronic obstructive pulmonary disease.

[0075] This invention discloses an early warning method and system for chronic obstructive pulmonary disease (COPD). By acquiring the patient's current environmental and physiological data, it identifies and assesses specific triggers and their current levels, such as ambient temperature difference, smog concentration, low ambient temperature, abnormal respiratory rhythm, and biomarker levels. Based on the trigger type and real-time level, it adaptively matches and applies corresponding early warning threshold adjustment strategies from a preset rule base, thereby dynamically generating individualized early warning thresholds that fit the current risk situation. Accurate early warning is achieved by comparing real-time physiological data with dynamic thresholds. This invention intelligently integrates adjustment strategies for multiple triggers, enabling a comprehensive response to complex risk situations, significantly improving the accuracy and timeliness of early warning, and providing an effective technical means for early intervention of COPD. Attached Figure Description

[0076] Figure 1 A flowchart illustrating the workflow of an early warning method for chronic obstructive pulmonary disease claimed in an embodiment of the present invention;

[0077] Figure 2 This is a second flowchart illustrating an early warning method for chronic obstructive pulmonary disease claimed in an embodiment of the present invention.

[0078] Figure 3 A third flowchart illustrating an early warning method for chronic obstructive pulmonary disease claimed in an embodiment of the present invention;

[0079] Figure 4 The fourth flowchart is a method for early warning of chronic obstructive pulmonary disease claimed in an embodiment of the present invention. Detailed Implementation

[0080] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0081] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include multiple such features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0082] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in multiple embodiments of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0083] According to the first embodiment of the present invention, referring to Figure 1 This invention claims protection for a method for early warning of chronic obstructive pulmonary disease, applied to a computer system, the method comprising:

[0084] S1, acquire the patient's current environmental data and current physiological data, and determine multiple target triggers for the onset of the disease based on the current environmental data;

[0085] S2, based on the current environmental data and the current physiological data, assess the current precipitating factor level of each of the multiple target disease-inducing factors;

[0086] S3, based on the type and current level of each of the multiple target disease triggers, adaptively determine multiple warning threshold adjustment strategies from a preset warning strategy rule base. The warning strategy rule base defines the mapping relationship between different trigger types and different trigger level ranges and warning threshold adjustment strategies. The warning threshold adjustment strategies are used to dynamically adjust the warning judgment threshold.

[0087] S4, according to the adaptively determined multiple early warning threshold adjustment strategies, adjust the benchmark threshold used for early warning determination to obtain the dynamic early warning threshold;

[0088] S5, compare the current physiological data with the dynamic early warning threshold, and when the current physiological data meets the early warning triggering condition based on the dynamic early warning threshold, generate and output early warning information for the patient's chronic obstructive pulmonary disease.

[0089] In this embodiment, the system continuously or periodically acquires input data from connected environmental sensors such as temperature and humidity sensors, particulate matter sensors, and patient-worn physiological monitoring devices such as smart bracelets and patch-type heart rate and respiration monitors via its data interface module. The acquired current environmental data includes at least geographic location information, ambient temperature, humidity, and concentrations of specific-sized particulate matter in the air, such as PM2.5 and PM10. The acquired current physiological data includes at least the patient's real-time respiratory rate, blood oxygen saturation, and heart rate data. This data is received in a structured format and temporarily stored in a data buffer.

[0090] Next, the trigger identification phase begins. The core processing unit analyzes the current environmental data and, based on a pre-built medical knowledge base and rules, determines whether there are any potential triggers that reach the warning level. Specifically, it compares each environmental parameter with its preset reference baseline or safety threshold. For example, it checks whether the current PM2.5 concentration exceeds a set air quality concern threshold, or whether the current ambient temperature is below a set low-temperature warning line, or calculates whether the recent fluctuation in ambient temperature is too large. If any one or more parameters exceed their corresponding concern threshold, the corresponding trigger type, such as smog concentration, low ambient temperature, or ambient temperature difference, is identified as the target trigger.

[0091] Next, a level assessment is performed on each identified target trigger. This assessment is not a simple yes or no judgment, but rather a quantitative classification of its intensity or risk level. For environmental temperature difference triggers, the assessment might be based on the difference between the highest and lowest temperatures over the past 24 hours; for smog concentration triggers, the assessment might be based on how many times the current concentration exceeds the standard value; for environmental low temperature triggers, the assessment might be based on the degree to which the current temperature is below the comfortable temperature. The assessment results are quantified into one or more comparable trigger level indicators, such as low, medium, and high levels, or a continuous numerical index.

[0092] Subsequently, based on the type of trigger and its specific level assessment results, an adaptive decision-making process for the early warning strategy is executed. An internally pre-built early warning strategy rule base is essentially a multidimensional lookup table or a set of conditional judgment rules. It uses the trigger type and trigger level as joint input keys. For example, the rule base might define: when the trigger type is ambient temperature difference and the level is high, strategy A is used; when the trigger type is smog concentration and the level is medium, strategy B is used. The core output of these strategies is an instruction to adjust the early warning threshold for one or more physiological parameters, such as respiratory rate, for example, lowering the respiratory rate early warning threshold by 10% or raising the blood oxygen saturation early warning threshold by 5%.

[0093] Next, the aforementioned adjustment strategy is applied. It retrieves the original, unadjusted baseline threshold for a specific physiological parameter, such as respiratory rate. This baseline threshold may be set based on the patient's individual baseline or population statistical standards. Then, it strictly adjusts the baseline threshold numerically according to the strategy instructions, thereby generating a dynamic warning threshold suitable for the current specific risk situation. If multiple adjustment strategies are applied simultaneously, these adjustment effects are combined according to preset priorities or composite calculation rules, ultimately generating a unified set of dynamic thresholds.

[0094] Next, the patient's real-time physiological data, such as instantaneous respiratory rate, is compared with the calculated dynamic warning threshold. This is a judgment process that checks whether the physiological data has exceeded the context-adjusted safety boundaries.

[0095] Finally, if the comparison results meet the warning triggering conditions, such as the respiratory rate being higher than the dynamic warning threshold for three consecutive monitoring cycles, a structured warning message is generated. This message includes at least the patient identifier, the physiological parameter that triggered the warning, the current dynamic threshold, relevant triggering factors, and a timestamp. This warning message is then output to healthcare personnel or the patient via an output module through one or more of the following methods: a graphical interface pop-up, SMS, application notification, or sound alarm.

[0096] Furthermore, the method also includes:

[0097] The target triggering factors include at least one of smog concentration, ambient temperature difference, and low ambient temperature.

[0098] The determination of multiple target triggering factors based on the current environmental data includes:

[0099] Extract the haze concentration monitoring value for a continuous first predetermined time period from the current environmental data;

[0100] Calculate the average haze concentration over the first predetermined time period;

[0101] If the average concentration of haze exceeds the first environmental reference value, then the haze concentration is determined as the target inducing factor for the disease.

[0102] In this embodiment, a dedicated analysis routine for haze concentration data is executed. This routine first extracts a continuous sequence of haze concentration monitoring values ​​from the input environmental data stream over a specific past time period, such as a first predetermined time interval (e.g., 6 hours). These monitoring values ​​are recorded sequentially, possibly one data point per minute or every ten minutes.

[0103] Next, an arithmetic mean was calculated for all valid monitoring values ​​within that time period to smooth out short-term fluctuations and obtain an average value that represents the overall haze exposure level during that period. This calculation process eliminates possible sensor transient error interference.

[0104] The calculated average haze concentration is then compared to a pre-stored first environmental reference value. This first environmental reference value, set according to environmental medicine research or public health standards, represents a threshold haze concentration level that should raise concern for COPD patients. It may be a fixed concentration value, such as PM2.5 at 75 micrograms per cubic meter, or a dynamic value adjusted according to the season or region.

[0105] Finally, a judgment is made based on the comparison results: if the calculated average value is significantly greater than the first environmental reference value, the current level of smog is logically determined to pose a significant risk, and smog concentration is officially marked as a trigger type and added to the target disease trigger set for this early warning calculation. Conversely, if the average value does not exceed the reference value, the current smog level is considered insufficient to be a high-priority trigger on its own and may not be included or may only be used as a secondary reference.

[0106] Furthermore, the method also includes:

[0107] The target triggers for the disease include environmental temperature differences;

[0108] Reference Figure 2 Based on the current environmental data, assessing the current precipitating factor level of each of the multiple target disease-causing factors includes:

[0109] Obtain the ambient temperature value at the current time point from the current environmental data, and obtain historical ambient temperature data within the past second predetermined time period;

[0110] Based on the historical ambient temperature data, calculate the average ambient temperature value within the past second predetermined time period;

[0111] Calculate the absolute difference between the current ambient temperature value and the average ambient temperature value, and use it as the current temperature difference range;

[0112] The current temperature difference is divided into multiple predefined level levels, which include at least low level temperature difference, medium level temperature difference and high level temperature difference;

[0113] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0114] Based on the type of cause, namely environmental temperature difference, and the level of the current temperature difference, query the warning strategy rule base;

[0115] The early warning strategy rule base contains the following pre-stored rules:

[0116] When the trigger type is ambient temperature difference and the trigger level is low temperature difference, a first adjustment strategy is associated, which indicates that the baseline threshold is lowered by a first amount; when the trigger type is ambient temperature difference and the trigger level is medium temperature difference, a second adjustment strategy is associated, which indicates that the baseline threshold is lowered by a second amount, the second amount being greater than the first amount; when the trigger type is ambient temperature difference and the trigger level is high temperature difference, a third adjustment strategy is associated, which indicates that the baseline threshold is lowered by a third amount, the third amount being greater than the second amount.

[0117] In this embodiment, during the assessment of the trigger level, the following detailed steps are performed: First, the ambient temperature value at the current moment or the most recent sampling moment is accurately read from the current environmental data and denoted as T_current. Simultaneously, all ambient temperature data recorded at fixed intervals over a specific period of time preceding the current moment (i.e., a second predetermined time period, such as the past 12 hours) is retrieved from the historical environmental database to form a historical temperature sequence.

[0118] Secondly, the historical temperature series is processed. The arithmetic mean of all temperature values ​​in the series is calculated to obtain the average ambient temperature value for that time period, denoted as T_average. This average value represents the recent background temperature level.

[0119] Next, the key performance indicator (KPI) is calculated. It calculates the absolute value of the difference between the current temperature value T_current and the recent average temperature T_average, i.e., |T_current - T_average|. This absolute value is defined as the current temperature difference magnitude, which quantifies the degree of deviation of the instantaneous temperature relative to the recent background.

[0120] Then, the temperature difference range is classified. A pre-defined temperature difference classification table divides the numerical range of temperature difference into several non-overlapping intervals, each corresponding to a predefined level. For example, a temperature difference range between 2 and 4 degrees Celsius is defined as a low-level temperature difference; between 4 and 6 degrees Celsius is defined as a medium-level temperature difference; and above 6 degrees Celsius is defined as a high-level temperature difference. Based on the calculated specific value of the current temperature difference range, it is assigned to the corresponding level.

[0121] During the strategy adaptation and determination phase, based on the determined trigger type (environmental temperature difference) and specific level (e.g., medium temperature difference), the early warning strategy rule base is queried. The rule base pre-associates a specific early warning threshold adjustment strategy for each level under the environmental temperature difference type. The core of these strategies is to specify how to adjust the baseline thresholds of physiological parameters.

[0122] In detail, the mapping relationships in the rule base are specifically defined as follows:

[0123] When the query conditions are "trigger type = ambient temperature difference" and "trigger level = low temperature difference", the associated first adjustment strategy will explicitly state that one or more physiological parameters used for early warning determination, such as the baseline threshold of resting respiratory rate, will be lowered. The magnitude of the reduction is a pre-set first magnitude, such as a 5% reduction of the baseline value.

[0124] When the query conditions are "trigger type = environmental temperature difference" and "trigger level = medium-level temperature difference", the associated second adjustment strategy will explicitly state: lower the baseline threshold of the same physiological parameter by a larger second magnitude, such as lowering the baseline value by 10%. Furthermore, the design ensures that the value of the second magnitude is greater than the first magnitude, reflecting the principle that the higher the risk, the more sensitive the warning.

[0125] When the query conditions are: trigger type = ambient temperature difference and trigger level = high-level temperature difference, the associated third adjustment strategy will explicitly indicate: lower the baseline threshold by a more significant third magnitude, such as lowering the baseline value by 15%, and the third magnitude is greater than the second magnitude.

[0126] Through such refined grading and strategy mapping, a precise and gradient response to environmental temperature difference-induced factors is achieved.

[0127] Furthermore, the method also includes:

[0128] The target triggers for the disease include smog concentration;

[0129] Reference Figure 3 Based on the current environmental data, assessing the current precipitating factor level of each of the multiple target disease-causing factors includes:

[0130] Obtain the real-time haze concentration value at the current time point from the current environmental data, and obtain the haze concentration change sequence within the third predetermined time period before the current time point;

[0131] Analyze the haze concentration change sequence to determine whether the real-time haze concentration value is in a continuous upward trend;

[0132] If the real-time haze concentration value is in a continuous upward trend, then calculate the duration of the continuous upward trend and the cumulative increase in concentration.

[0133] Based on the duration and the cumulative increase in concentration, combined with the real-time haze concentration value itself, a comprehensive level index of haze causes is obtained by weighted calculation.

[0134] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0135] Based on the type of cause, namely haze concentration, and the comprehensive level index of haze causes, query the early warning strategy rule base;

[0136] The early warning strategy rule base pre-stores: dividing the numerical range of the comprehensive level index of haze causes into multiple continuous intervals, with each interval associated with a different early warning threshold adjustment strategy. The early warning threshold adjustment strategy includes a reduction ratio of the baseline threshold, and the reduction ratio increases as the interval to which the comprehensive level index of haze causes belongs increases.

[0137] In this embodiment, when assessing the level of the trigger, the following complex analysis is performed: First, two key data are obtained from the current environmental data: one is the real-time haze concentration value at the current time point, denoted as C_now; the other is the haze concentration monitoring value sequence arranged in chronological order from the current time point back to a specific time period, namely the third predetermined time period, such as the past 8 hours, denoted as {C1, C2, ..., Cn}.

[0138] Secondly, trend analysis is performed on the historical sequence. This uses a trend-judgment logic, such as calculating the slope of the sequence or checking the monotonicity of multiple consecutive data points, to determine whether the current real-time concentration value C_now is at the end of a continuous upward trend. In other words, it looks not only at the current value but also at whether it is the result of a sustained upward process.

[0139] If it is determined to be in a continuous upward trend, then the two characteristics of this trend are further quantified: first, the duration, that is, how long the concentration value has been rising continuously, for example, 3 hours; second, the cumulative increase in concentration, that is, the concentration difference from the starting point of the upward trend to C_now.

[0140] Next, a comprehensive index is calculated. This involves weighting the real-time haze concentration value (C_now), duration, and cumulative concentration increase according to pre-defined weighting coefficients. For example, the current concentration value is given a higher weight, while the duration and cumulative increase are given some weight. A single comprehensive index of haze-causing factors is calculated through weighted summation. This index aims to integrate risk information from three dimensions: instantaneous haze intensity, duration of exposure, and rate of concentration increase, providing a more comprehensive picture than simply using the current concentration value.

[0141] During the strategy adaptive determination phase, using haze concentration as the trigger type and the specific value of the haze trigger comprehensive level index calculated above as input, the early warning strategy rule base is queried. The rule base processes this type of continuous numerical input as follows:

[0142] The database predefines the entire possible range of values ​​for the comprehensive index of smog-causing factors, dividing it into multiple consecutive numerical intervals. For example, the index range of 0-50 is interval one, 50-100 is interval two, and above 100 is interval three.

[0143] Each numerical range is uniquely associated with a specific warning threshold adjustment strategy. These strategies all involve a core action: adjusting the baseline threshold of one or more physiological parameters by a specific downward adjustment ratio.

[0144] The key design feature is that the downward adjustment percentage is not fixed, but increases in steps as the index range increases. For example, when the index is in range one, the downward adjustment percentage of the correlation strategy might be 5%; when the index is in range two, the downward adjustment percentage increases to 10%; and when the index is in range three, the downward adjustment percentage further increases to 20%. This design ensures that the warning sensitivity can be increased proportionally to the comprehensive assessment of the haze risk level, achieving more refined adaptive adjustment.

[0145] Furthermore, the method also includes:

[0146] Based on the current physiological data, multiple physiological target inducing factors are identified, including abnormal respiratory rhythm indicators;

[0147] The step of determining multiple physiological target precipitating factors based on the current physiological data includes: assessing the current precipitating factor level of the abnormal respiratory rhythm indicators based on the current physiological data;

[0148] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: determining the corresponding early warning threshold adjustment strategy from the early warning strategy rule base based on the trigger type of the respiratory rhythm abnormality index and its current trigger level.

[0149] In this embodiment, during the trigger identification phase, current physiological data is scanned and analyzed in parallel while environmental data is being analyzed. Built-in rules identify specific abnormal physiological patterns. For example, it continuously monitors respiratory waveforms or respiratory interval sequences and calculates their regularity indicators. When an abnormal pattern in the respiratory rhythm is detected, such as alternating tachypnea and bradypnea, or significantly irregular respiratory intervals, and the degree of this abnormality exceeds a preset physiological reference baseline, the abnormal respiratory rhythm indicator is identified as a physiological target trigger. This signifies that it can not only perceive external environmental risks but also identify potential risk signals from within the patient's body.

[0150] Secondly, in the precipitating factor assessment phase, the severity of the identified respiratory rhythm abnormality indicator needs to be evaluated. The assessment process focuses on the physiological indicator itself. This may involve extracting quantitative features reflecting the degree of respiratory rhythm abnormality from current physiological data, such as calculating the standard deviation of respiratory cycle variability over a recent period or statistically analyzing the frequency of abnormal respiratory events. Based on these quantitative features, they are mapped to a predefined precipitating factor level, such as mild, moderate, or severe abnormality.

[0151] Finally, in the strategy adaptation and determination phase, the early warning strategy rule base needs to support the handling of physiological triggers. Therefore, the rule base must store entries with respiratory rhythm abnormalities as the trigger type. Based on this trigger type and the specific trigger level assessed in the previous step (e.g., moderate abnormality), a unique corresponding early warning threshold adjustment strategy is retrieved from the rule base. This strategy will also explicitly indicate how to adjust the early warning thresholds of relevant physiological parameters. For example, when a moderate respiratory rhythm abnormality is detected, the strategy might instruct to simultaneously lower the blood oxygen saturation threshold and raise the heart rate alarm threshold to address potential compensatory physiological responses.

[0152] Furthermore, this method assesses the current precipitating factor level of respiratory rhythm abnormality indicators in the following ways:

[0153] Extract respiratory waveform data for a fourth predetermined time period from the current physiological data;

[0154] Identify the duration of each respiratory cycle in the respiratory waveform data;

[0155] Calculate the coefficient of variation between the durations of consecutive respiratory cycles;

[0156] If the coefficient of variation exceeds the first physiological reference value, then the causative level of the abnormal respiratory rhythm index is determined to be abnormal.

[0157] In this embodiment, the following steps are performed for accurate evaluation: First, the most recent continuous time period, i.e., the fourth predetermined time period, is extracted from the current physiological data, such as the high-quality raw respiratory waveform data within the past 5 minutes or the respiratory event timestamp sequence obtained after preliminary processing.

[0158] Next, the respiratory waveform data is divided into periods. It uses a peak / trough detection algorithm or a zero-crossing detection method to identify one inhalation and one exhalation for each complete respiratory cycle in the waveform. For each identified respiratory cycle, its duration is precisely calculated; that is, the time interval from the start of one inhalation to the start of the next inhalation, usually in seconds. This results in a sequence consisting of the durations of consecutive respiratory cycles.

[0159] Next, the stability of this duration sequence is analyzed. The coefficient of variation (COP) of the sequence is calculated. The COP is a statistical measure of the dispersion of data; it is calculated by dividing the standard deviation of the sequence by its mean. This coefficient eliminates the influence of the mean itself and purely reflects the regularity of the breathing rhythm. A smaller COP indicates more regular breathing; a larger COP indicates a more unstable and irregular breathing rhythm.

[0160] Finally, a threshold judgment is performed. This compares the calculated respiratory cycle variability coefficient with a pre-existing first physiological reference value. This first physiological reference value is set based on typical respiratory rhythm variability data from healthy individuals or stable COPD patients, representing the upper limit of normal variability. If the calculated variability coefficient exceeds this first physiological reference value, the current respiratory rhythm is determined to be statistically significant. Based on this determination, the current precipitating factor level of the abnormal respiratory rhythm indicator is officially marked as abnormal for subsequent strategy queries.

[0161] Furthermore, the method also includes:

[0162] Reference Figure 4 The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategies to obtain the dynamic early warning threshold includes:

[0163] When there are multiple target triggers for disease onset, obtain the early warning threshold adjustment strategy determined for each target trigger.

[0164] The adjustment operation on the baseline threshold indicated by each early warning threshold adjustment strategy is analyzed, and the adjustment operation includes the adjustment direction and adjustment amount;

[0165] All adjustment operations are integrated according to preset conflict resolution rules to generate a composite adjustment command;

[0166] The baseline threshold is adjusted once according to the composite adjustment instruction to obtain a unified dynamic early warning threshold.

[0167] In this embodiment, after identifying all target triggers and their corresponding early warning threshold adjustment strategies, these strategies are collected into a temporary strategy set. For example, there may be both strategy A targeting moderate environmental temperature differences and strategy B targeting high levels of smog concentration.

[0168] Secondly, each policy in the set is parsed. Each policy is essentially an instruction, typically containing two core elements: first, the direction of adjustment—whether the baseline threshold is increased, decreased, or left unchanged; and second, the amount of adjustment—the specific value or percentage of the increase or decrease. The parsing process transforms the semantic content of the policy into explicit mathematical operational instructions.

[0169] Then, conflict resolution and integration are performed. Since multiple strategies may act on the same physiological parameter—for example, all requiring adjustment of the respiratory rate threshold—but the direction and amount of adjustment may differ, direct application could lead to contradictions. Therefore, they are processed according to pre-defined conflict resolution rules. These rules may include:

[0170] Priority rule: Pre-set priorities for different incentive types. When multiple policies issue different instructions for the same threshold, the policy instruction corresponding to the incentive with the highest priority is adopted, and the others are ignored.

[0171] Stacking rule: Adjustments in the same direction for the same threshold are allowed to be stacked. For example, if strategy A requires a 5% reduction and strategy B also requires a 3% reduction, they will be combined into an 8% reduction. However, for adjustments in opposite directions, net asset value calculation or priority-based processing will be used.

[0172] Most Sensitive Rule: In the context of security alerts, the most sensitive principle is sometimes adopted. That is, for instructions that lower the threshold to make the alert more easily triggered, the one with the largest downward adjustment is adopted; for instructions that raise the threshold to make the alert more stringent, the one with the smallest upward adjustment may be adopted to ensure that no risk is overlooked.

[0173] Based on one or a combination of the above rules, logical operations are performed on all parsed operational instructions to ultimately generate a single composite adjustment instruction. This instruction provides a clear and unambiguous adjustment direction and amount for each physiological parameter baseline threshold that needs adjustment.

[0174] Finally, based on this composite adjustment instruction, all relevant baseline thresholds are adjusted in a one-time, synchronous manner to obtain a complete and unified set of dynamic early warning thresholds for subsequent real-time data comparison.

[0175] Furthermore, the method targets disease-causing factors including the levels of biomarkers characterizing the risk of respiratory inflammation, and based on the current physiological data, identifies multiple physiological target disease-causing factors, including:

[0176] Obtain the current biomarker concentration value measured by the detection device from the current physiological data;

[0177] Obtain the baseline biomarker concentration values ​​of the patients;

[0178] Calculate the ratio of the current biomarker concentration value to the baseline biomarker concentration value;

[0179] The current precipitating factor level of the biomarker is determined based on the ratio, and the level includes at least normal level, slightly elevated level, and significantly elevated level.

[0180] The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes:

[0181] In response to determining that the current level of the trigger for the biomarker is slightly elevated, a first biomarker adjustment strategy is obtained from the warning strategy rule base. The first biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter used for warning determination is lowered by a first proportion, while keeping the baseline threshold of the second physiological parameter used for warning determination unchanged.

[0182] In response to determining that the current trigger level of the biomarker is significantly elevated, a second biomarker adjustment strategy is obtained from the warning strategy rule base. The second biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter is lowered by a second proportion and the baseline threshold of the second physiological parameter is lowered by a third proportion, wherein the second proportion is greater than the first proportion.

[0183] Target triggers include low environmental temperature. Based on the current environmental data, the current trigger level of each of the target triggers is assessed, including:

[0184] Obtain the current ambient temperature value from the current environmental data;

[0185] If the current ambient temperature value is lower than the second ambient reference value, then the duration for which the current ambient temperature value is lower than the second ambient reference value is further obtained;

[0186] Calculate a low - temperature exposure intensity index based on the extent to which the current ambient temperature value is lower than the second ambient reference value and the duration of such low - temperature exposure.

[0187] The adaptively determining, from a preset early - warning strategy rule library, multiple early - warning threshold adjustment strategies according to the types of the multiple target disease onset inducements and their respective current inducement levels includes:

[0188] Query the early - warning strategy rule library according to the inducement type of environmental low temperature and the low - temperature exposure intensity index, and obtain the corresponding early - warning threshold adjustment strategy. The early - warning threshold adjustment strategy indicates that the adjustment amount for the baseline threshold is positively correlated with the low - temperature exposure intensity index.

[0189] Among them, in this embodiment, in the inducement level assessment stage, process a specific biomarker data. For example, the concentration of exhaled nitric oxide measured by a portable or bedside detection device, or an approximate index of blood C - reactive protein. The specific steps are as follows: First, obtain the current biomarker concentration value of the patient uploaded by a dedicated detection device. At the same time, retrieve the baseline biomarker concentration value during the patient's historical stable period from the patient's personal file. This baseline value represents the typical level of the patient when not in acute exacerbation.

[0190] Next, calculate a ratio: divide the current concentration value by the baseline concentration value. This ratio is denoted as R, which intuitively reflects the multiple of the increase in the current inflammation level relative to the personal baseline. The closer the ratio is to 1, the closer it is to the baseline; the larger the ratio, the more significant the increase.

[0191] Then, perform a grading judgment according to the ratio R. There are preset grading thresholds in it, for example:

[0192] If R ≤ 1.2, it is determined to be at the normal level;

[0193] If 1.2 < R ≤ 2.0, it is determined to be at the mildly elevated level;

[0194] If R > 2.0, it is determined to be at the significantly elevated level.

[0195] Determine the current inducement level grade of the biomarker level according to the interval in which the calculated R value falls.

[0196] In the strategy adaptive determination stage, refined strategies are predefined in the early - warning strategy rule library for different grades of the inducement type of biomarker level. The key feature of these strategies is that they may make different adjustments for multiple physiological parameters, reflecting the expected differential impact of the inducement on different physiques.

[0197] When the current level is determined to be slightly elevated, it retrieves a first biomarker adjustment strategy from the rule base. This strategy specifically instructs the lowering of the first physiological parameter used for warning determination, such as the baseline threshold for respiratory rate representing airway responsiveness, by a specific first proportion, for example, 8%. Simultaneously, the strategy explicitly instructs the maintenance of the baseline threshold for the second physiological parameter, such as blood oxygen saturation representing oxygenation status. This is because mild inflammation may primarily affect breathing patterns and has not yet significantly impacted gas exchange.

[0198] When the current level is determined to be significantly elevated, it retrieves a second biomarker adjustment strategy from the rule base. This strategy is more aggressive: first, it lowers the baseline threshold for the first physiological parameter, respiratory rate, by a larger second proportion, such as 15%. Furthermore, instead of ignoring the second parameter, it instructs that the baseline threshold for the second physiological parameter, blood oxygen saturation, also be lowered by a third proportion, such as 5%. By design, the second proportion must be greater than the first proportion, reflecting a comprehensive increase in warning sensitivity when the risk escalates, and beginning to include oxygen exchange efficiency in a more stringent monitoring scope.

[0199] When assessing the level of the trigger, perform the following steps: First, read the current ambient temperature value from the current environmental data and denote it as T_now.

[0200] Next, an initial assessment is performed. The temperature value T_now is compared with a pre-stored second environmental reference value, such as 10 degrees Celsius. This is a low-temperature warning line that easily triggers bronchospasm in COPD patients. If T_now is not lower than this reference value, low ambient temperature is not considered a primary trigger. Only when T_now falls below this second environmental reference value is further in-depth evaluation triggered.

[0201] Then, the in-depth evaluation process begins. A timing monitoring logic is initiated to continuously track the current temperature below the second environmental reference value. It records the duration from when the temperature first falls below the warning line to the current moment or the most recent sampling moment that is still below the warning line; for example, it has been below 10 degrees Celsius for 4 hours.

[0202] Next, a comprehensive low-temperature exposure intensity index is calculated. This index considers two factors: first, the magnitude of the low temperature, i.e., the specific difference between the current temperature T_now and the second environmental reference value (e.g., if the current temperature is 5 degrees Celsius, then the difference is 5 degrees); and second, the duration recorded in the previous step. Following a predetermined formula, for example, the magnitude of the low temperature and the duration are multiplied or added in a weighted manner, and then combined into a single numerical index. The higher this index, the greater the intensity of the low-temperature exposure—the lower the temperature and / or the longer the duration.

[0203] Finally, in the strategy adaptive determination phase, using low ambient temperature as the trigger type and the calculated low temperature exposure intensity index as input, the early warning strategy rule base is queried. The corresponding strategies stored in the rule base contain a core instruction that adjusts a baseline threshold. A key feature is that the rule base is designed so that the magnitude of this adjustment is positively correlated with the input low temperature exposure intensity index. This means that the higher the index, the more severe the low temperature exposure, and the greater the magnitude of the threshold adjustment indicated by the strategy. For example, this can be achieved through a linear function or a piecewise lookup table: a 5% reduction when the index is 10; a 10% reduction when the index is 20; and a 15% reduction when the index is 30. This positive correlation design ensures a precise match between the early warning response intensity and the degree of low temperature risk.

[0204] Furthermore, the method also includes:

[0205] Identify whether the current physiological data contains indicators that characterize the patient's emotional tension.

[0206] If included, the level of emotional stress triggers is assessed based on the data of the indicators representing the patient's emotional stress.

[0207] Based on the type of emotional stress trigger and the assessed level of emotional stress trigger, a corresponding emotional-related early warning threshold adjustment strategy is determined from the early warning strategy rule base;

[0208] The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategy to obtain the dynamic early warning threshold includes:

[0209] The baseline threshold is adjusted by combining the emotion-related early warning threshold adjustment strategy determined for triggers of emotional stress.

[0210] In this embodiment, an additional identification task is performed when analyzing physiological data. It checks whether the current physiological data stream contains indicators that can indirectly or directly characterize the patient's emotional stress. Such data might include: the low-frequency to high-frequency power ratio of heart rate variability obtained through photoplethysmography (PPG) analysis, the skin conductance response level measured by a skin conductance sensor, or stress self-reports actively submitted by the patient through an interactive interface. The presence of such data is determined by identifying specific data labels or formats.

[0211] If data containing such emotion-related indicators is identified, an assessment of the level of emotional stress triggers is initiated. The assessment is based on the indicator data itself. For example, if the indicator is the heart rate variability ratio, its recent average is calculated and compared with the individual's resting baseline, and the level of stress is categorized according to the degree of deviation, such as calm, mild stress, or high stress. If the indicator is a self-rated score, its numerical value or level is used directly.

[0212] Next, emotional stress is treated as an independent trigger type, along with the assessed specific trigger level, such as mild stress, as a query key, retrieved from a unified warning strategy rule base. The rule base pre-stores emotion-related warning threshold adjustment strategies corresponding to different levels of emotional stress. For example, for high stress levels, the strategy might instruct a downward adjustment of the heart rate warning threshold, or an additional increase in the downward adjustment of the respiratory rate warning threshold.

[0213] Finally, in the step of adjusting the baseline threshold to generate the dynamic early warning threshold, it is necessary to comprehensively process the adjustment strategies determined for all triggers identified in this early warning cycle, including environmental triggers, physiological triggers, and newly added emotional stress triggers. This means that the emotion-related early warning threshold adjustment strategy determined for emotional stress triggers will be input as an input, along with strategies for other triggers, into the strategy integration and conflict resolution process. After this process, a final comprehensive composite adjustment instruction is generated, and the baseline threshold is adjusted accordingly. Thus, emotional factors are formally incorporated into the adaptive regulation system affecting early warning sensitivity.

[0214] According to a second embodiment of the present invention, the present invention claims protection for an early warning system for chronic obstructive pulmonary disease, comprising:

[0215] One or more processors;

[0216] A memory having stored one or more programs that, when executed by one or more processors, enable the one or more processors to implement the aforementioned method for early warning of chronic obstructive pulmonary disease.

[0217] The following is a specific example:

[0218] This embodiment is implemented in a computer system of a simulated chronic disease management center in a hospital's respiratory department. The system is connected to an environmental monitoring station deployed in the patient's home environment to monitor temperature, humidity, and PM2.5 concentration, and a multi-parameter physiological monitor worn by the patient to monitor respiratory rate, blood oxygen saturation, heart rate, and respiratory waveform. The system background runs the early warning method described in this invention, conducting continuous monitoring and early warning tests on multiple simulated chronic obstructive pulmonary disease (COPD) patients for several weeks.

[0219] Test scenario: Daily monitoring of patient A.

[0220] Data Acquisition: The system synchronizes data every five minutes. At a certain moment, it acquires a set of current environmental data from Patient A's home environmental monitoring station, including the current temperature, temperature records from the past few hours, and real-time PM2.5 concentration. Simultaneously, it acquires current physiological data from the patient's wearable device, including instantaneous respiratory rate, blood oxygen saturation, and a summary of the respiratory waveform from the past minute.

[0221] Trigger Identification: The system analyzes current environmental data. First, it calculates the average temperature over the past twelve hours and compares it with the current temperature, finding that the temperature difference exceeds a preset temperature difference concern threshold. Second, it checks the current PM2.5 concentration and finds that it exceeds the air quality concern threshold. Therefore, the system identifies ambient temperature difference and haze concentration as the two target triggers for this assessment.

[0222] Level Assessment: The system assesses the two contributing factors separately. For ambient temperature difference, it categorizes it into the medium-level temperature difference category based on the specific magnitude of the difference. For haze concentration, it not only examines the current value but also analyzes its trend over the past few hours to calculate a comprehensive haze index, which is then used to classify the haze level as high.

[0223] Adaptive Strategy Determination: The system accesses the early warning strategy rule base. Using a moderate ambient temperature difference as the condition, strategy A is found, which instructs the respiratory rate baseline threshold to be lowered by one level. Using a high level of haze concentration as the condition, strategy B is found, which instructs the respiratory rate baseline threshold to be lowered by two levels and the blood oxygen saturation baseline threshold to be lowered by one level.

[0224] Threshold Adjustment: The system obtains patient A's personalized baseline respiratory rate thresholds (e.g., respiratory rates per minute) and baseline blood oxygen saturation thresholds (e.g., blood oxygen percentage). Applying strategies A and B: for the respiratory rate threshold, the system combines the two downward adjustment magnitudes to obtain a larger downward adjustment value; for the blood oxygen saturation threshold, it lowers it by one level. After adjustment, a set of more stringent dynamic warning thresholds tailored to the current situation is obtained.

[0225] Data Comparison: The system compares patient A's real-time respiratory rate and blood oxygen saturation with the adjusted dynamic warning thresholds one by one.

[0226] Warning Generation and Output: Comparison revealed that Patient A's real-time respiratory rate exceeded the dynamically adjusted respiratory rate threshold, while blood oxygen saturation remained above the dynamic threshold. Therefore, the system determined that the respiratory rate alone triggered a warning. Subsequently, the system generated a warning message: "Patient A, at [time], due to environmental temperature differences and smog, experienced a rise in respiratory rate to the warning level. Attention is advised." This message was simultaneously pushed to the nurses' station monitoring screen in the management center and to the attending physician's mobile application.

[0227] Test Results: This process validates a complete closed loop from multi-source data input, adaptive trigger analysis, dynamic threshold adjustment to intelligent early warning output. Compared to older systems using fixed thresholds, the method of this invention triggered the early warning earlier in this scenario. Subsequent simulations of disease progression showed that this early warning provided a valuable lead time for intervention.

[0228] Second test scenario: A series of days in spring when air diffusion conditions are poor.

[0229] Execution steps: The system performs trigger identification for Patient B. It retrieves all PM2.5 concentration monitoring records for Patient B's location over a specific time period, such as six hours. The system calculates the arithmetic mean of the PM2.5 concentration over these six hours. After calculation, this average value is compared with a preset first environmental reference value for COPD patients, a concentration limit set according to health recommendations. The comparison results show that the average value significantly exceeds this reference value.

[0230] Based on the rule that the average value exceeds the reference value, the system logically determines that the current environmental smog exposure has reached a level requiring the activation of an early warning response. Therefore, the system officially identifies the smog concentration as an active target trigger for the onset of illness in Patient B and inputs it into the subsequent processing flow.

[0231] Test results: This method ensures that the identification of haze causes is not based on a peak at a certain moment, but on the average exposure level over a period of time, avoiding misjudgments caused by short-term fluctuations, and making the identification of causes more stable and reliable.

[0232] Third test scenario: The area where Patient C lives experiences a rapid cooling process.

[0233] Data Extraction and Calculation: The system obtains the current ambient temperature at patient C's location. Simultaneously, it retrieves all historical temperature data from the previous twelve hours. The system calculates the historical average temperature over these twelve hours. Next, it calculates the absolute value of the difference between the current temperature and this historical average temperature to obtain the current temperature difference range.

[0234] Level Classification: The system has pre-defined classification rules. For example, a temperature difference within a certain range is considered low-level, a larger range is considered medium-level, and exceeding a certain significant value is considered high-level. Calculations show that the current temperature drop has caused the current temperature difference to fall within the medium-level range. Therefore, the system assesses the level of the environmental temperature difference as a medium-level temperature difference.

[0235] Strategy Library Query and Adaptation: The early warning strategy rule library pre-configures refined strategies for the environmental temperature difference type. The rules explicitly state: When the temperature difference is low, strategy L is applied, indicating a slight reduction in the respiratory rate baseline threshold, for example, by a small percentage. When the temperature difference is medium, which is the condition triggered in this instance, strategy M is applied, indicating a moderate reduction in the respiratory rate baseline threshold by a larger percentage. When the temperature difference is high, strategy H is applied, indicating a significant reduction by the largest percentage.

[0236] Application Results: Based on the assessment of the intermediate-level temperature difference, the system automatically selected and applied strategy M. This lowered the dynamic warning threshold for monitoring patient C's respiratory rate to a more sensitive level.

[0237] Test Results: In subsequent monitoring, because the threshold had been adaptively lowered to accommodate moderate temperature differences, the system successfully detected a slight increase in respiratory rate in patient C caused by a sudden change in temperature and issued an alert. The control group system, using the old fixed threshold, failed to trigger this alert. This fully demonstrates the effectiveness of precise strategy mapping based on temperature difference level grading.

[0238] Fourth test scenario: The PM2.5 concentration in the area where patient D is located continues to rise within half a day.

[0239] Trend Analysis: The system not only reads the current PM2.5 concentration value but also retrieves the historical concentration series from the past eight hours. By analyzing the series, the system determines that the current concentration is in a continuous upward trend, and this trend has been maintained for some time.

[0240] Comprehensive Index Calculation: The system performs a multi-factor fusion calculation. It doesn't simply use the final concentration value, but comprehensively considers three dimensions: A. the current peak concentration; B. the duration of the concentration increase trend; and C. the total cumulative increase in concentration during the entire increase period. The system combines these three factors into a single comprehensive index of haze-causing factors according to preset weights. The level of this index represents the overall risk level after comprehensively considering concentration, duration, and rate of increase.

[0241] Continuous Interval Strategy Matching: The early warning strategy rule base has designed a continuous interval-based response mechanism for haze triggers. It pre-divides the possible value range of the comprehensive level index of haze triggers into several continuous intervals, such as green intervals, yellow intervals, orange intervals, and red intervals.

[0242] Proportional Adjustment: Each interval is associated with a specific threshold adjustment strategy, the core of which is defining a downward adjustment ratio. The rule is: the higher the level of the interval in which the index is located, the larger the associated downward adjustment ratio. For example, when the index falls into the yellow interval, the respiratory rate threshold is adjusted downward by a level one ratio; when it falls into the orange interval, the downward adjustment ratio increases to a larger level two; and when it falls into the red interval, the largest level three downward adjustment ratio is used.

[0243] Application Results: In this scenario, due to the continuous increase in concentration, the calculated comprehensive index was high, falling into the orange zone. Therefore, the system automatically applied the corresponding level two reduction ratio, significantly lowering the warning threshold.

[0244] Test Results: This evaluation method makes the system more sensitive to slowly but steadily increasing smog pollution patterns. Test data shows that, compared to methods based solely on instantaneous concentration, this invention initiates a higher level of early warning preparation in the early stages of pollution accumulation, achieving earlier risk alerts.

[0245] Fifth test scenario: Patient E's physiological data did not show abnormal environmental exposure, but characteristic changes were observed in physiological signals.

[0246] Trigger Identification: In routine analysis, the system did not identify any significant triggers from the environmental data. However, during pattern scanning of patient E's real-time respiratory waveform, the system's built-in algorithm identified an irregular wheezing prodromal pattern in the respiratory rhythm. Based on preset rules, the system classified the abnormal respiratory rhythm indicator as an independent, physiological target trigger.

[0247] Level assessment: The system performs quantitative analysis on the abnormal rhythm, calculates the degree of its irregularity, and assesses it as a moderate level of abnormality based on the calculation results.

[0248] Strategy Determination: The system queries the rule base for strategies under the category of abnormal respiratory rhythm indicators. Based on the level of moderate abnormality, the corresponding strategy is retrieved. This strategy may indicate the need to adjust multiple parameters simultaneously, such as slightly lowering the blood oxygen saturation threshold and simultaneously slightly raising the warning threshold for heart rate variability, to address potential early changes in ventilation function and autonomic nervous system responses.

[0249] Test results: This shows that the method of the present invention does not depend on changes in the external environment, and can detect early signs of risk from internal signals of the body and activate corresponding early warning strategies, thus realizing the ability to respond to endogenous causes such as infection and inflammation.

[0250] Sixth test scenario: Precisely quantify the abnormal respiratory rhythm of patient E.

[0251] The system captures the complete respiratory waveform data of patient E over the most recent five minutes. Through the signal processing module, the starting point of each respiratory cycle in the waveform is accurately identified, and the time interval between adjacent starting points is calculated, thus obtaining a sequence composed of the duration of consecutive respiratory cycles.

[0252] The system calculates the standard deviation and mean of the sequence. To eliminate the influence of individual differences in baseline respiratory rate, the system further calculates the coefficient of variation, which is the standard deviation divided by the mean. This coefficient value purely reflects the stability and regularity of the respiratory rhythm.

[0253] The system compares the calculated coefficient of variation with a preset first physiological reference value. This reference value is derived from statistical data of a large number of stable COPD patients and represents the upper limit of normal rhythm variability. The comparison revealed that patient E's current coefficient of variation significantly exceeds this reference value.

[0254] Based on the objective quantitative standard of exceeding the coefficient of variation, the system officially confirmed that patient E currently has an abnormal respiratory rhythm and marked the causative level of this indicator as the exact abnormal level.

[0255] Test results: This quantitative method provides an objective and repeatable assessment standard for respiratory rhythm abnormalities, avoiding errors in subjective judgment and ensuring the accuracy and consistency of the assessment of physiological precipitating factors.

[0256] Seventh test scenario: The patient is simultaneously facing a low temperature difference in the environment and emotional stress due to a visitor's arrival.

[0257] Strategy Collection: The system identified two triggers: a moderate level of ambient low temperature and a mild level of emotional stress. Corresponding strategies were retrieved from the rule base: Strategy X lowers the respiratory rate threshold by A units in response to low temperature; Strategy Y lowers the respiratory rate threshold by B units in response to emotional stress, and simultaneously lowers the heart rate threshold by C units.

[0258] Command parsing and integration: The system detects that both strategies X and Y have issued adjustment commands to the respiratory rate threshold, and the direction is the same—a downward adjustment. According to the preset overlay rules, the system adds the downward adjustment magnitudes A and B to obtain the total downward adjustment command for respiratory rate, A+B. The command to lower the heart rate threshold by C units, unique to strategy Y, is directly adopted.

[0259] Generate a compound instruction: Finally, the system generates a compound adjustment instruction: 1) Lower the personal baseline threshold for respiratory rate by A+B units; 2) Lower the personal baseline threshold for heart rate by C units.

[0260] Unified Adjustment: Based on this composite instruction, the system makes a one-time, synchronous numerical modification to the baseline thresholds of the patient's two physiological parameters, generating a unified set of dynamic early warning thresholds.

[0261] Test Results: This mechanism effectively solves the problem of strategy coordination when multiple factors coexist. Tests show that by integrating rules, the dynamic threshold generated by the system can comprehensively reflect the cumulative effect of multiple risk factors, making the early warning behavior more reasonable and avoiding situations where other risks are ignored due to simply choosing one strategy.

[0262] Eighth test scenario: Patient G uses a portable inflammatory marker detector connected to the system for regular testing.

[0263] Data Acquisition and Ratio Calculation: The system receives the latest inflammatory marker test values ​​uploaded by the patient. The system retrieves the patient's baseline values ​​from their personal historical stable period from their records. The ratio R between the current value and the baseline value is calculated.

[0264] Level Classification: According to the preset classification criteria: if the R value is below a certain lower threshold, it is considered to be at a normal level; if the R value exceeds the lower threshold but does not reach a higher threshold, it is considered to be at a slightly elevated level; if the R value exceeds a higher threshold, it is considered to be at a significantly elevated level. The R value detected in this study fell within the significantly elevated level range.

[0265] Scenario Comparison A: If the elevation is mild: The system will apply the first biomarker strategy. This strategy only targets the first physiological parameter, such as respiratory rate, instructing its baseline threshold to be lowered by a relatively small first percentage, such as a few percent, while leaving the threshold for the second physiological parameter, such as blood oxygen saturation, unchanged. This reflects that mild inflammation may primarily affect ventilation function.

[0266] Scenario B shows a significant increase: The system applies a second biomarker strategy. This strategy comprises two parts: First, the baseline threshold for the first physiological parameter, respiratory rate, is lowered by a larger second proportion than the first. Second, the baseline threshold for the second physiological parameter, blood oxygen saturation, is also lowered by a third proportion. This reflects that significant inflammation may have simultaneously affected ventilation and oxygenation, necessitating a comprehensive improvement in monitoring sensitivity.

[0267] Application Results: Due to the significantly elevated level of patient G in this test, the system implemented strategy B, and at the same time significantly lowered the warning thresholds for respiratory rate and blood oxygen saturation.

[0268] Test Results: In subsequent simulations, patient G experienced a slow decrease in blood oxygen saturation. Because strategy B had been pre-applied to lower the blood oxygen threshold, the system issued an earlier warning about the decline in oxygenation than strategies using only a fixed threshold or adjusting only the respiratory rate. This demonstrates the necessity and superiority of implementing differentiated, multi-parameter adjustment strategies for different levels of the same trigger.

[0269] Ninth Test Scenario: Patient Xin was exposed to a low-temperature indoor environment for an extended period during a winter night. The system detected that the current indoor temperature was below the second environmental reference value, such as the low-temperature warning line. The system then initiated a deep assessment, which not only recorded the current low-temperature value but also began tracking the duration of this low-temperature exposure, i.e., how long the temperature had been below the warning line.

[0270] The system calculates a low-temperature exposure intensity index. This index calculation incorporates two key factors: the specific temperature difference between the current temperature and the warning threshold, and the duration of exposure. For example, the index is a weighted function of temperature difference × duration. This means that four hours of exposure at 5 degrees Celsius may produce different index values ​​than two hours of exposure at 0 degrees Celsius, thus differentiating the degree of risk.

[0271] The system queries the rule base using this index value. The rule base is designed so that the output adjustment strategy, such as the threshold reduction magnitude, is positively correlated with this index value. That is, a higher index value means a colder environment and / or longer exposure, and the strategy requires a larger threshold reduction magnitude.

[0272] Test Results: Through this positive correlation design, the system's warning response intensity for patient Xin was precisely matched to her actual exposure burden to low temperatures. During the test, for short-term, mild hypothermia, the system made only minor adjustments; for this prolonged and significant hypothermia exposure, the system applied a strategy of drastically lowering the threshold, achieving precise adaptive warnings.

[0273] Tenth test scenario: Patient Ren feels nervous while performing rehabilitation exercises.

[0274] Emotional Data Identification and Assessment: The system identifies heart rate variability data from the patient's physiological data stream, which is labeled as potentially useful for assessing autonomic nervous system tension. Analysis reveals a specific pattern of reduced low-frequency to high-frequency power ratio in the heart rate variability data, which the system assesses as a moderate level of emotional stress triggers based on built-in rules.

[0275] Emotional Strategy Retrieval: The system retrieves corresponding emotion-related early warning threshold adjustment strategies from the early warning strategy rule base, using "moderate" as the criteria. This strategy instructs the heart rate early warning threshold to be lowered by a specific amount.

[0276] Comprehensive Integration: Assuming the presence of mild smog as an environmental trigger, the system sends the strategy of lowering the heart rate threshold for emotional stress and the strategy of lowering the respiratory rate threshold for mild smog, together with these, to the strategy integration engine as described in Example 7. After integration according to preset rules, a final composite adjustment instruction is generated: simultaneously lowering both the heart rate and respiratory rate thresholds, but the magnitude of the reduction is rationally calculated based on the priority of the trigger and the superposition rules.

[0277] Unified Adjustment and Application: The system adjusts the baseline threshold based on this final instruction. During subsequent exercise monitoring, the patient's heart rate increased due to tension. Since the heart rate threshold had been adaptively lowered to accommodate the patient's emotional state, the system promptly issued a warning prompting the patient to relax or rest.

[0278] Test Results: This embodiment demonstrates how psychological factors can be organically integrated into the adaptive early warning framework of this invention. The system no longer focuses solely on the physical and chemical environment, but also considers the patient's psychophysiological state as an important risk moderating variable, making the early warning more aligned with the patient's overall physical and mental condition, embodying a truly individualized and contextualized early warning concept.

[0279] Through the aforementioned series of interconnected and logically rigorous test scenarios and execution steps, this specific embodiment demonstrates in detail the implementation and operational effects of the technical solution protected by this invention in a real-world system. The embodiment covers the entire process from external environment to internal physiology, from single triggers to combinations of multiple triggers, and from identification and assessment to strategy adjustment, fully demonstrating the effectiveness and ingenuity of this method in improving the accuracy, timeliness, and individualization of early warning for chronic obstructive pulmonary disease (COPD). All steps are completed through logical judgment and processing of data by a computer system, and do not involve disease diagnosis or treatment methods, thus complying with relevant patent regulations.

[0280] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0281] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0282] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for early warning of chronic obstructive pulmonary disease, characterized in that, Applied to a computer system, the method includes: S1, acquire the patient's current environmental data and current physiological data, and determine multiple target triggers for the onset of the disease based on the current environmental data; S2, based on the current environmental data and the current physiological data, assess the current precipitating factor level of each of the multiple target disease-inducing factors; S3, based on the type and current level of each of the multiple target disease triggers, adaptively determine multiple warning threshold adjustment strategies from a preset warning strategy rule base. The warning strategy rule base defines the mapping relationship between different trigger types and different trigger level ranges and warning threshold adjustment strategies. The warning threshold adjustment strategies are used to dynamically adjust the warning judgment threshold. S4, according to the adaptively determined multiple early warning threshold adjustment strategies, adjust the benchmark threshold used for early warning determination to obtain the dynamic early warning threshold; S5, compare the current physiological data with the dynamic early warning threshold, and when the current physiological data meets the early warning triggering condition based on the dynamic early warning threshold, generate and output early warning information for the patient's chronic obstructive pulmonary disease; When the trigger type is ambient temperature difference and the trigger level is low temperature difference, a first adjustment strategy is associated, and the first adjustment strategy indicates that the benchmark threshold is lowered by a first amount. When the trigger type is ambient temperature difference and the trigger level is medium-level temperature difference, a second adjustment strategy is associated, which indicates that the benchmark threshold is lowered by a second magnitude, which is greater than the first magnitude. When the trigger type is environmental temperature difference and the trigger level is high-level temperature difference, a third adjustment strategy is associated, which indicates that the benchmark threshold is lowered by a third magnitude, which is greater than the second magnitude. In the stage of assessing the level of triggers, perform the following detailed steps: The environmental temperature value at the current moment or the most recent sampling moment is accurately read from the current environmental data and recorded as T_current. All environmental temperature data recorded at fixed intervals from the historical environmental database are retrieved and extracted from the historical environmental database, which traces back a specific time period from the current moment, i.e., the second predetermined time period, to form a historical temperature sequence. The historical temperature series is processed, and the arithmetic mean of all temperature values ​​in the series is calculated to obtain the average ambient temperature value during this period, denoted as T_average. This average value represents the recent background temperature level of the environment. Calculate key indicators and find the absolute value of the difference between the current temperature value T_current and the recent average temperature T_average, i.e., |T_current - T_average|. This absolute value is defined as the current temperature difference amplitude, which quantifies the degree of deviation of the instantaneous temperature from the recent background. The temperature difference range is classified into levels. A pre-set temperature difference classification table is used to divide the numerical range of the temperature difference range into several non-overlapping intervals. Each interval corresponds to a predefined level. Based on the calculated specific value of the current temperature difference range, it is assigned to the corresponding level. In the strategy adaptive determination phase, the early warning strategy rule base is queried based on the determined trigger type, ambient temperature difference, and specific level. The rule base pre-associates early warning threshold adjustment strategies for each level under the type of ambient temperature difference. The mapping relationships in the rule base are specifically defined as follows: When the query conditions are: trigger type = ambient temperature difference, trigger level = low temperature difference, the first adjustment strategy of the association will explicitly state: The baseline threshold of one or more physiological parameters used for early warning determination will be lowered. When the query conditions are: trigger type = ambient temperature difference, trigger level = medium-level temperature difference, the associated second adjustment strategy will clearly indicate: lower the baseline threshold of the same physiological parameter by a second magnitude. When the query conditions are: trigger type = ambient temperature difference, trigger level = high-level temperature difference, the associated third adjustment strategy will clearly indicate: lower the baseline threshold by a third amount.

2. The method according to claim 1, characterized in that, Also includes: The target triggering factors include at least one of smog concentration, ambient temperature difference, and low ambient temperature. The determination of multiple target triggering factors based on the current environmental data includes: Extract the haze concentration monitoring value for a continuous first predetermined time period from the current environmental data; Calculate the average haze concentration over the first predetermined time period; If the average concentration of haze exceeds the first environmental reference value, then the haze concentration is determined as the target inducing factor for the disease.

3. The method according to claim 2, characterized in that, Also includes: The target triggers for the disease include smog concentration; Based on the current environmental data, assessing the current precipitating factor level of each of the target disease-causing factors includes: Obtain the real-time haze concentration value at the current time point from the current environmental data, and obtain the haze concentration change sequence within the third predetermined time period before the current time point; Analyze the haze concentration change sequence to determine whether the real-time haze concentration value is in a continuous upward trend; If the real-time haze concentration value is in a continuous upward trend, then calculate the duration of the continuous upward trend and the cumulative increase in concentration. Based on the duration and the cumulative increase in concentration, combined with the real-time haze concentration value itself, a comprehensive level index of haze causes is obtained by weighted calculation. The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: Based on the type of cause, namely haze concentration, and the comprehensive level index of haze causes, query the early warning strategy rule base; The early warning strategy rule base pre-stores: dividing the numerical range of the comprehensive level index of haze causes into multiple continuous intervals, with each interval associated with a different early warning threshold adjustment strategy. The early warning threshold adjustment strategy includes a reduction ratio of the baseline threshold, and the reduction ratio increases as the interval to which the comprehensive level index of haze causes belongs increases.

4. The method according to claim 2, characterized in that, The method further includes: Based on the current physiological data, multiple physiological target inducing factors are identified, including abnormal respiratory rhythm indicators; The step of determining multiple physiological target precipitating factors based on the current physiological data includes: assessing the current precipitating factor level of the abnormal respiratory rhythm indicators based on the current physiological data; The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: determining the corresponding early warning threshold adjustment strategy from the early warning strategy rule base based on the trigger type of the respiratory rhythm abnormality index and its current trigger level.

5. The method according to claim 2, characterized in that, The current precipitating factor level of abnormal respiratory rhythm indicators is assessed using the following methods: Extract respiratory waveform data for a fourth predetermined time period from the current physiological data; Identify the duration of each respiratory cycle in the respiratory waveform data; Calculate the coefficient of variation between the durations of consecutive respiratory cycles; If the coefficient of variation exceeds the first physiological reference value, then the causative level of the abnormal respiratory rhythm index is determined to be abnormal.

6. The method according to claim 2, characterized in that, Also includes: The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategy to obtain the dynamic early warning threshold includes: When there are multiple target triggers for disease onset, obtain the early warning threshold adjustment strategy determined for each target trigger. The adjustment operation on the baseline threshold indicated by each early warning threshold adjustment strategy is analyzed, and the adjustment operation includes the adjustment direction and adjustment amount; All adjustment operations are integrated according to preset conflict resolution rules to generate a composite adjustment command; The baseline threshold is adjusted once according to the composite adjustment instruction to obtain a unified dynamic early warning threshold.

7. The method according to claim 5, characterized in that, Target triggers include the levels of biomarkers characterizing the risk of respiratory inflammation. Based on the current physiological data, several physiological target triggers have been identified, including: Obtain the current biomarker concentration value measured by the detection device from the current physiological data; Obtain the baseline biomarker concentration values ​​of the patients; Calculate the ratio of the current biomarker concentration value to the baseline biomarker concentration value; The current precipitating factor level of the biomarker is determined based on the ratio, and the level includes at least normal level, slightly elevated level, and significantly elevated level. The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: In response to determining that the current level of the trigger for the biomarker is slightly elevated, a first biomarker adjustment strategy is obtained from the warning strategy rule base. The first biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter used for warning determination is lowered by a first proportion, while keeping the baseline threshold of the second physiological parameter used for warning determination unchanged. In response to determining that the current level of the trigger for the biomarker is significantly elevated, a second biomarker adjustment strategy is obtained from the warning strategy rule base. The second biomarker adjustment strategy indicates that the baseline threshold of the first physiological parameter is lowered by a second proportion and the baseline threshold of the second physiological parameter is lowered by a third proportion, wherein the second proportion is greater than the first proportion. Target triggers include low environmental temperature. Based on the current environmental data, the current trigger level of each of the target triggers is assessed, including: Obtain the current ambient temperature value from the current environmental data; If the current ambient temperature value is lower than the second ambient reference value, then the duration for which the current ambient temperature value is lower than the second ambient reference value is further obtained; Based on the extent to which the current ambient temperature is lower than the second environmental reference value and the duration of the difference, the low temperature exposure intensity index is calculated. The step of adaptively determining multiple early warning threshold adjustment strategies from a preset early warning strategy rule base based on the type and current level of each of the multiple target disease triggers includes: Based on the type of cause, namely low environmental temperature, and the low temperature exposure intensity index, the early warning strategy rule base is queried to obtain the corresponding early warning threshold adjustment strategy. The early warning threshold adjustment strategy indicates that the adjustment amount of the benchmark threshold is positively correlated with the low temperature exposure intensity index.

8. The method according to claim 2, characterized in that, The method further includes: Identify whether the current physiological data contains indicators that characterize the patient's emotional tension. If included, the level of emotional stress triggers is assessed based on the data of the indicators representing the patient's emotional stress. Based on the type of emotional stress trigger and the assessed level of emotional stress trigger, a corresponding emotional-related early warning threshold adjustment strategy is determined from the early warning strategy rule base; The step of adjusting the baseline threshold used for early warning determination according to the adaptively determined multiple early warning threshold adjustment strategies to obtain the dynamic early warning threshold includes: combining the emotion-related early warning threshold adjustment strategies determined as emotional tension triggers to comprehensively adjust the baseline threshold.

9. An early warning system for chronic obstructive pulmonary disease, characterized in that, include: One or more processors; A memory having stored one or more programs that, when executed by one or more processors, cause the one or more processors to implement a method for early warning of chronic obstructive pulmonary disease according to any one of claims 1 to 8.

Images