A clinical indicator-based illness assessment method and system

Abstract

This invention relates to the field of disease assessment technology, providing a method and system for disease assessment based on clinical indicators. The method includes assessing disease based on routine clinical vital signs parameters, defining normal, intermediate, and abnormal ranges for each parameter at the corresponding physiological stage based on the patient's age, establishing an abnormality judgment benchmark suitable for the individual, and avoiding the adaptation bias of general standards. Time-series data is decomposed into continuous, equal-duration minimum fluctuation units. Fluctuation characteristics are extracted and abnormal trigger units falling into the intermediate range are marked, accurately capturing early disease anomalies and overcoming the lag limitations of traditional monitoring. Abnormal trigger units are chained together according to time rules to form a transmission relationship chain. The time lag difference is calculated and linked to treatment plans to generate an emergency treatment library, reconstructing the disease development path and establishing standardized treatment comparison criteria. Real-time monitoring enables rapid alarms for critical situations, and early risk classification warnings and precise matching of treatment plans for early anomalies, improving the timeliness of disease assessment.

Description

Technical Field

[0001] This invention relates to the field of disease assessment technology, and more specifically, to a disease assessment method and system based on clinical indicators. Background Technology

[0002] The content in this section only provides background information related to this invention and may not constitute prior art.

[0003] In clinical intensive care and perioperative management, continuous monitoring of patients' vital signs is a core means of assessing changes in their condition. Clinicians rely on the values ​​of patients' vital signs to identify potential physiological deterioration. However, existing monitoring systems still heavily depend on threshold-driven static rules in their data processing and alerting logic; that is, an alarm is triggered when a parameter exceeds a preset normal range. This model has revealed significant technical limitations in clinical practice.

[0004] In existing technologies, threshold alarms typically treat vital signs as independent discrete variables, neglecting the coupling relationships and dynamic evolution patterns between multiple parameters over time. The compensatory mechanisms of physiological systems often manifest as sequential, subtle changes in multiple indicators, rather than a sudden collapse of a single indicator. Static single-parameter thresholds cannot capture the early, insidious deterioration stage characterized by coordinated fluctuations between parameters.

[0005] Meanwhile, alarm systems widely used in clinical practice generally lack the ability to perform structured analysis of temporal fluctuation patterns. Even when generating large amounts of continuous monitoring data, they can only identify instantaneous out-of-limit events and cannot extract the transmission relationships and cascade effects between fluctuations from the data stream. This lack of analytical capability causes the system to overlook slowly drifting but clearly deteriorating fluctuation combinations, only issuing an alarm when the condition progresses to significant decompensation, thus missing the window of opportunity for early intervention.

[0006] Therefore, there is an urgent need for a disease assessment method and system based on clinical indicators to accurately match early warnings and treatment plans for early abnormalities, thereby improving the timeliness of disease assessment. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for assessing disease conditions based on clinical indicators, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows: Firstly, this application provides a disease assessment method based on clinical indicators, including: Target parameters were obtained from multiple patients with the same case, including heart rate, respiratory rate, body temperature, and blood oxygen saturation. Based on the patient's age, normal, intermediate, and abnormal ranges for each target parameter under the corresponding physiological stage were defined. The full-time data of each target parameter is decomposed into continuous minimum fluctuation units of equal duration. The fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time are extracted as unit features. The minimum fluctuation unit whose amplitude falls into the middle interval is marked as the abnormal trigger unit. In chronological order, it is determined whether the time interval between two adjacent abnormal triggering units does not exceed the preset time. If it does not exceed the preset time, the two abnormal triggering units are connected in series to form a continuous transmission chain. Otherwise, the current transmission chain is terminated, and a new transmission chain is reconstructed with the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. For two adjacent abnormal triggering units in the transmission relationship chain, calculate the time delay difference between the previous abnormal triggering unit and the next abnormal triggering unit; bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency treatment library; The system acquires the time-series data of the current patient's target parameters. If any time-series data falls into an abnormal range, an alarm is issued directly. Otherwise, the time-series data is processed according to S102 to obtain the current patient's abnormal trigger units. If the number of abnormal trigger units is 1, continuous monitoring is performed. If the number is greater than 1, the time interval between two adjacent abnormal trigger units is calculated sequentially. If all adjacent time intervals exceed the preset duration, a minor abnormality prompt is sent to the terminal. If at least one set of adjacent abnormal trigger units has a time interval that does not exceed the preset duration, an alarm is triggered, and the set of abnormal trigger units is connected in series to form a real-time transmission chain. The corresponding time delay difference value in the real-time transmission chain is extracted, and the real-time transmission chain is matched with the corresponding time delay difference value in the emergency processing library. The successfully matched treatment plan is displayed on the terminal.

[0008] Furthermore, the normal, intermediate, and abnormal ranges for each target parameter under the corresponding physiological stage are defined, specifically including: Based on the physiological stage corresponding to the patient's age, the clinically known normal value range of each target parameter is obtained and determined as the normal range. All values ​​outside the normal range are defined as the abnormal range. For each target parameter, the historical monitoring records of multiple patients with the same case were reviewed, and the continuous monitoring curve of the parameter was extracted. The deviation segment where the parameter value changed continuously in one direction from the normal range and eventually entered the abnormal range was identified. In the deviation segment, the turning point where the trend first reversed and began to deviate continuously from the normal range was determined as the boundary point of the range. Calculate the mean of all the inner boundary points of the divergent segments, and use the mean to determine the inner boundary of the normal interval; the area between the inner boundary and the boundary of the normal interval closest to the inner boundary is determined as the middle interval.

[0009] Furthermore, the length of the smallest fluctuation unit of equal duration is set according to the following rules: Based on the physiological characteristics corresponding to the patient's age, the normal fluctuation frequency range of each parameter in the target parameters is determined. The duration required for the complete fluctuation cycle of the parameter with the lowest fluctuation frequency is taken as the benchmark. This benchmark is multiplied by a preset amplification factor to obtain the minimum fluctuation unit length, so that each unit covers at least one complete physiological fluctuation waveform.

[0010] Furthermore, the full-time data of each target parameter is decomposed into the smallest fluctuation units of continuous equal duration, specifically including: The time series waveform of the parameter with the lowest normal fluctuation frequency among the target parameters is identified. The inflection point identification algorithm is used to detect the local turning point where the time series waveform changes from continuous decline to continuous rise. The local turning point that appears first on the time axis of the data sequence is selected as the cutting starting point. Starting from this cutting starting point, the full time series data is continuously divided according to the length of the smallest fluctuation unit, so that each smallest fluctuation unit contains a complete waveform fluctuation segment.

[0011] Furthermore, the time delay difference between the previous and subsequent anomaly triggering units is calculated, specifically including: For two adjacent abnormal triggering units in the transmission relationship chain, extract the time series data of the target parameter corresponding to the previous abnormal triggering unit in that unit, identify the moment when the parameter amplitude first enters the corresponding intermediate interval in the time series data, and record it as the start time of the preceding abnormality. Extract the time series data of the target parameter corresponding to the next abnormal triggering unit within the unit, identify the moment when the parameter amplitude first enters the corresponding intermediate interval in the time series data, and record it as the start time of the subsequent abnormality. Calculate the time difference between the start time of the subsequent anomaly and the start time of the preceding anomaly, and determine this time difference as the time delay difference corresponding to the adjacent anomaly triggering units in this group.

[0012] Furthermore, the successfully matched treatment plans will be displayed on the terminal, specifically including: If the real-time transmission chain and its corresponding time delay difference successfully match any continuous local interval of any transmission relationship chain in the emergency treatment database and its bound corresponding time delay difference, an alarm is sent to the terminal and the corresponding treatment plan is displayed simultaneously.

[0013] Furthermore, after displaying the successfully matched treatment plan on the terminal, it also includes: Receive confirmation instructions or adjustments to the actual treatment plan submitted by the doctor through the terminal; if the actual treatment plan differs from the displayed plan, the actual treatment plan shall prevail. The real-time transmission chain, the calculated time delay differences, and the finally confirmed actual treatment plan are used as incremental learning samples and added to the emergency response library.

[0014] Secondly, this application also provides a disease assessment system based on clinical indicators, including: The interval definition module is used to obtain target parameters for multiple patients with the same case. The target parameters include heart rate, respiratory rate, body temperature and blood oxygen saturation. Based on the patient's age, the normal interval, intermediate interval and abnormal interval of each target parameter under the corresponding physiological stage are defined. The unit marking module is used to decompose the full-time data of each target parameter into continuous equal-duration minimum fluctuation units, extract the fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time as unit features; and mark the minimum fluctuation unit whose amplitude falls into the middle interval as an abnormal trigger unit. The relationship chain building module is used to determine whether the time interval between two adjacent abnormal triggering units does not exceed the preset time in chronological order. If it does not exceed the preset time, the two abnormal triggering units are connected in series to form a continuous transmission relationship chain. Otherwise, the current transmission relationship chain is terminated, and a new transmission relationship chain is reconstructed with the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. The emergency database construction module is used to calculate the time delay difference between two adjacent abnormal triggering units in the transmission relationship chain; and to bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency processing database. The evaluation and matching module is used to acquire the time-series data of the current patient's target parameters. If any time-series data falls into the abnormal range, an alarm is issued directly. Otherwise, the time-series data is processed according to S102 to obtain the current patient's abnormal trigger units. If the number of abnormal trigger units is 1, continuous monitoring is performed. If the number is greater than 1, the time interval between two adjacent abnormal trigger units is calculated sequentially. If all adjacent time intervals exceed the preset duration, a minor abnormality prompt is sent to the terminal. If there is at least one set of adjacent abnormal trigger units whose time intervals do not exceed the preset duration, an alarm is triggered and the set of abnormal trigger units is connected in series to form a real-time transmission chain. The corresponding time delay difference value in the real-time transmission chain is extracted, and the real-time transmission chain is matched with the corresponding time delay difference value in the emergency processing library. The successfully matched treatment plan is displayed on the terminal.

[0015] Thirdly, this application also provides an electronic device, including: Memory, used to store computer programs; A processor for implementing the method steps as described in the first aspect when executing the computer program.

[0016] Fourthly, this application also provides a readable storage medium storing a computer program that, when executed by a processor, implements the method steps of the first aspect.

[0017] The beneficial effects of this invention are as follows: This invention assesses a patient's condition based on readily available clinical vital signs parameters. First, it defines graded physiological intervals for each parameter within the corresponding physiological stage, based on the patient's age. This establishes an abnormality judgment standard tailored to the individual physiological characteristics of each patient, avoiding individual adaptation biases caused by general judgment rules and providing a core benchmark for the stratified identification of progressive abnormalities. Then, the full-time data of each parameter is decomposed into continuous, equal-duration minimum fluctuation units. The fluctuation amplitude and corresponding time within each minimum fluctuation unit are extracted as unit features. Simultaneously, minimum fluctuation units whose amplitude falls within the transition range are marked as abnormality trigger units. This accurately captures early anomalies in vital signs parameters that have deviated from the normal range but have not yet entered a clearly abnormal state, overcoming the lag limitation of traditional monitoring modes that only alarm for clearly abnormal values. Based on this, the time interval between adjacent abnormal triggering units is sequentially determined according to the time sequence to see if it exceeds the preset duration. Abnormal triggering units that meet the interval requirements are sequentially connected to form a continuous transmission chain. If the interval requirement is exceeded, the current chain is terminated and the chain is rebuilt starting from a new abnormal triggering unit. At the same time, the time delay difference between adjacent abnormal triggering units in the transmission chain is calculated. The complete transmission chain, the corresponding time delay difference, and the appropriate treatment plan are bound to generate an emergency treatment library. This not only fully restores the gradual development path of the abnormal condition through the transmission chain, but also provides a standardized and traceable reference for rapid matching and treatment in subsequent real-time monitoring. During real-time patient monitoring, an alarm is immediately triggered if any vital sign time-series data of the current patient falls into a clearly abnormal range, enabling rapid response to critical situations. For time-series data that does not fall into a clearly abnormal range, the same unit decomposition and feature extraction process is used to obtain the current patient's abnormal trigger unit. If only a single abnormal trigger unit exists, monitoring continues; if multiple abnormal trigger units exist, the time intervals of adjacent units are judged sequentially. If all adjacent intervals exceed the preset duration, a minor abnormality alert is sent; if adjacent units that meet the interval requirements exist, an alarm is triggered and connected in series to form a real-time transmission chain. The time delay difference corresponding to the real-time transmission chain is extracted and matched with an emergency treatment database. The successfully matched treatment plan is then simultaneously displayed to the terminal. This not only provides early warning of the risk of disease deterioration but also provides accurate and practical diagnostic and treatment references for clinical management, improving the timeliness, accuracy, and clinical applicability of disease assessment. Attached Figure Description

[0018] Figure 1 A flowchart of a disease assessment method based on clinical indicators provided by the present invention; Figure 2 A schematic diagram of a disease assessment system based on clinical indicators provided by the present invention; Figure 3 This is a schematic diagram of an electronic device provided by the present invention.

[0019] In the diagram: 201, Interval delineation module; 202, Unit marking module; 203, Relationship chain construction module; 204, Emergency database construction module; 205, Evaluation and matching module; 301, Processor; 302, Memory. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0021] like Figure 1 As shown in the embodiment of the present invention, a disease assessment method based on clinical indicators includes: S101: Obtain target parameters from multiple patients with the same case, including heart rate, respiratory rate, body temperature, and blood oxygen saturation; based on the patient's age, define the normal range, intermediate range, and abnormal range for each target parameter under the corresponding physiological stage.

[0022] Specifically, this invention selects four clinically universal, continuously monitorable, and non-invasive core vital signs as target parameters, covering the core states of circulation, respiration, and systemic metabolism to ensure comprehensiveness of the assessment dimensions. Simultaneously, based on patient age, it clinically categorizes patients into physiological stages such as neonatal, infancy, adulthood, and old age, matching the corresponding stage's baseline vital signs to avoid misinterpretation of normal values ​​across age groups, thus improving the clinical accuracy of interval definition from the outset. Specifically, defining the normal, intermediate, and abnormal intervals for each target parameter within the corresponding physiological stage includes: First, based on the physiological stage corresponding to the patient's age, the clinically known normal value ranges for each target parameter are obtained and defined as the normal range. All values ​​outside the normal range are defined as abnormal ranges. Here, the clinically known normal value ranges adopt recognized reference values ​​from medical data. Taking adult patients (18-65 years old) as an example, their clinically normal heart rate range is 60-100 beats / min, which is defined as the normal range for heart rate in this stage. All values ​​less than 60 beats / min or greater than 100 beats / min are defined as abnormal ranges. Directly using authoritative standards to define the ranges ensures the clinical compliance of the method, avoids the diagnostic and treatment risks of custom-defined ranges, and provides a recognized threshold for direct alarm of critical conditions.

[0023] Secondly, for each target parameter, the historical monitoring records of multiple patients with the same case were reviewed, and the continuous monitoring curve of the parameter was extracted. The deviation segment where the parameter value changed continuously in one direction from the normal range and eventually entered the abnormal range was identified. In the deviation segment, the turning point where the trend first reversed and began to deviate continuously from the normal range was determined as the boundary point of the range. Taking the same case of severe pneumonia as an example, we reviewed the continuous monitoring data of 100 such patients in the 72 hours before their condition deteriorated. For the continuous monitoring curve of the heart rate of a single patient, we identified the complete curve segment in which the heart rate started from the baseline value in the normal range and continued to rise in a unidirectional direction, eventually entering the abnormal range. This segment is the deviation segment. In this segment, we used the first-order difference trend identification method to determine the trend. When the difference results of three consecutive equally spaced monitoring cycles remain in the same direction, it is determined to be a continuous unidirectional change. The starting point of this continuous unidirectional change is the trend turning point, which is the boundary point of the interval. The core principle is that before the condition deteriorates, vital signs will first show a trend deviation rather than an instantaneous jump. Identifying this boundary point can capture early abnormal signals that traditional threshold alarms cannot detect, thus solving the pain point of delayed clinical early warning.

[0024] The formula for identifying continuous unidirectional changes using first-order difference is as follows: First, subtract the previous value from the current value. The sign of the difference directly represents the direction of change (positive = increasing, negative = decreasing), specifically: (1) In the formula, For the first patient no. Term parameter at time The first-order difference is used to characterize the changing trend of the parameter; For the first Patient 1 Term parameter at time The monitored values; This refers to the monitoring value at the previous sampling time. This represents the time interval between adjacent sampling points.

[0025] If three consecutive periodic differences maintain the same direction, it means that the signs of the consecutive differences are completely consistent (all positive / all negative). Mathematically, the product of multiple numbers with the same sign must be greater than 0. Therefore, the product formula is used to determine whether they are in the same direction, specifically: (2) In the formula, This represents the number of consecutive monitoring cycles. This is the loop variable for the multiplication operation; For the first Example: The starting point of a patient's parameters undergoing a continuous unidirectional change.

[0026] Finally, the mean of all the inner boundary points of the divergent segments is calculated, and the inner boundary of the normal interval is determined by the mean. The area between the inner boundary and the boundary of the normal interval that is closer to the inner boundary is determined as the middle interval. In the specific calculation, outliers in the inner boundary point data are first removed using the Grubbs criterion. Then, the arithmetic mean of the remaining effective inner boundary points is calculated. For example, in the heart rate data of the 100 patients with severe pneumonia mentioned above, after removing two outliers, the mean of the 98 effective inner boundary points is 94 beats / min. This mean is the inner boundary of the normal range. In the normal range of 60-100 beats / min, the inner boundary of 94 beats / min is close to the upper boundary of 100 beats / min. Therefore, the region of 94-100 beats / min is determined as the middle range. Calculating the inner boundary using the mean of a large sample can effectively eliminate the bias caused by individual differences and ensure the clinical universality of the inner boundary. The setting of the middle range fills the early warning gap between the normal and abnormal ranges. It can identify the potential risk of disease deterioration before the parameters reach the clinical abnormal threshold, greatly improving the advance and sensitivity of disease warning.

[0027] The outlier removal from the inner boundary point data using the Grubbs criterion is calculated as follows: The first step is to calculate the arithmetic mean of the sample, which is used to characterize the central tendency of the sample. The formula for the mean is: (3) In the formula, For the first The arithmetic mean of all inner boundary point samples for the term parameter; Total patient sample size; For the first Patient 1 The interval boundary points of the term parameters; This refers to the patient sample number.

[0028] The second step is to calculate the unbiased standard deviation of the sample, which measures the dispersion of the sample. The formula is: (4) In the formula, For the first The unbiased standard deviation of the inner boundary point samples of the parameter measures the degree of dispersion of the sample.

[0029] The third step is to calculate the Grubbs' statistic for each sample, which is the absolute deviation of the sample from the mean divided by the standard deviation. When the Grubbs' statistic is greater than the clinically accepted critical value, it is considered an outlier and is removed. The formula is: (5) In the formula, For the first The Grubbs statistic for each inner boundary point sample is used to identify outliers.

[0030] After removing outliers from the inner boundary data, the formula for calculating the inner boundary value is: (6) In the formula, Physiological stage Next, the The inner boundary value of the normal interval of the parameter, that is, the arithmetic mean of the effective inner boundary points; This represents the effective inner boundary point sample size after removing outliers. After removing outliers, the first Valid inner boundary point sample values.

[0031] Define physiological stages Next, the The middle range of the parameter is The upper bound of the normal range for the corresponding parameter is... The lower bound of the normal range for the corresponding parameter is... When the inner boundary is close to the upper boundary: When the inner boundary is close to the lower boundary: .

[0032] S102, decompose the full-time data of each target parameter into continuous equal-duration minimum fluctuation units, extract the fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time as unit features; mark the minimum fluctuation unit whose amplitude falls into the middle interval as the abnormal trigger unit.

[0033] Specifically, this step standardizes and modularizes continuous monitoring data based on physiological fluctuation cycles, avoiding the shortcomings of traditional single-point threshold determination which is susceptible to instantaneous interference and has a high false positive rate, thus providing a unified and accurate analysis unit for subsequent disease transmission relationship analysis.

[0034] First, the full-time series data is broken down into the smallest continuous fluctuation units of equal duration. The core principle is that fluctuations in human vital signs have inherent physiological periodicity. Dividing the units based on complete cycles avoids waveform distortion caused by cross-cycle segmentation, eliminates random errors from single-point sampling, and improves the anti-interference capability and accuracy of anomaly identification. The full-time series data consists of monitoring data for the four core vital signs—heart rate, respiratory rate, body temperature, and blood oxygen saturation—determined by S101. This is a complete, time-stamped monitoring data stream continuously collected by clinical monitoring equipment at a universal sampling frequency of 1Hz.

[0035] Secondly, the fluctuation amplitude of each parameter within each minimum fluctuation unit and the corresponding time are extracted as unit features. Specifically, for the time series data of a single target parameter within a single minimum fluctuation unit, the difference between the maximum and minimum values ​​of the parameter within that unit is calculated to obtain the fluctuation amplitude. Simultaneously, the acquisition time corresponding to the peak amplitude and the time when the parameter first deviates from the unit baseline value are extracted, together constituting the unit features. This calculation method characterizes the degree of parameter change with the overall fluctuation amplitude over a complete cycle, rather than relying on single-point values. This effectively filters out spurious data caused by electromagnetic interference from monitoring equipment and changes in patient positioning, avoiding misjudging transient fluctuations unrelated to the patient's condition as abnormal. Furthermore, the retained time dimension information provides accurate time anchors for subsequent abnormal time series correlation analysis.

[0036] Finally, the smallest fluctuation unit whose amplitude falls into the middle range is marked as the abnormal trigger unit. Specifically, when the peak value of the target parameter fluctuation amplitude falls into the middle range defined by S101 corresponding to the patient's physiological stage, the unit is marked as an abnormal trigger unit. This rule uses the peak value of the complete fluctuation unit falling into the warning range as the trigger condition, which can ensure that the abnormality is a continuous physiological trend change rather than a transient artifact. While reducing the false positive rate, it can accurately lock the time node of early abnormality of the disease.

[0037] The length of the minimum fluctuation unit of equal duration is set according to the following rules: Based on the physiological characteristics corresponding to the patient's age, the normal fluctuation frequency range of each parameter in the target parameters is determined. The duration required for the complete fluctuation cycle of the parameter with the lowest fluctuation frequency is taken as the benchmark. This benchmark is multiplied by a preset amplification factor to obtain the minimum fluctuation unit length, ensuring that each unit covers at least one complete physiological fluctuation waveform. Specifically, taking adult patients (18-65 years old) as an example, their normal respiratory rate has a maximum complete fluctuation cycle of 5 seconds, their heart rate has a maximum complete fluctuation cycle of 1 second, and their body temperature and blood oxygen saturation have the lowest fluctuation frequencies, with a maximum complete fluctuation cycle of 10 seconds. Therefore, the longest cycle of 10 seconds corresponding to body temperature is taken as the benchmark, the preset amplification factor is set to 2, and the final minimum fluctuation unit length is determined to be 20 seconds. This setting ensures that each unit completely covers at least one complete physiological fluctuation cycle of all target parameters, avoiding problems such as incomplete waveform segmentation and inaccurate amplitude calculation caused by excessively short unit durations. At the same time, the amplification factor reserves redundancy to adapt to individual physiological fluctuation differences of patients, ensuring the clinical universality and data stability of the unit division.

[0038] Furthermore, the full-time data of each target parameter is decomposed into continuous, equal-duration minimum fluctuation units. Specifically, this involves: identifying the time-series waveform of the parameter with the lowest normal fluctuation frequency among the target parameters; using an inflection point identification algorithm to detect the local turning point where the time-series waveform changes from continuous decline to continuous rise; and selecting the first local turning point appearing on the time axis of the data sequence as the cutting starting point. Starting from this cutting starting point, the full-time data is continuously segmented according to the length of the minimum fluctuation unit, ensuring that each minimum fluctuation unit contains a complete waveform fluctuation segment. This inflection point identification algorithm is implemented through first-order difference calculation. The time-series data of the parameter with the lowest fluctuation frequency is subdivided point by point using first-order difference. When the difference results of multiple consecutive (e.g., 8) equally spaced sampling points change from negative to positive, it is determined to be a local turning point where the waveform changes from decline to rise, and this point is the starting point of the complete physiological fluctuation cycle. Choosing the earliest inflection point in the time series as the starting point for cutting, rather than the first data acquisition point, ensures that each segmentation unit contains a complete "decline-rise" physiological fluctuation waveform, avoiding incomplete waveforms caused by cutting from the halfway point of the waveform, ensuring the accuracy of fluctuation amplitude calculation within each unit, and laying a reliable data foundation for the accurate marking of subsequent abnormal triggering units.

[0039] S103: Following the chronological order, sequentially determine whether the time interval between two adjacent abnormal trigger units does not exceed a preset duration. If it does not, connect the two abnormal trigger units in series to form a continuous transmission chain. Otherwise, terminate the current transmission chain and use the next abnormal trigger unit exceeding the preset interval as the new transmission starting point to reconstruct a new transmission chain. The core of this step is to integrate discrete early abnormal signals into a transmission path that reflects the pattern of disease progression through temporal correlation verification, addressing the clinical pain point that a single abnormal trigger unit cannot reflect the progression of the disease and is prone to isolated false positive warnings.

[0040] Specifically, the time interval between two adjacent abnormal triggering units is determined sequentially according to time chronological order to ensure it does not exceed a preset duration. In practice, all abnormal triggering units output by S102 are first arranged in ascending order of acquisition time sequence based on their built-in timestamp information, forming an abnormal triggering unit time sequence queue to avoid judgment errors caused by out-of-order data transmission. The time interval between two adjacent abnormal triggering units is calculated as follows: the start time of the subsequent abnormal triggering unit is extracted, and the end time of the preceding abnormal triggering unit is subtracted to obtain the time interval value between the two adjacent units. Using the start and end times of the units to calculate the interval, rather than the abnormal start time within the unit, fully covers the blank time of the two abnormal events, avoiding overly short interval calculations due to differences in the occurrence times of abnormalities within the unit, and ensuring the rigor of the time sequence correlation judgment. The preset duration is set based on the clinical deterioration pattern of the target cases. The core principle is as follows: for the same cases selected in S101, the transmission time sequence data of abnormal signs before the deterioration of multiple patients are traced back, and the maximum time interval between the early abnormalities of adjacent signs is statistically analyzed. The 95th percentile of this statistical value is taken as the preset duration. Taking adult severe pneumonia cases as an example, after retrospectively analyzing the data of 100 patients, the maximum transmission interval of early abnormalities of adjacent vital signs is 120 seconds. The 95th percentile is used to determine the preset duration as 100 seconds. This setting will not interrupt the real disease transmission link due to the duration being too short, nor will it include unrelated isolated abnormalities in the same link due to the duration being too long. It balances the sensitivity and specificity of the link construction.

[0041] If the time interval between two adjacent abnormal triggering units is less than or equal to a preset duration, the two units are determined to have a clinically significant correlation in disease progression, rather than being isolated random interference abnormalities. In this case, the preceding abnormal triggering unit is taken as the upstream node and the following abnormal triggering unit as the downstream node, and they are sequentially connected in chronological order and incorporated into the currently constructed progression chain. Through this chaining logic, discrete early abnormal signals can be integrated into a complete deterioration path according to the temporal pattern of disease progression, clearly reconstructing the progressive process of vital sign abnormalities, and providing a complete logical basis for subsequent risk level assessment and emergency plan matching.

[0042] Conversely, if the interval exceeds the preset interval, the current transmission chain is terminated, and a new transmission chain is reconstructed using the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. If, after calculation, the time interval between two adjacent abnormal triggering units is greater than the preset duration, it is determined that the two units have no clinical transmission correlation, and the subsequent abnormality is a new abnormal event independent of the preceding abnormality. At this time, the currently constructed transmission chain is immediately terminated and sealed to ensure the temporal integrity of the constructed link. At the same time, the next abnormal triggering unit that exceeds the preset interval is set as the starting node of the new transmission chain, and the correlation verification and link construction of subsequent adjacent abnormal triggering units are carried out again according to the aforementioned temporal judgment rules. This branch logic can effectively isolate unrelated abnormal events, avoid forcibly including abnormalities from different time periods and without transmission relationships into the same link, and ensure the accuracy of subsequent emergency treatment library matching.

[0043] S104. For two adjacent abnormal triggering units in the transmission relationship chain, calculate the time delay difference between the previous abnormal triggering unit and the next abnormal triggering unit; bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency processing library.

[0044] Specifically, the core principle of calculating the time lag difference is that in the process of disease deterioration of the same case, the early abnormal transmission of vital signs has a fixed temporal pattern. The time difference of the onset of adjacent abnormal signs can accurately quantify the transmission speed and progression level of disease deterioration. This value is not affected by individual baseline differences and can be used as the core quantitative indicator for determining the risk level of the disease. At the same time, it provides a unique feature matching anchor point for subsequent emergency plan matching, solving the clinical pain point that traditional early warning can only identify abnormalities but cannot predict the deterioration process.

[0045] Specifically, calculating the time delay difference between the previous and subsequent anomaly triggering units includes: First, for two adjacent abnormal triggering units in the transmission chain, the time-series data of the target parameter corresponding to the previous abnormal triggering unit within that unit is extracted. The moment when the parameter amplitude first enters the corresponding intermediate interval in the time-series data is identified and recorded as the start time of the preceding abnormality. The extracted time-series data is the continuous monitoring data of the target parameter within the abnormal triggering unit with a 1Hz sampling frequency timestamp. The parameter amplitude is the real-time monitoring value of the parameter at a single sampling point within the unit. The moment when it first enters the corresponding intermediate interval is the sampling point acquisition time in the unit's time-series data that satisfies the condition that "the parameter amplitude falls into the intermediate interval defined by S101, corresponding to the patient's physiological stage." This moment is the true starting point of the early abnormality, not the unit's starting time. Its core principle is to accurately remove the interference of normal fluctuation periods within the unit, lock the starting time anchor point of the abnormal signal, eliminate the time calculation error caused by unit division, and significantly improve the calculation accuracy of the time delay difference.

[0046] Secondly, the time-series data of the target parameter corresponding to the subsequent abnormal triggering unit is extracted within that unit. The moment when the parameter amplitude first enters the corresponding intermediate interval in the time-series data is identified and recorded as the start time of the subsequent abnormality. The subsequent abnormal triggering unit is a downstream node in the transmission chain that has a time-series transmission relationship with the previous abnormal triggering unit. Its corresponding target parameter belongs to a different vital sign dimension than that of the previous unit. This allows for the reconstruction of the abnormal transmission sequence of different vital signs. The principle is to accurately capture the transmission process of the disease between different physiological systems such as circulation and respiration by calibrating the abnormal start time of different upstream and downstream vital signs, thus providing a pathophysiological basis for the targeted matching of subsequent treatment plans.

[0047] Finally, the time difference between the start time of the subsequent anomaly and the start time of the preceding anomaly is calculated, and this time difference is determined as the time lag difference corresponding to the adjacent anomaly triggering units in that group. Specifically, the standard timestamp value of the start time of the preceding anomaly is subtracted from the standard timestamp value of the start time of the subsequent anomaly to obtain the time difference value in seconds. This calculation method uses absolute timestamps as a benchmark and is not affected by the sampling frequency of the monitoring equipment or the unit division length. The calculation result is unique and repeatable. The principle is to transform discrete anomaly events into quantifiable and matchable feature values ​​through standardized time difference calculation, providing a unified quantitative standard for feature matching in subsequent emergency response databases. The expression is as follows: (7) In the formula, In the transmission relationship chain, the first The and the first +1 time delay difference between adjacent abnormal triggering units; For the first Each abnormal triggering unit corresponds to the moment when the target parameter amplitude first enters the middle range; For the first +1 abnormal triggering unit corresponds to the moment when the target parameter amplitude first enters the middle interval; This is the node number of the abnormal triggering unit in the propagation relationship chain.

[0048] After calculating the time delay difference, each transmission chain, its corresponding time delay difference, and the corresponding treatment plan are bound together to generate an emergency response library. For each complete transmission chain, the time delay difference corresponding to all adjacent abnormal triggering units within the chain is bound one-to-one with the node sequence of the transmission chain and the target parameter type corresponding to each node. Simultaneously, a clinically validated standardized treatment plan corresponding to that transmission chain is matched, forming a complete set of emergency response entries. All emergency response entries corresponding to the same cases are aggregated and integrated to generate an emergency response library with search and matching functions. The principle is to transform clinically validated patterns of disease deterioration and treatment experience into standardized tools that can be automatically matched and invoked through a three-in-one binding of "transmission link - time delay characteristics - treatment plan," significantly shortening clinical response time while avoiding non-standardized treatment plans due to differences in clinical experience.

[0049] S105: Obtain the time-series data of the current patient's target parameters. If any time-series data falls into the abnormal range, issue an alarm directly. Otherwise, process the time-series data according to S102 to obtain the current patient's abnormal trigger units. If the number of abnormal trigger units is 1, continue monitoring. If the number is greater than 1, calculate the time interval between two adjacent abnormal trigger units in sequence. If all adjacent time intervals exceed the preset duration, send a minor abnormality prompt to the terminal. If there is at least one set of adjacent abnormal trigger units whose time intervals do not exceed the preset duration, issue an alarm and connect the set of abnormal trigger units in series to form a real-time transmission chain. Extract the corresponding time delay difference value in the real-time transmission chain and match the real-time transmission chain with the corresponding time delay difference value in the emergency processing library. Display the successfully matched treatment plan on the terminal.

[0050] Specifically, the abnormal range refers to all values ​​outside the clinically normal range defined by S101 based on the patient's age and corresponding physiological stage. The principle is that once a parameter exceeds the abnormal range, it indicates that the patient has experienced clear physiological decompensation, which is a critical condition requiring immediate intervention. This directly triggers the highest priority audible and visual alarm without the need for subsequent link analysis, thus avoiding delays in the process and the loss of the treatment window.

[0051] If the abnormal range is not breached, the time-series data is processed according to S102 to obtain the current patient's abnormal trigger unit. When all real-time monitoring data does not exceed the normal range boundary, the time-series data is continuously segmented according to the rules of S102, using the minimum fluctuation unit length matching the patient's physiological stage. The fluctuation amplitude characteristics of each unit are extracted, and units whose amplitude falls into the corresponding middle range are marked as abnormal trigger units. The principle is to filter out interference from monitoring equipment and instantaneous artifacts caused by changes in patient position through unitized processing, retaining only early abnormal signals with trends. The beneficial effect is to avoid false positive warnings caused by single-point data interference and to provide reliable abnormal nodes for subsequent link analysis.

[0052] During the judgment process, if the number of abnormal triggering units is 1, monitoring continues. That is, when only one abnormal triggering unit is identified in the full-time monitoring data, it represents only a single, isolated early fluctuation of vital signs, without clear characteristics of disease progression. Therefore, no warning is triggered, and the parameter type and corresponding time period of the abnormal unit are simply labeled on the terminal, while subsequent monitoring data continues to be collected. The principle is that a single, isolated abnormality is likely due to individual physiological fluctuations or minor environmental disturbances, and does not have clinical indication of disease deterioration; while the solution in this embodiment can reduce the frequency of meaningless warnings and reduce the ineffective workload of clinical medical staff.

[0053] During the judgment process, if the number is greater than 1, the time interval between two adjacent abnormal triggering units is calculated sequentially. When the number of abnormal triggering units is ≥ 2, all abnormal triggering units are first sorted in ascending order of timestamps, and then the difference between the start time of the next unit and the end time of the previous unit is calculated sequentially as the time interval between adjacent units. The principle is that using the start and end times of the unit to calculate the interval can completely cover the blank time period between two abnormal events, avoid the difference in the occurrence time of the abnormality within the unit causing the interval calculation to be too short, and ensure the rigor of the time sequence correlation judgment.

[0054] During the judgment process, if all adjacent time intervals exceed a preset duration, a minor anomaly alert is sent to the terminal. The preset duration is a fixed threshold obtained in S103 based on the statistical analysis of the clinical deterioration patterns of corresponding cases. Taking severe pneumonia in adults as an example, the preset duration is 100 seconds. When the intervals of all adjacent abnormal units exceed this threshold, it indicates that all abnormalities are isolated events with no disease progression correlation, and only a minor anomaly alert is pushed to the terminal. The principle is that unrelated isolated abnormalities only represent slight fluctuations in vital signs and have no risk of progressive deterioration. Its beneficial effect is to achieve graded risk alerts and avoid decreased clinical vigilance caused by excessive warnings.

[0055] During the judgment process, if the time interval between at least one set of adjacent abnormal triggering units does not exceed a preset duration, an alarm is triggered, and the set of abnormal triggering units is connected in series to form a real-time transmission chain. As long as the interval between any set of adjacent units is less than or equal to the preset duration, it is determined that there are progressive characteristics of disease transmission across physiological systems, triggering a medium-level warning. Simultaneously, the associated abnormal triggering units are sequentially connected in series according to time sequence to form a real-time transmission chain. The principle is that adjacent abnormalities that meet the preset interval correspond to the transmission process of the disease between systems such as the circulatory and respiratory systems, possessing a clear indication of deterioration. The beneficial effect is that it integrates discrete abnormal signals into an interpretable disease progression path, providing a complete logical basis for subsequent solution matching.

[0056] Extract the corresponding time delay difference values ​​from the real-time transmission chain and match the real-time transmission chain with the corresponding time delay difference values ​​in the emergency handling database. First, extract the moment when the parameter amplitude of the adjacent abnormal triggering unit first enters the middle interval in the real-time transmission chain, and calculate the time difference between the start time of the subsequent abnormality and the start time of the preceding abnormality to obtain the real-time time delay difference value. In the matching process, the time delay ratio feature matching method is adopted: first, calculate the ratio of the time delay difference values ​​of all adjacent nodes in the real-time transmission chain according to the transmission order to form a real-time ratio value sequence; at the same time, for each transmission relationship chain in the emergency handling database, the time delay ratio value sequence of the entire segment and all continuous local intervals is pre-generated. During matching, the cosine similarity between the real-time ratio value sequence and the sequence in the database is calculated. When the similarity is ≥90%, the match is considered successful. The principle is that during the deterioration of the same case, the relative speed of the transmission of abnormal signs follows a fixed pathophysiological law, which is not affected by the patient's individual baseline or the overall rate of disease progression. Compared with absolute numerical matching, it can adapt to the individual differences of different patients, avoid matching failure due to the patient's disease progressing too fast or too slow, and improve the compatibility and clinical accuracy of matching.

[0057] The formula for the cosine similarity of the time delay ratio sequences is: (8) In the formula, Real-time ratio sequence The sequence of ratios within the library The cosine similarity, with a value range of [-1, 1], is used to determine a successful match when the similarity is ≥0.9. This is a sequence of time delay ratios for real-time conduction chains. For the first in the sequence Each time delay ratio; This is a sequence of time delay ratios for the corresponding conduction chains within the emergency response repository. For the first in the sequence Each time delay ratio.

[0058] Then, the successfully matched treatment plan is displayed on the terminal. This includes: if the real-time transmission chain and its corresponding time delay difference match any continuous local interval of any transmission relationship chain in the emergency treatment database and its bound corresponding time delay difference, an alarm is sent to the terminal and the corresponding treatment plan is displayed simultaneously. The principle is that the deterioration of the condition is a gradual process, and the corresponding intervention plan can be matched in advance without waiting for the complete chain to be formed. The beneficial effect is to further improve the lead time for early warning and treatment, and seize the best clinical intervention window.

[0059] Furthermore, after displaying the successfully matched treatment plan on the terminal, the process also includes: receiving confirmation instructions or adjustments to the actual treatment plan submitted by the doctor through the terminal; if the actual treatment plan differs from the displayed plan, the actual treatment plan takes precedence; and adding the real-time transmission chain, the calculated time delay differences, and the finally confirmed actual treatment plan as incremental learning samples to the emergency treatment database. The terminal supports doctors in confirming, modifying, or completely replacing the pushed plan. The modified actual plan will overwrite the original pushed plan and serve as the final record of this treatment. Simultaneously, the real-time transmission chain, time delay ratio sequence, and final treatment plan are updated to the emergency treatment database as incremental samples. The principle is to continuously enrich the treatment scenarios in the database through continuous iteration of real clinical treatment experience. The beneficial effect is to achieve a self-optimizing closed loop of the method, continuously improving the accuracy and clinical applicability of subsequent matching.

[0060] like Figure 2 As shown, based on the same inventive concept, this embodiment provides a disease assessment system based on clinical indicators, including: The interval definition module 201 is used to obtain target parameters for multiple patients with the same case. The target parameters include heart rate, respiratory rate, body temperature and blood oxygen saturation. Based on the patient's age, the normal interval, intermediate interval and abnormal interval of each target parameter under the corresponding physiological stage are defined. The unit marking module 202 is used to decompose the full-time data of each target parameter into continuous equal-duration minimum fluctuation units, extract the fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time as unit features; and mark the minimum fluctuation unit whose amplitude falls into the middle interval as an abnormal trigger unit. The relationship chain building module 203 is used to determine, in chronological order, whether the time interval between two adjacent abnormal triggering units does not exceed the preset time. If it does not exceed the preset time, the two abnormal triggering units are connected in series to form a continuous transmission relationship chain. Otherwise, the current transmission relationship chain is terminated, and a new transmission relationship chain is reconstructed with the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. The emergency database construction module 204 is used to calculate the time delay difference between the previous and next abnormal triggering units for two adjacent abnormal triggering units in the transmission relationship chain; and to bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency processing database. The evaluation and matching module 205 is used to acquire the time-series data of the current patient's target parameters. If any time-series data falls into the abnormal range, an alarm is issued directly. Otherwise, the time-series data is processed according to S102 to obtain the current patient's abnormal trigger units. If the number of abnormal trigger units is 1, continuous monitoring is performed. If the number is greater than 1, the time interval between two adjacent abnormal trigger units is calculated sequentially. If all adjacent time intervals exceed the preset duration, a minor abnormality prompt is sent to the terminal. If there is at least one set of adjacent abnormal trigger units whose time intervals do not exceed the preset duration, an alarm is triggered and the set of abnormal trigger units is connected in series to form a real-time transmission chain. The corresponding time delay difference value in the real-time transmission chain is extracted, and the real-time transmission chain is matched with the corresponding time delay difference value in the emergency processing library. The successfully matched treatment plan is displayed on the terminal.

[0061] like Figure 3 As shown, based on the same inventive concept, this embodiment provides an electronic device, including: Memory 302 is used to store computer programs; Processor 301 is used to implement the method steps as described in the first aspect when executing a computer program.

[0062] Based on the same inventive concept, this embodiment provides a readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, it implements the method steps as described in the first aspect.

[0063] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0064] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A disease assessment method based on clinical indicators, characterized in that, include: Obtain target parameters from multiple patients with the same case, including heart rate, respiratory rate, body temperature, and blood oxygen saturation; Based on the patient's age, the normal range, intermediate range, and abnormal range of each target parameter under the corresponding physiological stage are defined; The full-time data of each target parameter is decomposed into continuous equal-duration minimum fluctuation units, and the fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time are extracted as unit features. The smallest fluctuation unit whose amplitude falls within the intermediate range is designated as the abnormal trigger unit; In chronological order, determine whether the time interval between two adjacent abnormal triggering units does not exceed the preset duration; If the threshold is not exceeded, the two abnormal triggering units will be connected in series to form a continuous transmission chain. Conversely, the current transmission chain is terminated, and a new transmission chain is reconstructed with the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. For two adjacent abnormal triggering units in the aforementioned transmission relationship chain, calculate the time delay difference between the preceding and following abnormal triggering units; bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency processing library; The time-series data of the current patient's target parameters are acquired. If any time-series data falls into the abnormal range, an alarm is issued directly. Otherwise, the time-series data is processed according to S102 to obtain the current patient's abnormal trigger unit. If the number of abnormal trigger units is 1, continuous monitoring is performed. If the number is greater than 1, the time interval between two adjacent abnormal triggering units is calculated sequentially; if all adjacent time intervals exceed the preset duration, a minor abnormality prompt is sent to the terminal; if at least one set of adjacent abnormal triggering units has a time interval that does not exceed the preset duration, an alarm is triggered and the set of abnormal triggering units is connected in series to form a real-time transmission chain; the corresponding time delay difference value in the real-time transmission chain is extracted, and the real-time transmission chain is matched with the corresponding time delay difference value in the emergency processing library; the successfully matched treatment plan is displayed on the terminal.

2. The method for assessing a patient's condition based on clinical indicators according to claim 1, characterized in that, The definition of the normal range, intermediate range, and abnormal range for each target parameter under the corresponding physiological stage specifically includes: Based on the physiological stage corresponding to the patient's age, the clinically known normal value range of each target parameter is obtained and determined as the normal range. All values ​​outside the normal range are defined as the abnormal range. For each target parameter, the historical monitoring records of multiple patients with the same case are traced back, and the continuous monitoring curve of the parameter is extracted. The deviation segment where the parameter value changes continuously in one direction from the normal range and eventually enters the abnormal range is identified. In the deviation segment, the turning point where the trend first reverses and begins to deviate continuously from the normal range is determined as the boundary point within the range. Calculate the mean of all the inner boundary points of the divergent segments, and use the mean to determine the inner boundary of the normal interval; the region between the inner boundary and the boundary of the normal interval closest to the inner boundary is determined as the middle interval.

3. The method for assessing a patient's condition based on clinical indicators according to claim 1, characterized in that, The length of the minimum fluctuation unit of equal duration is set according to the following rules: Based on the physiological characteristics corresponding to the patient's age, the normal fluctuation frequency range of each parameter in the target parameters is determined. The duration required for the complete fluctuation cycle of the parameter with the lowest fluctuation frequency is taken as the benchmark. This benchmark is multiplied by a preset amplification factor to obtain the minimum fluctuation unit length, so that each unit covers at least one complete physiological fluctuation waveform.

4. The method for assessing a patient's condition based on clinical indicators according to claim 3, characterized in that, The step of decomposing the full-time data of each target parameter into the smallest fluctuation units of continuous equal duration specifically includes: Identify the time-series waveform of the parameter with the lowest normal fluctuation frequency among the target parameters, use an inflection point identification algorithm to detect the local turning point where the time-series waveform changes from continuous decline to continuous rise, and select the local turning point that appears first on the time axis of the data sequence as the cutting starting point; starting from this cutting starting point, continuously divide the full time-series data according to the length of the minimum fluctuation unit, so that each minimum fluctuation unit contains a complete waveform fluctuation segment.

5. The method for assessing a patient's condition based on clinical indicators according to claim 1, characterized in that, The calculation of the time delay difference between the previous and subsequent anomaly triggering units specifically includes: For two adjacent abnormal triggering units in the transmission relationship chain, extract the time series data of the target parameter corresponding to the previous abnormal triggering unit within the unit, identify the moment when the parameter amplitude first enters the corresponding intermediate interval in the time series data, and record it as the start time of the preceding abnormality; Extract the time series data of the target parameter corresponding to the next abnormal triggering unit within the unit, identify the moment when the parameter amplitude first enters the corresponding intermediate interval in the time series data, and record it as the start time of the subsequent abnormality. Calculate the time difference between the start time of the subsequent abnormality and the start time of the preceding abnormality, and determine the time difference as the time delay difference corresponding to the adjacent abnormality triggering units in this group.

6. The method for assessing a patient's condition based on clinical indicators according to claim 1, characterized in that, The process of displaying the successfully matched treatment plan on the terminal specifically includes: If the real-time transmission chain and its corresponding time delay difference successfully match any continuous local interval of any transmission relationship chain in the emergency treatment database and its bound corresponding time delay difference, an alarm is sent to the terminal and the corresponding treatment plan is displayed simultaneously.

7. The method for assessing a patient's condition based on clinical indicators according to claim 1, characterized in that, After displaying the successfully matched treatment plan on the terminal, the process also includes: Receive confirmation instructions or adjustments to the actual treatment plan submitted by the doctor through the terminal; if the actual treatment plan differs from the displayed plan, the actual treatment plan shall prevail. The real-time conduction chain, the calculated time delay differences, and the finally confirmed actual treatment plan are used as incremental learning samples and added to the emergency treatment library.

8. A disease assessment system based on clinical indicators, based on the disease assessment method based on clinical indicators as described in claim 1, characterized in that, include: The interval delineation module is used to obtain target parameters for multiple patients with the same case, including heart rate, respiratory rate, body temperature and blood oxygen saturation. Based on the patient's age, the normal range, intermediate range, and abnormal range of each target parameter under the corresponding physiological stage are defined; The unit labeling module is used to decompose the full-time data of each target parameter into continuous equal-duration minimum fluctuation units, and extract the fluctuation amplitude of each parameter in each minimum fluctuation unit and the corresponding time as unit features. The smallest fluctuation unit whose amplitude falls within the intermediate range is designated as the abnormal trigger unit; The relationship chain building module is used to determine, in chronological order, whether the time interval between two adjacent abnormal triggering units does not exceed a preset duration; if it does not exceed the preset duration, the two abnormal triggering units are connected in series to form a continuous transmission relationship chain. Conversely, the current transmission chain is terminated, and a new transmission chain is reconstructed with the next abnormal triggering unit that exceeds the preset interval as the new transmission starting point. An emergency database construction module is used to calculate the time delay difference between two adjacent abnormal triggering units in the transmission relationship chain; and to bind each transmission relationship chain, the corresponding time delay difference, and the corresponding treatment plan to generate an emergency processing database. The evaluation and matching module is used to acquire the time series data of the current patient's target parameters. If any time series data falls into the abnormal range, an alarm is issued directly; otherwise, the time series data is processed according to S102 to obtain the current patient's abnormal trigger unit; if the number of abnormal trigger units is 1, continuous monitoring is performed. If the number is greater than 1, the time interval between two adjacent abnormal triggering units is calculated sequentially; if all adjacent time intervals exceed the preset duration, a minor abnormality prompt is sent to the terminal; if at least one set of adjacent abnormal triggering units has a time interval that does not exceed the preset duration, an alarm is triggered and the set of abnormal triggering units is connected in series to form a real-time transmission chain; the corresponding time delay difference value in the real-time transmission chain is extracted, and the real-time transmission chain is matched with the corresponding time delay difference value in the emergency processing library; the successfully matched treatment plan is displayed on the terminal.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement a clinical indicator-based disease assessment method as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program that, when executed by a processor, implements a clinical indicator-based disease assessment method as described in any one of claims 1 to 7.

Images