Vital sign-behavior-psychological multi-parameter linkage monitoring system for common-patient in general ward

Abstract

The invention discloses a vital sign-behavior-psychological multi-parameter linkage monitoring system for common patients in general wards, which relates to the technical field of intelligent medical monitoring and comprises a multi-parameter data acquisition module, a data preprocessing module, a multi-parameter association analysis module, a self-adaptive threshold calculation module, an early warning level judgment module and a database. Vital signs, behaviors and psychological parameters are collected in an integrated mode, and comprehensive coverage of health states of patients with common diseases is achieved; through correlation analysis, time-delay correlation analysis and causal relationship analysis, a weighted incidence matrix is constructed, an individual reference value is calculated based on patient historical data in combination with age and a basic disease correction coefficient, and individual differences of co-disease patients are dynamically updated and adapted to improve early warning accuracy. Through single-parameter abnormity judgment, multi-parameter linkage analysis and trend prediction, multi-level early warning from prompt to emergency is realized, intervention measures are accurately matched, and the safety of patients is improved.

Description

Technical Field

[0001] This invention relates to the field of smart medical monitoring technology, specifically to a multi-parameter linkage monitoring system for vital signs, behavior, and psychology of comorbid patients in general practice wards. Background Technology

[0002] In today's medical environment, the coexistence of multiple diseases has become a common challenge in general practice wards, especially among elderly patients. The clinical management complexity of these comorbid patients increases exponentially, and traditional single-disease management models are no longer sufficient to meet their diagnostic and treatment needs. These patients often need to address multiple health issues simultaneously, including abnormal physiological indicators, limited daily activities, and emotional and psychological disorders. However, most existing medical systems use decentralized monitoring methods, resulting in fragmented collection and analysis of key parameters such as vital signs, behavioral activities, and psychological states. This makes it difficult for medical staff to obtain a comprehensive and continuous health profile of the patients. Therefore, a multi-parameter linkage monitoring system for vital signs, behavior, and psychological states of comorbid patients in general practice wards has been invented.

[0003] Existing technologies have the following shortcomings, specifically: They employ decentralized monitoring methods, resulting in fragmented collection and analysis of key parameters such as vital signs, behavioral activities, and psychological states, making it difficult for healthcare professionals to obtain a comprehensive and continuous profile of patient health; traditional single-disease-based management models cannot adapt to the multi-dimensional health problems faced by comorbid patients, including abnormal physiological indicators, limited mobility, and emotional and psychological disorders, leading to high complexity in clinical management; they lack systematic analysis of the correlation and causal relationships between parameters, making it difficult to capture potential health risks under the linkage of multiple parameters, resulting in insufficient accuracy and timeliness of early warnings; and they fail to consider individual patient differences, using uniform thresholds for early warning, resulting in low personalization and a high risk of false alarms or missed alarms. Summary of the Invention

[0004] To address the aforementioned technical shortcomings, the present invention aims to provide a multi-parameter linkage monitoring system for vital signs, behavior, and psychological states of comorbid patients in general practice wards.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a multi-parameter linkage monitoring system for comorbid patients in general practice wards, including: a multi-parameter data acquisition module, used to collect relevant parameters of comorbid patients in general practice wards, specifically including vital sign parameters, behavioral parameters and psychological parameters.

[0006] The data preprocessing module is used to clean, handle missing values, standardize, and align time series data of relevant parameters collected from patients with comorbidities in general practice wards.

[0007] The multi-parameter correlation analysis module is used to analyze the correlation and causal relationship between various parameters of patients with comorbidities in general practice wards, and to construct a weighted correlation matrix between parameters.

[0008] The adaptive threshold calculation module is used to calculate individual baseline values ​​for various parameters of patients with comorbidities in general practice wards based on historical data of such patients.

[0009] The warning level determination module is used to determine the warning level and generate warning information based on the comprehensive analysis results of the adaptive threshold calculation module.

[0010] Preferably, the multi-parameter data acquisition module includes a vital signs acquisition unit, a behavior trajectory acquisition unit, and a psychological state acquisition unit.

[0011] The vital signs collection unit is used to collect heart rate, blood pressure, and respiratory rate of patients with comorbidities in the general ward.

[0012] The behavior trajectory acquisition unit is used to collect the activity level, sleep patterns, and daily activity abilities of comorbid patients in general practice wards.

[0013] The psychological state acquisition unit is used to collect the emotional state, cognitive function, and social interaction characteristics of patients with comorbidities in general practice wards.

[0014] Preferably, the analysis of the correlation and causal relationship among various parameters of patients with comorbidities in general practice wards adopts a multi-dimensional analysis method that combines correlation analysis, time-delay correlation analysis and causal relationship analysis.

[0015] The correlation analysis used Pearson correlation coefficient and Spearman rank correlation coefficient to obtain the correlation coefficient matrix among various parameters of comorbid patients in the general practice ward.

[0016] The time-delay correlation analysis utilizes the cross-correlation function to obtain the time-delay correlation coefficient matrix among various parameters of comorbid patients in the general ward.

[0017] The causal relationship analysis employed Granger causality test and information entropy transfer method to obtain the causal correlation coefficient matrix among various parameters of comorbid patients in the general ward.

[0018] Based on the above analysis results, a weighted correlation matrix among the parameters is constructed.

[0019] Preferably, the specific method for obtaining the correlation coefficient matrix among various parameters of patients with comorbidities in the general practice ward is as follows: Normality is tested on the preprocessed time-series data of various parameters of patients with comorbidities in the general practice ward to determine whether the time-series data of various parameters conform to a normal distribution. If the time-series data of a certain parameter of the patients with comorbidities in the general practice ward conforms to a normal distribution, the linear correlation coefficient among various parameters of the patients with comorbidities in the general practice ward is calculated based on the Pearson correlation coefficient formula. If the time-series data of a certain parameter of the patients with comorbidities in the general practice ward does not conform to a normal distribution, the nonlinear correlation coefficient among various parameters of the patients with comorbidities in the general practice ward is calculated based on the Spearman rank correlation coefficient formula. The correlation coefficients of all parameter pairs are integrated to obtain the correlation coefficient matrix among various parameters of patients with comorbidities in the general practice ward.

[0020] Preferably, the specific method for obtaining the time-lag correlation coefficient matrix among various parameters of patients with comorbidities in the general practice ward is as follows: Based on the cross-correlation coefficient calculation formula, the time-lag correlation coefficient of each parameter with different time lags is calculated for the preprocessed time series data of patients with comorbidities in the general practice ward. The absolute values ​​are processed, and the time lag corresponding to the maximum value of the time-lag correlation coefficient of each parameter is selected as the optimal time lag among the various parameters of patients with comorbidities in the general practice ward. The time-lag correlation coefficients of each parameter with the optimal time lag are integrated to obtain the time-lag correlation coefficient matrix among the various parameters of patients with comorbidities in the general practice ward.

[0021] Preferably, the method for calculating the individual baseline values ​​of each parameter for comorbid patients in the general practice ward is as follows: Relevant parameters of comorbid patients within a preset time period are obtained from the database; the mean and standard deviation of each parameter are calculated according to the time window; a weighted average is taken as the baseline value of each parameter for comorbid patients in the general practice ward; the age of comorbid patients in the general practice ward is obtained from the database, and an age correction coefficient corresponding to the age of the comorbid patients in the general practice ward is obtained; the underlying diseases of comorbid patients in the general practice ward are obtained from the database, and an underlying disease correction coefficient corresponding to the underlying diseases of the comorbid patients in the general practice ward is obtained; the baseline values ​​of each parameter of comorbid patients in the general practice ward are multiplied by the age correction coefficient corresponding to the age of the comorbid patients in the general practice ward, and by the underlying disease correction coefficient corresponding to the underlying diseases of the comorbid patients in the general practice ward, to obtain the individual baseline values ​​of each parameter of comorbid patients in the general practice ward.

[0022] Preferably, the early warning level determination module includes a single-parameter anomaly determination unit, a multi-parameter linkage determination unit, a trend prediction unit, and an early warning level classification unit.

[0023] The single-parameter anomaly determination unit is used to detect anomalies in a single parameter and quantify the degree of anomaly.

[0024] The multi-parameter linkage judgment unit is used to analyze the linkage abnormalities between multiple parameters and calculate the comprehensive abnormal index of each parameter pair for patients with comorbidities in the general ward.

[0025] The trend prediction unit is used to predict future abnormal risks based on historical data and current trends.

[0026] The warning level classification unit is used to classify warnings into alert levels, warning levels, emergency levels, and critical levels, and automatically match intervention measures according to different levels.

[0027] Preferably, the method for detecting and quantifying the abnormality of a single parameter is as follows: extract a parameter from a comorbid patient in a general ward, compare it with the individual baseline value of the same parameter for the comorbid patient in the general ward, calculate the deviation between the real-time data of the parameter and the individual baseline value for the comorbid patient in the general ward, and divide it by the standard deviation corresponding to the parameter for the comorbid patient in the general ward to obtain the abnormality index of the parameter for the comorbid patient in the general ward.

[0028] The abnormal index of this parameter for patients with comorbidities in the general practice ward is compared with the preset abnormal index threshold for this parameter in the database. If the abnormal index of this parameter for patients with comorbidities in the general practice ward is lower than the preset abnormal index threshold for this parameter in the database, then the parameter for patients with comorbidities in the general practice ward is determined to be normal. If the abnormal index of this parameter for patients with comorbidities in the general practice ward is higher than the preset abnormal index threshold for this parameter in the database, then the parameter for patients with comorbidities in the general practice ward is determined to be abnormal. This process yields the parameters of patients with abnormal comorbidities in the general practice ward, and the level of abnormality is determined based on the abnormal index of each parameter for patients with comorbidities in the general practice ward.

[0029] Preferably, the analysis of abnormal linkages between multiple parameters and the calculation of the comprehensive abnormality index of each parameter pair for comorbid patients in the general ward are specifically implemented as follows: extracting the abnormality index of each parameter of the abnormal comorbid patients in the general ward, performing standardization processing, determining the weight of each parameter pair based on a weighted correlation matrix, adding them together to obtain the comprehensive abnormality index of each parameter pair for the comorbid patients in the general ward, comparing the comprehensive abnormality index of each parameter pair for the comorbid patients in the general ward with the comprehensive abnormality index of the same parameter pair for comorbid patients in the general ward preset in the database, if the comorbid patients in the general ward... If the overall abnormality index of a certain parameter pair of a patient is lower than the overall abnormality index of the same parameter pair of patients in the general practice ward as preset in the database, then the linkage of that parameter in the general practice ward is determined to be normal. If the overall abnormality index of a certain parameter pair of a patient in the general practice ward is higher than the overall abnormality index of the same parameter pair of patients in the general practice ward as preset in the database, then the linkage of that parameter in the general practice ward is determined to be abnormal. The abnormal parameter pairs of the general practice ward patients are obtained, and the abnormality level is determined based on the overall abnormality index of the parameter pairs of the general practice ward patients.

[0030] Preferably, it also includes a database for storing relevant parameters of comorbid patients in the general practice ward within a preset time period, the age of the comorbid patients in the general practice ward, the underlying diseases of the comorbid patients in the general practice ward, the standardized abnormality index of each parameter of the comorbid patients in the preset general practice ward, and the comprehensive abnormality index of each parameter pair of the comorbid patients in the preset general practice ward.

[0031] The beneficial effects of this invention are as follows: It integrates the collection of vital signs, behavioral, and psychological parameters, achieving comprehensive coverage of the health status of comorbid patients; through correlation analysis, time-lag correlation analysis, and causal relationship analysis, a weighted correlation matrix is ​​constructed to reveal the complex relationships between parameters, avoiding the limitations of single-parameter monitoring; based on patient historical data, combined with age and underlying disease correction coefficients, individual baseline values ​​are calculated and dynamically updated to adapt to individual differences among comorbid patients, improving the accuracy of early warning; through single-parameter abnormality judgment, multi-parameter linkage analysis, and trend prediction, multi-level early warnings ranging from alert to critical conditions are achieved, accurately matching intervention measures, improving clinical management efficiency, and enhancing patient safety. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of the system structure connection of the present invention;

[0034] Figure 2 This is a schematic diagram of the multi-parameter data acquisition module of the present invention;

[0035] Figure 3 This is a schematic diagram of the early warning level determination module of the present invention. Detailed Implementation

[0036] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] according to Figure 1 As shown, this invention provides a multi-parameter linkage monitoring system for vital signs, behavior, and psychological states of comorbid patients in a general ward, comprising:

[0038] The multi-parameter data acquisition module is used to collect relevant parameters of comorbid patients in general practice wards, specifically including vital signs, behavioral parameters, and psychological parameters.

[0039] In one specific embodiment, the multi-parameter data acquisition module includes a vital signs acquisition unit, a behavior trajectory acquisition unit, and a psychological state acquisition unit.

[0040] The vital signs collection unit is used to collect heart rate, blood pressure, and respiratory rate of patients with comorbidities in the general ward.

[0041] It should be noted that the heart rate, blood pressure, and respiratory rate of the comorbid patients in the general ward were obtained through heart rate sensors, blood pressure sensors, and respiratory flow sensors, respectively.

[0042] The behavior trajectory acquisition unit is used to collect the activity level, sleep patterns, and daily activity abilities of comorbid patients in general practice wards.

[0043] It should be noted that the activity level specifically includes, but is not limited to, activity intensity, activity frequency, and activity duration; the sleep pattern specifically includes, but is not limited to, sleep duration, sleep cycle, and sleep efficiency; and the daily activities specifically include, but are not limited to, eating, dressing, and washing.

[0044] The psychological state acquisition unit is used to collect the emotional state, cognitive function, and social interaction characteristics of patients with comorbidities in general practice wards.

[0045] The data preprocessing module is used to clean, handle missing values, standardize, and align time series data of relevant parameters collected from patients with comorbidities in general practice wards.

[0046] It should be noted that the data cleaning step removes obviously abnormal or erroneous data points by setting reasonable parameter value ranges, ensuring the accuracy of subsequent analysis. The missing value handling step uses interpolation, mean imputation, or prediction methods based on historical data to complete missing data, ensuring data integrity. The standardization step uses methods such as Z-score standardization or min-max normalization to transform data with different parameters to a unified dimension and range, eliminating the scale effect between parameters. The time series alignment step synchronizes multi-source data according to the collection frequency and timestamp of each parameter, ensuring the comparability of each parameter at the same point in time, providing a solid data foundation for subsequent multi-parameter analysis and modeling.

[0047] The multi-parameter correlation analysis module is used to analyze the correlation and causal relationship between various parameters of patients with comorbidities in general practice wards, and to construct a weighted correlation matrix between parameters.

[0048] In one specific embodiment, the analysis of the correlation and causal relationship among various parameters of comorbid patients in general practice wards adopts a multi-dimensional analysis method that combines correlation analysis, time-delay correlation analysis and causal relationship analysis.

[0049] The correlation analysis used Pearson correlation coefficient and Spearman rank correlation coefficient to obtain the correlation coefficient matrix among various parameters of comorbid patients in the general practice ward.

[0050] In one specific embodiment, the method for obtaining the correlation coefficient matrix among parameters of comorbid patients in a general practice ward is as follows: A normality test is performed on the preprocessed time-series data of each parameter of the comorbid patients in the general practice ward to determine whether the time-series data of each parameter conforms to a normal distribution. If the time-series data of a certain parameter of the comorbid patients in the general practice ward conforms to a normal distribution, then the linear correlation coefficient among the parameters of the comorbid patients in the general practice ward is calculated based on the Pearson correlation coefficient formula. If the time-series data of a certain parameter of the comorbid patients in the general practice ward does not conform to a normal distribution, then the nonlinear correlation coefficient among the parameters of the comorbid patients in the general practice ward is calculated based on the Spearman rank correlation coefficient formula. The correlation coefficients of all parameter pairs are integrated to obtain the correlation coefficient matrix among the parameters of the comorbid patients in the general practice ward.

[0051] It should be noted that the Pearson correlation coefficient and Spearman rank correlation coefficient are existing technologies and will not be elaborated upon here.

[0052] The time-delay correlation analysis utilizes the cross-correlation function to obtain the time-delay correlation coefficient matrix among various parameters of comorbid patients in the general ward.

[0053] In one specific embodiment, the method for obtaining the time-lag correlation coefficient matrix among various parameters of comorbid patients in the general practice ward is as follows: Based on the cross-correlation coefficient calculation formula, the time-lag correlation coefficient of each parameter with different time lags is calculated for the preprocessed time series data of each parameter of comorbid patients in the general practice ward. The absolute values ​​are processed, and the time lag corresponding to the maximum value of the time-lag correlation coefficient of each parameter is selected as the optimal time lag among various parameters of comorbid patients in the general practice ward. The time-lag correlation coefficients of each parameter with the optimal time lag are integrated to obtain the time-lag correlation coefficient matrix among various parameters of comorbid patients in the general practice ward.

[0054] It should be noted that the cross-correlation coefficient is existing technology and will not be elaborated upon here.

[0055] The causal relationship analysis employed Granger causality test and information entropy transfer method to obtain the causal correlation coefficient matrix among various parameters of comorbid patients in the general ward.

[0056] It should be noted that the specific process for obtaining the causal correlation coefficient matrix among various parameters of patients with comorbidities in the general practice ward is as follows: First, the preprocessed multi-parameter time series data is subjected to a stationarity test to ensure that the Granger causality test conditions are met; an autoregressive model is constructed for each pair of parameters, and the models with and without the lagged term of the other parameter are estimated respectively. The F statistic is calculated to test the null hypothesis (X does not Granger cause Y). If rejected, a Granger causal relationship is determined to exist and the causality index is calculated; at the same time, the continuous parameters are transformed into a state sequence through discretization, and the KL divergence of the conditional probabilities p(y_{t+1}|y_t,x_t) and p(y_{t+1}|y_t) is calculated to obtain the information entropy transfer amount, quantifying the information flow intensity from X to Y; the stability of the results is verified by Bootstrap resampling, and statistically insignificant causal relationships are eliminated; finally, the strength (Granger causality index or standardized transfer entropy) and direction (X→Y or Y→X) of significant causal relationships are arranged into an antisymmetric matrix to obtain the causal correlation coefficient matrix among various parameters of patients with comorbidities in the general practice ward.

[0057] Where p(y_{t+1}|y_t,x_t) represents the transition probability of Y given the current state of X, x_t represents the self-transition probability of Y improving the prediction of Y based on additional information provided by X, and p(y_{t+1}|y_t) represents the uncertainty in predicting the future state based solely on the historical state of Y. X and Y represent the hypothesized parameters, {Xt} and {Yt} represent the hypothesized parameter time series, and t represents time.

[0058] Based on the above analysis results, a weighted correlation matrix among the parameters is constructed.

[0059] The adaptive threshold calculation module is used to calculate individual baseline values ​​for various parameters of patients with comorbidities in general practice wards based on historical data of such patients.

[0060] In one specific embodiment, the method for calculating the individual baseline values ​​of various parameters of comorbid patients in the general practice ward is as follows: Relevant parameters of comorbid patients within a preset time period are obtained from the database; the mean and standard deviation of each parameter are calculated according to the time window; a weighted average is taken as the baseline value of each parameter of the comorbid patients in the general practice ward; the age of the comorbid patients in the general practice ward is obtained from the database, and an age correction coefficient corresponding to the age of the comorbid patients in the general practice ward is obtained; the underlying diseases of the comorbid patients in the general practice ward are obtained from the database, and an underlying disease correction coefficient corresponding to the underlying diseases of the comorbid patients in the general practice ward is obtained; the baseline values ​​of each parameter of the comorbid patients in the general practice ward are multiplied by the age correction coefficient corresponding to the age of the comorbid patients in the general practice ward, and the underlying disease correction coefficient corresponding to the underlying diseases of the comorbid patients in the general practice ward, to obtain the individual baseline values ​​of each parameter of the comorbid patients in the general practice ward.

[0061] The warning level determination module is used to determine the warning level and generate warning information based on the comprehensive analysis results of the adaptive threshold calculation module.

[0062] In one specific embodiment, the early warning level determination module includes a single-parameter anomaly determination unit, a multi-parameter linkage determination unit, a trend prediction unit, and an early warning level classification unit.

[0063] The single-parameter anomaly determination unit is used to detect anomalies in a single parameter and quantify the degree of anomaly.

[0064] In one specific embodiment, the method for detecting and quantifying the degree of abnormality of a single parameter is as follows: extract a parameter from a comorbid patient in a general ward, compare it with the individual baseline value of the same parameter for the comorbid patient in the general ward, calculate the deviation between the real-time data of the parameter and the individual baseline value for the comorbid patient in the general ward, and divide it by the standard deviation corresponding to the parameter for the comorbid patient in the general ward to obtain the abnormality index of the parameter for the comorbid patient in the general ward.

[0065] The abnormal index of this parameter for patients with comorbidities in the general practice ward is compared with the preset abnormal index threshold for this parameter in the database. If the abnormal index of this parameter for patients with comorbidities in the general practice ward is lower than the preset abnormal index threshold for this parameter in the database, then the parameter for patients with comorbidities in the general practice ward is determined to be normal. If the abnormal index of this parameter for patients with comorbidities in the general practice ward is higher than the preset abnormal index threshold for this parameter in the database, then the parameter for patients with comorbidities in the general practice ward is determined to be abnormal. This process yields the parameters of patients with abnormal comorbidities in the general practice ward, and the level of abnormality is determined based on the abnormal index of each parameter for patients with comorbidities in the general practice ward.

[0066] The multi-parameter linkage judgment unit is used to analyze the linkage abnormalities between multiple parameters and calculate the comprehensive abnormal index of each parameter pair for patients with comorbidities in the general ward.

[0067] In one specific embodiment, the analysis of abnormal linkages between multiple parameters and the calculation of the comprehensive abnormality index of each parameter pair for comorbid patients in the general practice ward are implemented as follows: The abnormality indices of each parameter for abnormal comorbid patients in the general practice ward are extracted, standardized, and the weights of each parameter pair are determined based on a weighted association matrix. These weights are then summed to obtain the comprehensive abnormality index of each parameter pair for the comorbid patients in the general practice ward. The comprehensive abnormality index of each parameter pair for the comorbid patients in the general practice ward is compared with the preset comprehensive abnormality index of the same parameter pair for comorbid patients in the general practice ward in the database. If the general practice ward... If the overall abnormality index of a parameter pair of a comorbid patient in the general practice ward is lower than the pre-set overall abnormality index of the same parameter pair for comorbid patients in the general practice ward, then the linkage of that parameter in the comorbid patient in the general practice ward is determined to be normal. If the overall abnormality index of a parameter pair of a comorbid patient in the general practice ward is higher than the pre-set overall abnormality index of the same parameter pair for comorbid patients in the general practice ward, then the linkage of that parameter in the comorbid patient in the general practice ward is determined to be abnormal. This results in obtaining the abnormal parameter pairs of comorbid patients in the general practice ward, and simultaneously determining the abnormality level based on the overall abnormality index of each parameter pair of comorbid patients in the general practice ward.

[0068] The trend prediction unit is used to predict future abnormal risks based on historical data and current trends.

[0069] It should be noted that the prediction of future abnormal risks based on historical data and current trends specifically involves: firstly, preprocessing the patient's historical data (such as vital signs, behavior, and psychological parameters over the past 24-72 hours) and extracting multi-dimensional features: time features (such as circadian rhythm cycles and time-specific fluctuation patterns), trend features (such as the rate of change of the mean within a sliding window and slope trends), and correlation features (parameter linkage patterns based on a weighted correlation matrix); subsequently, a fusion model is used for prediction, employing an LSTM neural network to capture the long-term and short-term dependencies between parameters (input is the historical sequence, output is the predicted parameter values ​​for the next 4-8 hours), combined with an ARIMA model to supplement linear trend prediction, and introducing personalized correction factors (…). (e.g., the influence weight of age and underlying diseases on parameter trends); during the prediction process, the model input is updated in real time through a sliding window (the latest data is included every 30 minutes), and the prediction results are dynamically adjusted; finally, the predicted value is compared with the adaptive threshold (the threshold adjusted by combining individual baseline values ​​and circadian rhythm) to calculate the probability of future abnormal risks (e.g., the probability that the predicted value of a certain parameter exceeds the threshold), and a comprehensive risk index is generated based on the synergistic probability of multiple parameter linkage abnormalities (e.g., the combined risk of increased heart rate and decreased blood oxygen). The index is mapped to four risk levels: low, medium, high, and very high, to achieve early warning of future abnormal risks, and each level corresponds to a specific clinical attention priority (e.g., high risk prompts a re-examination within 1 hour, and very high risk triggers immediate intervention preparation).

[0070] The warning level classification unit is used to classify warnings into alert levels, warning levels, emergency levels, and critical levels, and automatically match intervention measures according to different levels.

[0071] It should be noted that the early warning system is divided into four levels: alert, warning, emergency, and critical. Intervention measures are automatically matched according to each level, and these are set by professionals. For example, a quantitative classification standard for the four-level early warning system is first established, using a comprehensive anomaly index, the degree of anomaly of a single parameter, the characteristics of multi-parameter linkage, and trend prediction risk as core indicators. The alert level corresponds to a comprehensive anomaly index of 0.3-0.5 (Z-score absolute value 1-2), indicating a slight anomaly in a single parameter and no multi-parameter linkage, with an anomaly risk probability of <20% in the next 4-8 hours; the warning level corresponds to a comprehensive anomaly index of 0.5-0.7 (Z-score absolute value 1-2). Z-score absolute value 2-3: 2-3 non-core parameters are abnormal or a single core parameter is mildly abnormal, with a risk probability of 20%-40%; Emergency level corresponds to a comprehensive abnormality index of 0.7-0.9 (Z-score absolute value 3-4): moderate abnormality of core parameters (such as blood pressure, blood oxygen), and more than 3 parameters showing coordinated abnormality, with a risk probability of 40%-70%; Critical level corresponds to a comprehensive abnormality index >0.9 (Z-score absolute value >4): severe abnormality of core parameters (such as systolic blood pressure >200 mmHg, blood oxygen <85%), and synergistic deterioration of multiple parameters, with a risk probability >70%. Based on this, the system automatically matches intervention measures: For alert levels, a patient status summary is pushed to the responsible nurse's terminal via the hospital intranet, requiring review within 2 hours; for warning levels, an audio-visual alert is triggered at the nurses' station, along with details of abnormal parameters and preliminary analysis, requiring nurses to assess within 1 hour; for emergency levels, the alert is simultaneously pushed to the attending physician's terminal, along with a trend prediction chart and intervention suggestions (such as adjusting medication dosage), requiring a response within 30 minutes; for critical levels, the ward emergency response system is activated, automatically retrieving the patient's emergency plan, triggering audio-visual alarms for the medical team (including bedside device linkage reminders), and simultaneously notifying the on-duty head nurse, requiring arrival within 15 minutes. Furthermore, the system tracks intervention feedback in real time, automatically escalating unprocessed warnings (such as warning levels exceeding 1 hour or emergency levels exceeding 30 minutes) to the terminals of higher-level medical staff, ensuring a closed-loop intervention process.

[0072] In one specific embodiment, a database is also included to store relevant parameters of comorbid patients in the general practice ward within a preset time period, the age of the comorbid patients in the general practice ward, the underlying diseases of the comorbid patients in the general practice ward, the standardized abnormality index of each parameter of the comorbid patients in the preset general practice ward, and the comprehensive abnormality index of each parameter pair of the comorbid patients in the preset general practice ward.

[0073] It should be noted that the multi-parameter data acquisition module is connected to the data preprocessing module, the data preprocessing module is connected to the multi-parameter correlation analysis module, the multi-parameter correlation analysis module is connected to the adaptive threshold calculation module, the adaptive threshold calculation module is connected to the early warning level determination module, and the multi-parameter correlation analysis module and the adaptive threshold calculation module are both connected to the database.

[0074] This invention integrates the collection of vital signs, behavioral, and psychological parameters to achieve comprehensive coverage of the health status of comorbid patients. Through correlation analysis, time-lag correlation analysis, and causal relationship analysis, a weighted correlation matrix is ​​constructed to reveal the complex relationships between parameters, avoiding the limitations of single-parameter monitoring. Based on patients' historical data, individual baseline values ​​are calculated and dynamically updated using age and underlying disease correction coefficients to adapt to individual differences among comorbid patients and improve early warning accuracy. Through single-parameter anomaly detection, multi-parameter linkage analysis, and trend prediction, multi-level early warnings ranging from alert to critical conditions are achieved, enabling precise matching of intervention measures, improving clinical management efficiency, and enhancing patient safety.

[0075] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A general ward co-morbidity patient vital signs-behavior-psychological multi-parameter linkage monitoring system, characterized in that, The application relates to a multi-parameter data acquisition module for acquiring relevant parameters of a general ward comorbidity patient, specifically including vital sign parameters, behavior parameters and psychological parameters. A data preprocessing module is used for cleaning, missing value processing, standardization and time sequence alignment of the acquired relevant parameters of the general ward comorbidity patient. A multi-parameter correlation analysis module is used for analyzing the correlation and causality between the parameters of the general ward comorbidity patient and constructing a weighted correlation matrix between the parameters. An adaptive threshold calculation module is used for calculating individual baseline values of the parameters of the general ward comorbidity patient based on historical data of the general ward comorbidity patient. An early warning level determination module is used for determining the early warning level and generating early warning information according to the comprehensive analysis result of the adaptive threshold calculation module. The multi-parameter data acquisition module comprises a vital sign acquisition unit, a behavior trajectory acquisition unit and a psychological state acquisition unit.

2. The multi-parameter linkage monitoring system for vital signs-behavior- psychology of patients with comorbidities in a general ward according to claim 1, characterized in that, The vital sign acquisition unit is used for acquiring the heart rate, blood pressure and respiratory rate of the general ward comorbidity patient. The behavior trajectory acquisition unit is used for acquiring the activity amount, sleep mode and daily activity ability of the general ward comorbidity patient. The psychological state acquisition unit is used for acquiring the emotional state, cognitive function and social interaction characteristics of the general ward comorbidity patient. The correlation and causality between the parameters of the general ward comorbidity patient are analyzed by adopting a multi-dimensional analysis mode combining correlation analysis, time lag correlation analysis and causality analysis.

3. The system according to claim 1, wherein, The correlation analysis obtains a correlation coefficient matrix between the parameters of the general ward comorbidity patient by a Pearson correlation coefficient and a Spearman rank correlation coefficient. The time lag correlation analysis obtains a time lag correlation coefficient matrix between the parameters of the general ward comorbidity patient by a cross-correlation function. The causality analysis obtains a causality correlation coefficient matrix between the parameters of the general ward comorbidity patient by Granger causality test and information entropy transfer method. The weighted correlation matrix between the parameters is constructed based on the above analysis results. The correlation coefficient matrix between the parameters of the general ward comorbidity patient is obtained by the following method:

4. The multi-parameter linkage monitoring system for vital signs-behavior- psychology of patients with comorbidities in a general ward according to claim 1, characterized in that, Normality test is conducted on the preprocessed parameter time sequence data of the general ward comorbidity patient to determine whether the parameter time sequence data of the general ward comorbidity patient conforms to normal distribution. If the parameter time sequence data of the general ward comorbidity patient conforms to normal distribution, the linear correlation coefficient between the parameters of the general ward comorbidity patient is calculated based on a Pearson correlation coefficient calculation formula. If the parameter time sequence data of the general ward comorbidity patient does not conform to normal distribution, the nonlinear correlation coefficient between the parameters of the general ward comorbidity patient is calculated based on a Spearman rank correlation coefficient calculation formula. The correlation coefficients of all parameter pairs are integrated to obtain the correlation coefficient matrix between the parameters of the general ward comorbidity patient. The time lag correlation coefficient matrix between the parameters of the general ward comorbidity patient is obtained by the following method:

5. The general ward comorbidity patient vital signs-behavior-psychological multi-parameter linkage monitoring system according to claim 4, characterized in that, ​ The time series data of each parameter of the pre-processed general ward comorbidity patient is processed to calculate the time lag correlation coefficient of each parameter at different time lags based on the cross-correlation coefficient calculation formula, the absolute value is processed, the time lag corresponding to the maximum time lag correlation coefficient of each parameter is screened to obtain the optimal time lag between each parameter of the general ward comorbidity patient, and the time lag correlation coefficient of each parameter at the optimal time lag is integrated to obtain the time lag correlation coefficient matrix between each parameter of the general ward comorbidity patient.

6. The multi-parameter linkage monitoring system for vital signs-behavior- psychology of patients with comorbidities in a general ward according to claim 4, characterized in that, The individual reference value of each parameter of the general ward comorbidity patient is calculated, and the specific implementation method is: The related parameters of the general ward comorbidity patient in a preset time period are obtained from the database, the mean and standard deviation of each parameter are calculated according to the time window, the weighted average is taken as the reference value of each parameter of the general ward comorbidity patient, the age of the general ward comorbidity patient is obtained from the database, and then the age correction coefficient corresponding to the age of the general ward comorbidity patient is obtained, the basic disease of the general ward comorbidity patient is obtained from the database, and then the basic disease correction coefficient corresponding to the basic disease of the general ward comorbidity patient is obtained, the reference value of each parameter of the general ward comorbidity patient is multiplied by the age correction coefficient corresponding to the age of the general ward comorbidity patient and the basic disease correction coefficient corresponding to the basic disease of the general ward comorbidity patient to obtain the individual reference value of each parameter of the general ward comorbidity patient.

7. The general ward comorbidity patient vital signs-behavior-psychological multi-parameter linkage monitoring system according to claim 6, characterized in that, The early warning level determination module includes a single parameter abnormality determination unit, a multi-parameter linkage determination unit, a trend prediction unit and a early warning level division unit; The single parameter abnormality determination unit is used for detecting the abnormality of a single parameter and quantifying the abnormality degree; The multi-parameter linkage determination unit is used for analyzing the linkage abnormality between multiple parameters and calculating the comprehensive abnormality index of each parameter pair of the general ward comorbidity patient; The trend prediction unit is used for predicting future abnormal risk based on historical data and current trend; The early warning level division unit is used for dividing the early warning into prompt level, warning level, emergency level and critical level, and automatically matching intervention measures according to different levels.

8. The general ward comorbidity patient vital signs-behavior-psychological multi-parameter linkage monitoring system according to claim 7, characterized in that, The specific implementation method for detecting the abnormality of a single parameter and quantifying the abnormality degree is: The parameter of the general ward comorbidity patient is extracted, which is compared with the individual reference value of the parameter of the general ward comorbidity patient, the deviation value of the real-time data of the parameter of the general ward comorbidity patient and the individual reference value is calculated, and then divided by the standard deviation corresponding to the parameter of the general ward comorbidity patient to obtain the abnormality index of the parameter of the general ward comorbidity patient; The abnormal index of the parameter of the general ward comorbidity patient is compared with the abnormal index threshold of the parameter of the general ward comorbidity patient in the database, if the abnormal index of the parameter of the general ward comorbidity patient is lower than the abnormal index threshold of the parameter of the general ward comorbidity patient in the database, it is determined that the parameter of the general ward comorbidity patient is normal, if the abnormal index of the parameter of the general ward comorbidity patient is higher than the abnormal index threshold of the parameter of the general ward comorbidity patient in the database, it is determined that the parameter of the general ward comorbidity patient is abnormal, and each parameter of the abnormal general ward comorbidity patient is obtained, and the abnormal level is determined based on the abnormal index of each parameter of the general ward comorbidity patient.

9. The general ward comorbidity patient vital signs-behavior-psychological multi-parameter linkage monitoring system according to claim 8, characterized in that, The linkage abnormality between the parameters is analyzed, and the comprehensive abnormal index of each parameter pair of the general ward comorbidity patient is calculated, and the specific implementation method is as follows: The abnormal index of each parameter of the abnormal general ward comorbidity patient is extracted, and standardized processing is performed, the weight of each parameter pair is determined based on the weighted correlation matrix, and the comprehensive abnormal index of each parameter pair of the general ward comorbidity patient is obtained by addition, the comprehensive abnormal index of each parameter pair of the general ward comorbidity patient is compared with the comprehensive abnormal index of the parameter pair of the general ward comorbidity patient in the database, if the comprehensive abnormal index of a certain parameter pair of the general ward comorbidity patient is lower than the comprehensive abnormal index of the parameter pair of the general ward comorbidity patient in the database, it is determined that the linkage of the parameter of the general ward comorbidity patient is normal, if the comprehensive abnormal index of a certain parameter pair of the general ward comorbidity patient is higher than the comprehensive abnormal index of the parameter pair of the general ward comorbidity patient in the database, it is determined that the linkage of the parameter of the general ward comorbidity patient is abnormal, and each parameter pair of the abnormal general ward comorbidity patient is obtained, and the abnormal level is determined based on the comprehensive abnormal index of each parameter pair of the general ward comorbidity patient.

10. The multi-parameter linkage monitoring system for vital signs-behavior- psychology of patients with comorbidities in a general ward according to claim 1, characterized in that, The database is also included, which is used for storing the related parameters of the general ward comorbidity patient in a preset time period, the age of the general ward comorbidity patient, the basic disease of the general ward comorbidity patient, the standardized abnormal index of each parameter of the general ward comorbidity patient, and the comprehensive abnormal index of each parameter pair of the general ward comorbidity patient.

Images