Intensive care medicine department patient monitoring and management system based on multi-modal data fusion

Abstract

The invention relates to the technical field of medical severe illness monitoring and management systems, in particular to an intensive illness medical patient monitoring and management system based on multi-modal data fusion, and the system comprises a physiological feature extraction module which collects a first feature set and processes the first feature set into a second feature set containing an electroencephalogram entropy value and cerebral perfusion pressure; the pathological state decoupling module is used for inversely calculating an observation vector, calculating a pharmacological stress vector in combination with a pharmacological baseline and a drug dosage, obtaining a pathological vector after stripping, and generating a pathological index; the evolution trend prediction module is used for predicting risk factors according to the pathological vector time sequence data; the closed-loop intervention decision-making module is used for combining the pathological indexes and the risk factors, calculating the medicine dosage adjustment amount according to a second mapping relation, and updating the infusion rate to achieve closed-loop regulation and control, and according to multi-modal data fusion and module cooperation, pharmacological stress is stripped, and the pathological indexes are generated so as to accurately judge the illness state; the risk can be actively predicted, and the drug dosage is automatically adjusted in combination with pathological indexes, so that the management efficiency and safety of critical patients are improved.

Description

Technical Field

[0001] This invention relates to the field of medical critical care monitoring and management systems, specifically to a critical care medicine patient monitoring and management system based on multimodal data fusion. Background Technology

[0002] In intensive care units, particularly neurocritical care units, patient monitoring relies heavily on isolated interpretations of multiple physiological parameters by clinicians and experience-based adjustments to treatment interventions. The presentation of physiological states is often the result of the combined effects of internal disease progression and external pharmacological interventions. This leads to confusion between the true signals of disease deterioration and physiological fluctuations caused by drug effects, making precise differentiation difficult. Clinical decision-making thus faces challenges, as it becomes impossible to clearly determine whether changes in physiological indicators are primarily due to pathological or pharmacological factors. This ambiguity results in treatment interventions, such as the titration of sedative drug dosages, lacking precise quantitative evidence and potentially delaying effective responses to actual disease deterioration. Furthermore, existing methods generally lack the ability to predict future trends in pathological states, leaving monitoring and intervention largely in a reactive mode.

[0003] The root of these problems lies in the fact that current monitoring technologies have failed to establish dynamic models capable of integrating multimodal data and deconstructing their inherent causal relationships. Specifically, this manifests as a lack of a unified mathematical framework to describe the complex coupling relationships between physiological subsystems and between them and drug interventions. Simultaneously, existing technologies struggle to extract pharmacological stress components from observed mixed physiological signals in real time, thus failing to isolate purely pathological stress information. Ultimately, this results in clinicians being unable to obtain clear, quantitative insights into the nature of disease deterioration when a patient's physiological state fluctuates. This directly impacts the accuracy and foresight of closed-loop treatment and control strategies, posing a potential risk to the prevention of serious consequences such as secondary brain injury. Summary of the Invention

[0004] The purpose of this invention is to provide a critical care medicine patient monitoring and management system and method based on multimodal data fusion, which solves the problems existing in the background technology.

[0005] To address the aforementioned technical problems, this invention provides a critical care medicine patient monitoring and management system based on multimodal data fusion, comprising: a physiological feature extraction module, used to collect a first feature set representing the multidimensional physiological state of patients, and based on a preset first mapping relationship, to process the first feature set into a second feature set containing EEG entropy value and cerebral perfusion pressure;

[0006] The pathological state decoupling module is used to back-calculate the observed physiological state change vector based on the second feature set, and calculate the predicted pharmacological stress vector based on the system's preset pharmacological and physiological response baseline and the current drug dose change; by extracting the predicted pharmacological stress vector from the observed physiological state change vector, the pathological stress vector is obtained, and then combined with the dynamically adjusted pathological stress threshold, a pathological index characterizing the degree of disease deterioration is generated.

[0007] The evolution trend prediction module is used to predict and output risk prediction factors characterizing the patient's future pathological state based on the time-series data of the pathological stress vector output by the pathological state decoupling module.

[0008] The closed-loop intervention decision module is used to combine the pathological index with the risk prediction factor and calculate the drug dosage adjustment amount according to the preset second mapping relationship. The drug dosage adjustment amount is then used to update the current drug infusion rate to achieve closed-loop control.

[0009] Preferably, the first feature set collected by the physiological feature extraction module further includes mean arterial pressure and intracranial pressure; the first mapping relationship is specifically: the cerebral perfusion pressure is generated by subtracting the mean arterial pressure from the intracranial pressure.

[0010] Preferably, the process by which the pathological state decoupling module calculates the predicted pharmacological stress vector includes: firstly, obtaining the preset pharmacological physiological response baseline to characterize individual drug sensitivity; then, combining the current drug dose change, predicting the theoretical physiological state change caused by the dose change, and setting the theoretical physiological state change as the predicted pharmacological stress vector.

[0011] Preferably, the process by which the pathological state decoupling module generates the pathological index includes: firstly, calculating the magnitude of the pathological stress vector; secondly, obtaining the dynamically adjusted pathological stress threshold; and finally, obtaining the pathological index by calculating the ratio of the magnitude of the pathological stress vector to the pathological stress threshold.

[0012] Preferably, it also includes a monitoring status classification module, which is used to: classify the patient's current stage as a monitoring status stage, such as a stable stage, an observation stage, or an alarm stage, based on the historical fluctuation range of the second feature set; and the pathological state decoupling module dynamically adjusts the pathological stress threshold based on the monitoring status stage output by the monitoring status classification module.

[0013] Preferably, the evolution trend prediction module has a built-in long short-term memory network model; the evolution trend prediction module is also used to: integrate the time series data of the pathological stress vector with the time series data of auxiliary clinical parameters such as body temperature and inflammatory indicators to construct a pathological evolution feature dataset; input the pathological evolution feature dataset into the long short-term memory network model, and its output is the risk prediction factor.

[0014] Preferably, the closed-loop intervention decision module performs calculations based on the second mapping relationship, including: setting a high-priority warning threshold and a medium-priority warning threshold; when the pathological index is higher than the high-priority warning threshold, a protective strategy is triggered, and the drug dosage adjustment amount is set to a first negative value; when the pathological index is lower than the high-priority warning threshold but higher than the medium-priority warning threshold, a conservative strategy is triggered, and the drug dosage adjustment amount is set to a second negative value.

[0015] Preferably, when the pathological index is not higher than the medium-priority warning threshold, the closed-loop intervention decision module is further configured to: obtain the stage risk threshold corresponding to the current monitoring status stage; calculate the deviation between the risk prediction factor and the stage risk threshold by comparing the risk prediction factor and the stage risk threshold, and generate the drug dosage adjustment amount based on the deviation.

[0016] Preferably, it also includes a baseline calibration module, which is used to: record the response changes of the second feature set after the patient receives a standardized test drug dose in the early stage of hospital admission; process the response changes and set them as the pharmacological and physiological response baseline for use by the pathological state decoupling module.

[0017] Preferably, after obtaining the drug dosage adjustment amount, the closed-loop intervention decision module is further configured to: sum the current drug infusion rate with the drug dosage adjustment amount to generate a new recommended drug infusion rate; and output the new recommended drug infusion rate to the drug infusion device to perform closed-loop control.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] (1) By working collaboratively with the physiological feature extraction module, the pathological state decoupling module, and the baseline calibration module, the core technical challenge of confusing the true state of the patient's condition with the effects of drugs in traditional monitoring was solved. The system utilizes the pathological state decoupling module to extract the predicted pharmacological stress vector from the observed physiological state change vector, thereby obtaining a pure pathological stress vector. This process can accurately distinguish between pathological and pharmacological factors, and the resulting pathological index provides clinicians with an unprecedented objective quantitative indicator of the degree of disease deterioration. This quantitative assessment capability significantly improves the accuracy of disease assessment and provides a solid data foundation for subsequent treatment decisions.

[0020] (2) By integrating the evolutionary trend prediction module and the closed-loop intervention decision module, clinical monitoring is elevated from a passive response mode to a new level of proactive prediction and automated intervention. The evolutionary trend prediction module can generate forward-looking risk predictors, enabling the system to anticipate the risk of disease deterioration. The closed-loop intervention decision module combines these risk predictors with real-time pathological indices and automatically calculates drug dosage adjustments based on a preset second mapping relationship to achieve closed-loop control. This proactive and automated intervention approach ensures the timeliness and accuracy of treatment adjustments, effectively improving the management efficiency and safety of critically ill patients. Attached Figure Description

[0021] 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.

[0022] Figure 1 This is a logic block diagram of the system of the present invention. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0024] Example 1:

[0025] Please see Figure 1 The present invention provides a critical care patient monitoring and management system based on multimodal data fusion, comprising: a physiological feature extraction module, used to collect a first feature set representing the multidimensional physiological state of patients, and based on a preset first mapping relationship, process the first feature set into a second feature set containing EEG entropy value and cerebral perfusion pressure;

[0026] The pathological state decoupling module is used to back-calculate the observed physiological state change vector based on the second feature set, and calculate the predicted pharmacological stress vector based on the system's preset pharmacological and physiological response baseline and the current drug dose change; by extracting the predicted pharmacological stress vector from the observed physiological state change vector, the pathological stress vector is obtained, and then combined with the dynamically adjusted pathological stress threshold, a pathological index characterizing the degree of disease deterioration is generated.

[0027] The evolution trend prediction module is used to predict and output risk predictors that characterize the patient's future pathological state based on the time-series data of the pathological stress vector output by the pathological state decoupling module.

[0028] The closed-loop intervention decision module is used to combine pathological indices and risk prediction factors, and calculate the drug dosage adjustment amount based on the preset second mapping relationship. The drug dosage adjustment amount is then used to update the current drug infusion rate to achieve closed-loop control.

[0029] The physiological feature extraction module collects a first feature set from patients with traumatic brain injury in the neuro-intensive care unit through a multimodal sensor network. This first feature set includes multi-dimensional physiological parameters such as EEG signal complexity, mean arterial pressure (MAP), intracranial pressure (ICP), body temperature (T), and inflammatory marker I. Based on a preset first mapping relationship, the physiological feature extraction module subtracts MAP from ICP to generate cerebral perfusion pressure.

[0030]

[0031] Cerebral perfusion pressure;

[0032] Mean arterial pressure;

[0033] Intracranial pressure;

[0034] Simultaneously, the brainwave entropy value was extracted using an EEG signal complexity analysis algorithm. This forms the first feature set, which includes brain entropy values ​​and brain perfusion pressure.

[0035] The pathological state analysis module receives the second feature set output by the physiological feature extraction module and solves the equation using the physiological state vector:

[0036]

[0037] For individual patients, a physiological resilience matrix is ​​provided.

[0038] This is a vector representing changes in physiological state.

[0039] The total stress vector experienced by the patient;

[0040] Inverse calculation yields the vector of observed physiological state changes. The individual physiological resilience matrix of a patient is obtained in the following ways: In the early stage of a patient's admission, the second feature set response data after receiving a standardized test drug dose is recorded through the baseline calibration module. The mapping relationship between the changes in physiological parameters and the total stress vector is fitted using linear regression or machine learning algorithms to generate an individualized physiological resilience matrix; or the matrix parameters are dynamically updated through a recursive algorithm based on the temporal fluctuations of the patient's real-time physiological data.

[0041] The pathological state analysis module is based on the system's preset pharmacological and physiological response baseline. Changes in current drug dosage Through pharmacokinetic / pharmacodynamic response functions:

[0042]

[0043] This serves as the baseline for pharmacological and physiological responses.

[0044] This represents the current change in drug dosage;

[0045] The sensitivity coefficient of patients to sedative drugs;

[0046] Calculate the predicted pharmacological stress vector ;

[0047] The pathological state analysis module calculates:

[0048]

[0049] This is a pathological stress vector;

[0050] To observe the total stress vector of physiological state;

[0051] To predict pharmacological stress vectors;

[0052] Obtain the pathological stress vector and combine it with a dynamically adjusted pathological stress threshold. Generate pathological indices that characterize the degree of disease progression.

[0053]

[0054] This is a pathological index, which quantitatively represents the degree of deterioration in the patient's current condition;

[0055] The magnitude of the pathological stress vector;

[0056] The pathological stress threshold is dynamically adjusted.

[0057] The evolutionary trend prediction module is based on the pathological stress vector output by the pathological state analysis module. The system integrates time-series data with auxiliary clinical parameters such as body temperature (T) and inflammatory marker I to construct a pathological evolution trend dataset. The evolution trend prediction module incorporates a Long Short-Term Memory (LSTM) network model. The pathological evolution trend dataset is input into the LSTM model, which learns and identifies change patterns through time-series feature learning, predicting and outputting risk predictors characterizing the patient's future pathological state.

[0058]

[0059] As a risk predictor;

[0060] Risk weights set for clinical experience;

[0061] For the LSTM model, various physiological parameters are... The predicted value;

[0062] This refers to the change in brainwave entropy.

[0063] This represents the change in cerebral perfusion pressure.

[0064] Body temperature;

[0065] Inflammation markers;

[0066] Closed-loop intervention decision-making module combined with pathological indices Risk predictors Based on the preset second mapping relationship, hierarchical decision calculation is performed; when the pathological index At that time, the closed-loop intervention decision-making module triggers a protective strategy, calculating the drug dosage adjustment amount as follows:

[0067] ,

[0068] in This is for adjusting the drug dosage;

[0069] The coefficient is the high-priority negative feedback coefficient.

[0070] This represents the current drug infusion rate;

[0071] The dynamic threshold for the alarm period;

[0072] The closed-loop intervention decision module uses the drug dosage adjustment amount to update the current drug infusion rate, through:

[0073]

[0074] This is for adjusting the drug dosage;

[0075] This represents the current drug infusion rate;

[0076] For the new recommended drug infusion rate;

[0077] Calculate new recommended drug infusion rates to achieve closed-loop control;

[0078] In this embodiment, the physiological feature extraction module, through multimodal data fusion technology, can accurately collect and process multidimensional physiological state information of patients, significantly improving the comprehensiveness and accuracy of physiological state assessment compared to traditional single-parameter monitoring methods. The pathological state analysis module, through decoupling analysis technology, can effectively distinguish between pharmacological stress and pathological stress, avoiding the problem of confusion between drug effects and pathological changes in traditional methods, and providing clinicians with a more accurate basis for disease assessment. The evolution trend prediction module, through an LSTM deep learning model, can predict the future pathological state evolution trend of patients based on historical time-series data, realizing the transformation from passive monitoring to active early warning compared to traditional static assessment methods, effectively reducing the risk of secondary brain injury. The closed-loop intervention decision module, through intelligent decision-making algorithms, can automatically adjust drug infusion strategies according to the real-time status of patients, significantly improving the accuracy and timeliness of treatment compared to traditional manual adjustment methods, reducing the workload of medical staff, and improving the overall efficiency and safety of intensive care.

[0079] Example 2:

[0080] The physiological feature extraction module collects a first feature set that further includes mean arterial pressure and intracranial pressure; the first mapping relationship is specifically generated by subtracting mean arterial pressure from intracranial pressure to generate cerebral perfusion pressure.

[0081] The pathological state decoupling module is used to calculate the predicted pharmacological stress vector. The process includes: first, obtaining the preset pharmacological and physiological response baseline to characterize individual drug sensitivity; then, combining the current drug dose change, predicting the theoretical physiological state change caused by the dose change, and setting the theoretical physiological state change as the predicted pharmacological stress vector.

[0082] The first feature set collected by the physiological feature extraction module further includes mean arterial pressure (MAP) obtained through an invasive arterial catheter and intracranial pressure (ICP) obtained through an intracranial pressure monitor. The sampling frequency is set to 100-1000Hz to ensure the real-time performance and accuracy of the data. The first mapping relationship in the physiological feature extraction module is specifically implemented by subtracting the mean arterial pressure (MAP) from the intracranial pressure (ICP) to generate the cerebral perfusion pressure.

[0083]

[0084] Cerebral perfusion pressure;

[0085] Mean arterial pressure;

[0086] Intracranial pressure;

[0087] The calculation process is executed at a millisecond frequency in the real-time data processing unit to ensure the continuity and accuracy of the cerebral perfusion pressure values;

[0088] The process of calculating and predicting the pharmacological stress vector in the pathological state analysis module first obtains the pre-set pharmacological and physiological response baseline to characterize the overall drug sensitivity. This baseline, defined using data from the initial 4-6 hours of stable period after admission, reflects the individual patient's baseline response to standard doses of sedative medication; the pathological status analysis module incorporates current drug dosage changes. Through pharmacokinetic models:

[0089]

[0090] This represents the pharmacological stress vector.

[0091] This serves as the baseline for pharmacological and physiological responses.

[0092] This represents the current change in drug dosage;

[0093] The sensitivity coefficient of patients to sedative drugs;

[0094] Calculate the predicted pharmacological stress vector To predict the theoretical physiological state changes caused by this dose change, among which This represents the PK / PD response function based on the Michaelis equation or linear regression. This represents the patient's hemodynamic and neurodepressive sensitivity coefficients to sedative drugs; the pathological state analysis module sets this theoretical physiological state change as a predictive pharmacological stress vector. This is used for subsequent decoupling analysis from the observed total stress vector;

[0095] In this embodiment, the physiological feature extraction module accurately collects mean arterial pressure and intracranial pressure, and calculates cerebral perfusion pressure in real time based on the first mapping relationship. This provides critical care physicians with key indicators for assessing cerebral blood flow perfusion. Compared with traditional indirect estimation methods, it significantly improves the accuracy and reliability of cerebral perfusion status monitoring, playing a particularly important role in the management of intracranial pressure in patients with traumatic brain injury. The pathological state analysis module obtains individualized pharmacological and physiological response baselines and combines them with current drug dosage changes to predict theoretical physiological state changes. This effectively establishes patient-specific drug effect models. Compared with traditional drug effect assessment methods based on population averages, it significantly improves the individualized accuracy of pharmacological stress prediction, providing important technical support for precision medicine. It effectively avoids drug effect assessment bias caused by individual differences and improves the safety and effectiveness of drug treatment in intensive care.

[0096] Example 3:

[0097] The process of generating pathological indices by the pathological state decoupling module includes: first, calculating the magnitude of the pathological stress vector; second, obtaining the dynamically adjusted pathological stress threshold; and finally, obtaining the pathological index by calculating the ratio of the magnitude of the pathological stress vector to the pathological stress threshold.

[0098] It also includes a monitoring status classification module, which is used to: classify the patient's current stage as a monitoring status stage, such as a stable stage, an observation stage, or an alarm stage, based on the historical fluctuation range of the second feature set; and a pathological status decoupling module to dynamically adjust the pathological stress threshold based on the monitoring status stage output by the monitoring status classification module.

[0099] The process of generating pathological indices by the pathological state analysis module first calculates the pathological stress vector. Length of the module This calculation is performed using vector norm operations:

[0100]

[0101] Pathological stress vector The modulus length;

[0102] Let i be the component of the pathological stress vector in the i-th dimension.

[0103] n is the number of dimensions of the vector;

[0104] The pathological state analysis module then obtains the dynamically adjusted pathological stress threshold. This threshold is adjusted in real time according to the monitoring status stage, and the calculation formula is as follows:

[0105] Base threshold

[0106] in The pathological stress threshold is dynamically adjusted.

[0107] This represents the threshold coefficient corresponding to the current monitoring status stage.

[0108] The pathological state analysis module ultimately obtains the pathological index by calculating the ratio of the magnitude of the pathological stress vector to the pathological stress threshold.

[0109]

[0110] This is a pathological index, which quantitatively represents the degree of deterioration in the patient's current condition;

[0111] The magnitude of the pathological stress vector;

[0112] The pathological stress threshold is dynamically adjusted.

[0113] The monitoring status classification module assesses the stability of patients' physiological parameters through statistical analysis based on the historical fluctuation range of the second feature set. The module employs a moving window standard deviation algorithm, where the EEG entropy value... or cerebral perfusion pressure When the standard deviation of fluctuation within 72 hours is less than 20% of the baseline value, the patient's current stage is defined as the stable stage. When the standard deviation of fluctuation is between 20% and 50% of the baseline value, it is defined as the observation phase. When the standard deviation of the fluctuation exceeds 50% of the baseline value, an alarm phase is defined as follows: The pathological state analysis module dynamically adjusts the pathological stress threshold based on the monitoring state stages output by the monitoring state division module, with the baseline threshold set during the stable phase. Base threshold during the observation phase Basic threshold during the alarm phase ;

[0114] In this embodiment, the pathological state analysis module generates a quantitative pathological index by calculating the magnitude of the pathological stress vector and performing a ratio operation with a dynamically adjusted threshold. Compared with traditional qualitative assessment methods, this significantly improves the objectivity and accuracy of assessing the degree of disease deterioration, providing clinicians with quantifiable disease monitoring indicators and effectively supporting precise diagnosis and treatment decisions in the intensive care unit. The monitoring state classification module automatically identifies the patient's monitoring state stage through intelligent analysis based on historical fluctuation ranges. Compared with traditional manual judgment methods, this significantly improves the consistency and timeliness of monitoring state assessment. At the same time, by dynamically adjusting the pathological stress threshold, it realizes individualized and staged pathological assessment standards, effectively adapting to the dynamic changes in the condition of critically ill patients, improving the accuracy and effectiveness of monitoring management, and providing differentiated monitoring strategy support for patients at different disease stages.

[0115] Example 4:

[0116] The evolution trend prediction module has a built-in long short-term memory network model. The evolution trend prediction module is also used to: integrate the time series data of pathological stress vectors with the time series data of auxiliary clinical parameters such as body temperature and inflammatory indicators to construct a pathological evolution feature dataset; input the pathological evolution feature dataset into the long short-term memory network model, and its output is the risk prediction factor.

[0117] The evolutionary trend prediction module incorporates a Long Short-Term Memory (LSTM) network model. This model employs a multi-layer LSTM architecture, including an input layer, hidden layers, and an output layer. The hidden layers have 128 neurons, effectively capturing long-term dependencies and short-term fluctuations in time-series data. The evolutionary trend prediction module also incorporates pathological stress vectors. The time-series data of patients were integrated with the time-series data of auxiliary clinical parameters such as body temperature (T) and inflammatory marker I to construct a pathological evolution trend dataset. This dataset contains continuous monitoring data of patients over the past 24-48 hours, with a sampling interval of 15 minutes to ensure the temporal continuity and representativeness of the data.

[0118] When constructing the pathological evolution trend dataset, the evolutionary trend prediction module includes pathological stress vectors. The time series data is used as the primary feature, while time-series changes in body temperature (T), timed detection results of inflammatory marker I, and monitoring status stage labels are integrated as auxiliary features. The evolutionary trend prediction module standardizes the input data using the Z-score standardization method.

[0119]

[0120] The original data value;

[0121] The mean;

[0122] Standard deviation;

[0123] These are the standardized data values;

[0124] The evolutionary trend prediction module inputs the standardized pathological evolution trend dataset into the Long Short-Term Memory (LSTM) network model. The LSTM model learns and identifies patterns of change in pathological states through temporal features and outputs risk prediction factors. ,

[0125] As a risk predictor;

[0126] Risk weights set for clinical experience ;

[0127] For the LSTM model, various physiological parameters are... The predicted value;

[0128] This refers to the change in brainwave entropy.

[0129] This represents the change in cerebral perfusion pressure.

[0130] Body temperature;

[0131] Inflammation markers;

[0132] In this embodiment, the evolutionary trend prediction module, through its built-in long short-term memory network model, can effectively handle the complex temporal changes in the physiological parameters of critically ill patients. Compared with traditional linear prediction methods, it significantly improves the accuracy and reliability of pathological state evolution trend prediction, especially in capturing nonlinear change patterns and long-term dependencies. By integrating pathological stress vector temporal data with auxiliary clinical parameters, the evolutionary trend prediction module constructs a multi-dimensional pathological evolution trend dataset. Compared with single-parameter prediction methods, it significantly enhances the comprehensiveness and robustness of the prediction model, and can more accurately reflect the complex evolutionary process of the patient's pathological state. The risk predictor output by the evolutionary trend prediction module provides critical care physicians with a prospective disease assessment tool. Compared with traditional passive monitoring methods, it realizes the transformation from reactive treatment to preventive intervention, effectively reducing the risk of serious complications such as secondary brain injury, improving the initiative and predictability of intensive care, and providing important scientific basis for clinical decision-making.

[0133] Example 5:

[0134] The closed-loop intervention decision module calculates based on the second mapping relationship as follows: setting high-priority warning thresholds and medium-priority warning thresholds; when the pathological index is higher than the high-priority warning threshold, a protective strategy is triggered, and the drug dosage adjustment amount is set to the first negative value; when the pathological index is lower than the high-priority warning threshold but higher than the medium-priority warning threshold, a conservative strategy is triggered, and the drug dosage adjustment amount is set to the second negative value.

[0135] When the pathological index is not higher than the medium-priority warning threshold, the closed-loop intervention decision module is also used to: obtain the stage risk threshold corresponding to the current monitoring status stage; calculate the deviation between the risk prediction factor and the stage risk threshold by comparing the risk prediction factor and the stage risk threshold, and generate the drug dosage adjustment amount based on the deviation.

[0136] The closed-loop intervention decision-making module calculates based on the second mapping relationship, setting a high-priority warning threshold of 1.8 and a medium-priority warning threshold of 1.0. These two thresholds are determined based on statistical analysis of a large amount of clinical data and can effectively distinguish pathological states of different severity. The closed-loop intervention decision-making module uses pathological indices... When the value exceeds the high-priority warning threshold of 1.8, a protective strategy is triggered, setting the drug dosage adjustment to the first negative value.

[0137]

[0138] This is for adjusting the drug dosage;

[0139] The high-priority negative feedback coefficient. ;

[0140] This is the current infusion rate of the sedative medication;

[0141] The dynamic threshold for the alarm period. ;

[0142] This is a pathological index, which quantitatively represents the degree of deterioration in the patient's current condition;

[0143] This strategy aims to rapidly reduce drug dosage to alleviate pharmacological inhibition; the closed-loop intervention decision module considers pathological indices. When the value is below the high-priority warning threshold of 1.8 but above the medium-priority warning threshold of 1.0, a conservative strategy is triggered, and the drug dosage adjustment is set to the second negative value.

[0144]

[0145] This is for adjusting the drug dosage;

[0146] This represents the coefficient for moderate-intensity negative feedback. ;

[0147] The dynamic threshold for the observation period.

[0148] This is a pathological index, which quantitatively represents the degree of deterioration in the patient's current condition;

[0149] This strategy employs moderate dose adjustments to balance therapeutic efficacy and safety;

[0150] Closed-loop intervention decision module when pathological index When the risk level is not higher than the medium-priority warning threshold of 1.0, obtain the phased risk threshold corresponding to the current monitoring status phase; the stable phase corresponds to... The observation phase corresponds to Alarm phase corresponding The closed-loop intervention decision-making module compares risk prediction factors. With phased risk threshold Calculate the deviation between the two. Based on this deviation, a drug dosage adjustment amount is generated.

[0151]

[0152] This is for adjusting the drug dosage;

[0153] Indicates the predictive adjustment coefficient. ;

[0154] Deviation between risk predictor factors and stage thresholds;

[0155] : The threshold coefficient corresponding to the current monitoring status stage;

[0156] when The adjustment amount is set to a negative value to reduce the dosage. The adjustment value is set to a positive value to appropriately increase the dosage;

[0157] In this embodiment, the closed-loop intervention decision-making module establishes a graded response intelligent decision-making mechanism by setting high-priority and medium-priority warning thresholds. Compared with the traditional single-threshold judgment method, this significantly improves the precision and adaptability of the intervention strategy, and can provide differentiated treatment plans according to the severity of the condition. The closed-loop intervention decision-making module realizes a gradient treatment mode from aggressive intervention to mild adjustment through graded triggering of protective and conservative strategies. Compared with the traditional fixed-dose adjustment method, this significantly improves the safety and effectiveness of drug treatment and effectively avoids the problems of overtreatment or undertreatment. When the pathological index is low, the closed-loop intervention decision-making module calculates the deviation by comparing the risk predictor factor with the stage risk threshold and generates the adjustment amount, realizing a prospective intervention based on predictive assessment. Compared with the traditional reactive treatment mode, this significantly improves the initiative and foresight of intensive care, and can adjust the treatment strategy in time before the condition deteriorates, effectively preventing the occurrence of serious complications and improving the treatment effect and prognosis of patients.

[0158] Example 6:

[0159] It also includes a baseline calibration module, which is used to: record the response changes of the second feature set after the patient receives a standardized test drug dose in the early stage of hospital admission; process the response changes and set them as the pharmacological and physiological response baseline for use by the pathological state decoupling module;

[0160] After determining the drug dosage adjustment amount, the closed-loop intervention decision module is also used to: sum the current drug infusion rate with the drug dosage adjustment amount to generate a new recommended drug infusion rate; and output the new recommended drug infusion rate to the drug infusion device to execute closed-loop control.

[0161] The baseline calibration module records changes in the second characteristic set of patients' responses after receiving a standardized test drug dose during the 4-6 hour stabilization period following admission. The baseline calibration module uses a standard dose of propofol (1 mg / kg intravenously) as the standardized test drug dose and monitors EEG entropy values ​​within 30 minutes after administration. Cerebral perfusion pressure The changes in physiological parameters such as mean arterial pressure (MAP) and intracranial pressure (ICP) are analyzed and their patterns observed. The baseline calibration module standardizes and models these response changes using data processing algorithms, processes them with a pharmacokinetic / pharmacodynamic model, and sets them as the baseline for the pharmacological and physiological response.

[0162]

[0163] in : Baseline of pharmacological and physiological response;

[0164] Patient drug sensitivity coefficient;

[0165] : Vector of physiological parameter changes induced by standard dose;

[0166] : Response time constant;

[0167] This baseline data is used by the pathological state analysis module for subsequent calculation of pharmacological stress vectors.

[0168] The closed-loop intervention decision-making module obtains the drug dosage adjustment amount Then, the current drug infusion rate will be... The new recommended drug infusion rate is generated by summing the drug dosage adjustment amount with the drug dosage adjustment amount.

[0169]

[0170] For the new recommended drug infusion rate;

[0171] This represents the current drug infusion rate;

[0172] This is for adjusting the drug dosage;

[0173] The closed-loop intervention decision module performs safety verification on the calculation results to ensure that the new recommended drug infusion rate is within the clinically safe range. The minimum infusion rate is set at 0.5 mg / kg / h and the maximum infusion rate is set at 4.0 mg / kg / h. When the calculation result exceeds the safe range, it is automatically limited to the boundary value. The closed-loop intervention decision module outputs the new recommended drug infusion rate to the drug infusion device through a standardized communication protocol. After receiving the instruction, the infusion device automatically adjusts the infusion parameters to achieve precise control and closed-loop regulation of drug dosage. The entire adjustment process is completed within 5 minutes, ensuring the timeliness and continuity of treatment.

[0174] In this embodiment, the baseline calibration module establishes an individualized pharmacological and physiological response baseline by conducting standardized drug dosage response assessments at the initial stage of patient admission. Compared with traditional drug effect assessment methods based on population averages, this significantly improves the individualized accuracy and reliability of pharmacological stress prediction, providing an accurate reference standard for subsequent decoupling analysis and effectively solving the problem of individual differences affecting drug effect assessment. The closed-loop intervention decision module generates a new suggested drug infusion rate by summing the current drug infusion rate with the calculated adjustment amount and outputting it to the drug infusion device, realizing a complete closed loop from decision calculation to execution control. Compared with traditional manual adjustment methods, this significantly improves the accuracy, timeliness, and consistency of drug dosage adjustment, reduces human error and delay, and enhances the automation level and treatment efficiency of intensive care. The entire system, through the complete process of baseline calibration, status monitoring, trend prediction, intelligent decision-making, and automatic execution, constructs a highly integrated intelligent intensive care management platform. Compared with traditional decentralized monitoring methods, this significantly improves the overall monitoring quality and patient safety level of the intensive care unit, providing important technical support for the digital transformation of intensive care medicine and the development of precision medicine.

[0175] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A critical care patient monitoring and management system based on multimodal data fusion, characterized in that, include: The physiological feature extraction module is used to collect a first feature set that represents the patient's multi-dimensional physiological state, and based on a preset first mapping relationship, process the first feature set into a second feature set that includes EEG entropy value and cerebral perfusion pressure. The pathological state decoupling module is used to back-calculate the observed physiological state change vector based on the second feature set, and calculate the predicted pharmacological stress vector based on the system's preset pharmacological and physiological response baseline and the current drug dose change; by extracting the predicted pharmacological stress vector from the observed physiological state change vector, the pathological stress vector is obtained, and then combined with the dynamically adjusted pathological stress threshold, a pathological index characterizing the degree of disease deterioration is generated. The evolution trend prediction module is used to predict and output risk prediction factors characterizing the patient's future pathological state based on the time-series data of the pathological stress vector output by the pathological state decoupling module. The closed-loop intervention decision module is used to combine the pathological index with the risk prediction factor and calculate the drug dosage adjustment amount according to the preset second mapping relationship. The drug dosage adjustment amount is then used to update the current drug infusion rate to achieve closed-loop control.

2. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1, characterized in that, The first feature set collected by the physiological feature extraction module further includes mean arterial pressure and intracranial pressure; the first mapping relationship is specifically: the cerebral perfusion pressure is generated by subtracting the mean arterial pressure from the intracranial pressure.

3. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1 or 2, characterized in that, The process by which the pathological state decoupling module calculates the predicted pharmacological stress vector includes: first, obtaining the preset pharmacological physiological response baseline to characterize individual drug sensitivity; then, combining the current drug dose change, predicting the theoretical physiological state change caused by the dose change, and setting the theoretical physiological state change as the predicted pharmacological stress vector.

4. A critical care patient monitoring and management system based on multimodal data fusion according to claim 3, characterized in that, The process by which the pathological state decoupling module generates the pathological index includes: firstly, calculating the magnitude of the pathological stress vector; secondly, obtaining the dynamically adjusted pathological stress threshold; and finally, obtaining the pathological index by calculating the ratio of the magnitude of the pathological stress vector to the pathological stress threshold.

5. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1 or 4, characterized in that, It also includes a monitoring status classification module, which is used to: classify the patient's current stage as a monitoring status stage, such as a stable stage, an observation stage, or an alarm stage, based on the historical fluctuation range of the second feature set; and the pathological state decoupling module dynamically adjusts the pathological stress threshold based on the monitoring status stage output by the monitoring status classification module.

6. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1, characterized in that, The evolution trend prediction module has a built-in long short-term memory network model; the evolution trend prediction module is also used to: integrate the time series data of the pathological stress vector with the time series data of auxiliary clinical parameters such as body temperature and inflammatory indicators to construct a pathological evolution feature dataset; input the pathological evolution feature dataset into the long short-term memory network model, and its output is the risk prediction factor.

7. A critical care patient monitoring and management system based on multimodal data fusion according to claim 6, characterized in that, The closed-loop intervention decision module performs calculations based on the second mapping relationship, including: setting a high-priority warning threshold and a medium-priority warning threshold; when the pathological index is higher than the high-priority warning threshold, a protective strategy is triggered, and the drug dosage adjustment amount is set to a first negative value; when the pathological index is lower than the high-priority warning threshold but higher than the medium-priority warning threshold, a conservative strategy is triggered, and the drug dosage adjustment amount is set to a second negative value.

8. A critical care patient monitoring and management system based on multimodal data fusion according to claim 7, characterized in that, When the pathological index is not higher than the medium-priority warning threshold, the closed-loop intervention decision module is also used to: obtain the stage risk threshold corresponding to the current monitoring status stage; By comparing the risk prediction factor with the staged risk threshold, the deviation between the two is calculated, and the drug dosage adjustment amount is generated based on the deviation.

9. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1, characterized in that, It also includes a baseline calibration module, which is used to: record the response changes of the second feature set after the patient receives a standardized test drug dose in the early stage of hospital admission; process the response changes and set them as the pharmacological and physiological response baseline for use by the pathological state decoupling module.

10. A critical care patient monitoring and management system based on multimodal data fusion according to claim 1, characterized in that, After obtaining the drug dosage adjustment amount, the closed-loop intervention decision module is further configured to: sum the current drug infusion rate with the drug dosage adjustment amount to generate a new recommended drug infusion rate; and output the new recommended drug infusion rate to the drug infusion device to perform closed-loop control.

Images