Postoperative care risk warning system for neurosurgical patients

Abstract

This invention relates to the field of medical monitoring and early warning technology, and discloses a postoperative nursing risk early warning system for neurosurgical patients. This system continuously receives real-time monitoring streams of the patient's electroencephalogram (EEG), intracranial pressure (ICP), and surface electromyography (EMG) to construct a dynamic interactive network of neurophysiological and intracranial biomechanical states, characterizing the coupling relationship between EEG rhythm, pressure transmission, and tissue compliance in the postoperative cranial fenestration area. The system extracts three core parameters from this network: cortical excitability, cerebrospinal fluid circulation load, and intracranial compensatory space. It then calculates the risk of brain tissue perfusion and neural structure compression, generating corresponding nursing early warning signals and physiological regulatory pathways. This approach achieves deep coupling of multiple physiological signals and mechanistic risk early warning, enabling earlier and more accurate identification of postoperative complication risks.

Description

Technical Field

[0001] This invention relates to the field of medical monitoring and early warning technology, specifically to a postoperative care risk early warning system for neurosurgical patients. Background Technology

[0002] Postoperative neurosurgical patients require close monitoring to prevent serious complications such as cerebral edema, intracranial rebleeding, and cerebral ischemia. Currently, widely used clinical monitoring techniques primarily rely on setting fixed thresholds for single or multiple physiological parameters to trigger independent alarms. These techniques treat brain electrical activity, intracranial pressure, and biomechanical status as isolated indicators for parallel monitoring and assessment.

[0003] Existing independent parameter threshold alarm methods have shortcomings. They sever the inherent, dynamic interaction between neurophysiological signals and intracranial biomechanical states. The monitoring method remains at the level of simple capture of surface phenomena, with delayed warnings and unclear mechanisms, making it difficult to identify early, coordinated deviations from intracranial homeostasis before irreversible brain tissue damage occurs.

[0004] A method is needed to deeply integrate and analyze the dynamic interactions of multi-source physiological signals to overcome the limitations of existing technologies in assessing complication risks through superficial and fragmented methods. This invention aims to address how to construct a coupled network of neurophysiology and intracranial biomechanics in real time, and how to extract deeper, derived parameters reflecting the core dimensions of intracranial homeostasis from this network. This would enable earlier, more precise, and more mechanism-specific postoperative risk warnings and nursing interventions. Summary of the Invention

[0005] The purpose of this invention is to provide a postoperative care risk warning system for neurosurgical patients to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a postoperative care risk warning system for neurosurgical patients, the system comprising:

[0007] The multi-source stream receiving module continuously receives real-time physiological monitoring streams of the patient from multiple monitoring devices, including electroencephalogram (EEG) waveforms, intracranial pressure waveforms, and surface electromyography (EMG) signals.

[0008] A dynamic interactive network module is constructed to establish a dynamic interactive network between neurophysiology and intracranial biomechanical state. The dynamic interactive network is used to characterize the coupling relationship between brain electrical rhythm fluctuations, intracranial pressure transmission and tissue compliance in the cranial fenestration area.

[0009] The state parameter parsing module parses three types of core state parameters from the dynamic interactive network, reflecting the excitability of the cerebral cortex, the cerebrospinal fluid circulation load, and the cranial cavity compensation space.

[0010] The risk warning and analysis module inputs the three types of core state parameters into the intracranial steady-state deviation warning unit to calculate the brain tissue perfusion risk and neural structure compression risk at the current moment.

[0011] The early warning path generation module generates nursing early warning signals and corresponding physiological regulation paths based on the brain tissue perfusion risk and neural structure compression risk.

[0012] Preferably, the construction of the dynamic interactive network between neurophysiology and intracranial biomechanical state includes:

[0013] The EEG waveform was rhythmically decoupled to separate the slow wave oscillation energy, fast wave rhythm synchronization index, and epileptiform discharge density.

[0014] Simultaneously, morphological decomposition of the intracranial pressure waveform was performed to identify pulse wave transmission delay, respiratory wave coupling strength, and frequency of abnormal high-amplitude waves.

[0015] Extract the spatiotemporal distribution characteristics of background muscle tone activity and abnormal myoclonic bursts from the surface electromyography signals;

[0016] Establish a correlation mapping between the slow wave oscillation energy of EEG and the transmission delay of intracranial pressure pulse waves, and generate a regulatory loop model that reflects the automatic regulation function of cerebral blood flow.

[0017] Cross-analysis of intracranial pressure respiratory wave coupling strength and muscle tone background activity level generates a respiratory mechanics coupling model that reflects the influence of changes in thoracic and abdominal pressure on intracranial pressure.

[0018] By aligning epileptiform discharge density, abnormal high-amplitude wave frequency of intracranial pressure, and abnormal myoclonic burst characteristics with multi-source events and making causal inferences, a pathological event chain model reflecting the transmission of abnormal nerve excitability to intracranial mechanical state is generated.

[0019] By integrating the regulatory loop model, respiratory mechanics coupling model, and pathological event chain model, a dynamic interactive network of neurophysiology and intracranial mechanical state is formed.

[0020] Preferably, the three types of core state parameters reflecting cortical excitability, cerebrospinal fluid circulation load, and cranial cavity compensatory space are extracted from the dynamic interactive network, including:

[0021] Based on the phase relationship between slow wave oscillation energy and pulse wave delay in the aforementioned regulatory loop model, the cerebral blood flow regulatory reserve index is calculated. The cerebral blood flow regulatory reserve index is used to quantify the metabolic basis of cerebral cortex excitability.

[0022] Based on the coordinated change pattern of respiratory wave coupling intensity and muscle tone level in the respiratory mechanics coupling model, the indirect estimate of central venous pressure in the thoracic cavity is calculated, and combined with the baseline pressure level of intracranial pressure waveform, the pressure gradient load of cerebrospinal fluid circulation is calculated.

[0023] By analyzing the sequence and intensity ratio of neural electrical events and intracranial pressure events in the pathological event chain model, the elasticity coefficient of intracranial volumetric pressure response is identified. The elasticity coefficient is used to characterize the remaining capacity of the current cranial cavity compensatory space.

[0024] Preferably, the step of inputting the three types of core state parameters into the intracranial steady-state deviation early warning unit to calculate the current brain tissue perfusion risk and neural structure compression risk includes:

[0025] The cerebral blood flow regulation reserve index is input into the perfusion risk assessment subunit. Combined with the patient's baseline cerebral blood flow velocity data, the stability boundary of cerebral perfusion pressure under different blood pressure fluctuation scenarios is simulated, thereby calculating the risk probability of deviating from the stability boundary, which is used as the cerebral tissue perfusion risk.

[0026] The pressure gradient load of the cerebrospinal fluid circulation is input into the circulation load assessment subunit. Combined with the preoperative ventricular size data, the balance between the current cerebrospinal fluid absorption resistance and the generation rate is calculated, and the trend of ventricular volume change is predicted.

[0027] The remaining capacity of the cranial cavity compensation space is input into the space compensation assessment subunit. Combined with the degree of brain tissue displacement shown in postoperative imaging, a cranial volume pressure relationship curve is constructed, and the current state is located on the cranial volume pressure relationship curve.

[0028] By combining the trend of changes in ventricular volume with the position of the current state on the intracranial volume-pressure relationship curve, the probability of irreversible structural displacement or brain herniation of brain tissue is calculated as the risk of compression of the neural structure.

[0029] Preferably, the step of inputting the cerebral blood flow regulatory reserve index into the perfusion risk assessment subunit, combined with the patient's baseline cerebral blood flow velocity data, simulates the stability boundary of cerebral perfusion pressure under different blood pressure fluctuation scenarios, thereby calculating the risk probability of deviating from the stability boundary as the brain tissue perfusion risk, including:

[0030] In the perfusion risk assessment subunit, a set of blood pressure fluctuation simulation parameters are pre-set, which cover a continuous range of variation from low blood pressure to high blood pressure.

[0031] The baseline cerebral blood flow velocity data measured at the time of admission or before surgery is used as the baseline data and coupled with the real-time acquired cerebral blood flow regulatory reserve index.

[0032] Based on the physiological model of automatic regulation of cerebral blood flow, the cerebral blood flow regulation reserve index is used as the core regulation capability parameter of the model to simulate the corresponding dynamic change trajectory of cerebral perfusion pressure when the system blood pressure changes within a preset fluctuation parameter range.

[0033] The upper and lower limits of blood pressure that can maintain cerebral perfusion pressure within a safe range are identified from the simulation results, and these upper and lower limits of blood pressure are defined as the stability boundaries of cerebral perfusion pressure.

[0034] Calculate the closest distance between the patient's actual blood pressure value at the current moment and the stability boundary, and map the distance value to a probability value through a preset risk transformation function. The probability value is the risk probability of deviating from the stability boundary, and is used as the final brain tissue perfusion risk output.

[0035] Preferably, the step of inputting the pressure gradient load of the cerebrospinal fluid circulation into the circulation load assessment subunit, combining it with preoperative ventricular size data, calculating the current balance between cerebrospinal fluid absorption resistance and generation rate, and predicting the trend of ventricular volume changes includes:

[0036] In the cyclic load assessment subunit, the ventricular size data measured by the patient's preoperative imaging examination are retrieved and quantified as the initial ventricular volume;

[0037] The pressure gradient load value of cerebrospinal fluid circulation obtained from real-time analysis is substituted into the circulation model based on the differential equation of cerebrospinal fluid dynamics.

[0038] The cyclic model uses cerebrospinal fluid absorption resistance and generation rate as key variables. By iteratively fitting the current pressure gradient load value with the theoretical pressure value calculated by the model, the most matching cerebrospinal fluid absorption resistance value and generation rate value at the current moment can be derived.

[0039] By comparing the relationship between the most suitable cerebrospinal fluid absorption resistance value and the generation rate value at the current moment, if the absorption resistance is significantly greater than the generation rate, the equilibrium state is determined to be inclined towards cerebrospinal fluid accumulation; if the generation rate is significantly greater than the absorption resistance, the equilibrium state is determined to be inclined towards excessive cerebrospinal fluid loss; if the two are similar, it is determined to be basically balanced.

[0040] Based on the determined equilibrium state tendency and the specific numerical difference between absorption resistance and generation rate, the expected change in ventricular volume and direction of change in a preset time period are calculated using an empirical formula for the change of ventricular volume over time. The expected change and direction of change together constitute the predicted trend of ventricular volume change.

[0041] Preferably, the step of generating nursing early warning signals and corresponding physiological regulation pathways based on the brain tissue perfusion risk and neural structure compression risk includes:

[0042] The brain tissue perfusion risk is compared with a preset perfusion risk threshold. When the threshold is exceeded, a brain perfusion maintenance warning is triggered, and an adjustment path is generated with the goal of increasing brain perfusion pressure. The adjustment path includes a recommended target mean arterial pressure range and blood pressure adjustment rate.

[0043] The risk of neural structure compression is compared with a preset structural compression risk threshold. When the threshold is exceeded, an intracranial pressure control warning is triggered, and an adjustment path aimed at reducing intracranial volume is generated. The adjustment path includes suggested head position angle, hyperventilation parameters, and the timing of activation of osmotic dehydrating agents.

[0044] When the risk of brain perfusion and the risk of neural structure compression both exceed their respective thresholds, an intracranial homeostasis imbalance warning is triggered, and a composite regulatory pathway that takes into account the contradiction between brain perfusion and intracranial volume is generated. The composite regulatory pathway specifies the priority and operation window for implementing measures to reduce intracranial pressure in stages while maintaining the minimum permissible brain perfusion pressure.

[0045] Preferably, the generation of the regulatory pathway aimed at increasing cerebral perfusion pressure includes:

[0046] Based on the specific value of the cerebral blood flow regulation reserve index, it is divided into three levels: sufficient reserve, reduced reserve, and depleted reserve.

[0047] For levels of adequate blood reserve, a plan is set to gradually increase the mean arterial pressure, and it is stipulated that the increase in cerebral blood flow velocity should be monitored simultaneously during the adjustment process to ensure that it is proportional.

[0048] For the level of reduced reserve, a step-by-step plan to gradually increase the mean arterial pressure is set up, and it is stipulated that after each step increase, the improvement of the slow wave oscillation energy of the EEG should be observed to determine whether to proceed to the next step.

[0049] For each level of reserve depletion, an emergency volume expansion and vasopressor regimen is generated, and it is stipulated that central venous pressure and surface electromyography signals must be monitored in conjunction to prevent heart failure or worsening of myocardial edema caused by rapid volume expansion.

[0050] Preferably, the generation of the regulatory pathway aimed at reducing intracranial volume includes:

[0051] The main sources of the pressure gradient load in the cerebrospinal fluid circulation were analyzed to distinguish whether the main cause was impaired cerebrospinal fluid absorption or obstructed venous return.

[0052] For cases where the main issue is impaired cerebrospinal fluid absorption, a pathway is developed that focuses on regulating the rate of cerebrospinal fluid production. This includes recommendations on the types of drugs that reduce cerebrospinal fluid production, initial doses, and methods for adjusting the dose based on changes in the coupling strength of respiratory waves in intracranial pressure waveforms.

[0053] For cases where venous return is primarily obstructed, a pathway is generated with the optimization of intracranial venous return as its core. This includes specific operational steps and parameter combinations for reducing central venous pressure by adjusting the head of the bed, neck position, and controlling intrathoracic pressure.

[0054] Preferably, the specific operational steps and parameter combinations for reducing central venous pressure by adjusting the head of the bed, neck position, and controlling intrathoracic pressure include:

[0055] Raise the head of the bed from its initial position to the first target angle, and maintain this angle for the first preset observation time to monitor the decrease in the intensity of respiratory wave coupling in the intracranial pressure waveform;

[0056] If the reduction is not as expected, further adjust the patient's neck position to the midline and slightly extend it backward, while monitoring the tension of the neck and shoulder muscles in the electromyography signal on the body surface to ensure that it does not increase significantly.

[0057] After adjusting the patient's position, instruct them to use a slow ventilation mode that primarily involves abdominal breathing, and use a respiratory monitoring device to provide real-time feedback on the decrease in intrathoracic pressure.

[0058] Based on feedback data from intracranial pressure respiratory wave coupling intensity, neck and shoulder electromyographic tension, and intrathoracic pressure, the head of the bed angle, neck extension angle, and breathing instructions are dynamically fine-tuned to form an optimal combination of body position and respiratory parameters that reduces the estimated central venous pressure to the target range.

[0059] Compared with the prior art, the beneficial effects of the present invention are:

[0060] By constructing a dynamic interactive network between neurophysiology and intracranial biomechanics, the system achieves real-time mathematical modeling and simulation of the coupling relationship between brainwave rhythm fluctuations, intracranial pressure transmission, and tissue compliance in the cranial fenestration area. Instead of simply processing multiple signals in parallel, it performs real-time mathematical modeling of brainwave rhythm fluctuations, intracranial pressure transmission, and tissue compliance in the cranial fenestration area, simulating their interactions and feedback mechanisms. The system continuously receives multi-source data streams to dynamically update this coupled model. This enables the system to capture early coordinated change patterns that cannot be identified by single-parameter threshold alarms. Deep integration transforms previously isolated monitoring indicators into an organic system that reflects the overall evolution of intracranial physiological state.

[0061] By extracting three higher-order derived parameters from the aforementioned dynamic interactive network—cerebral cortical excitability, cerebrospinal fluid circulation load, and intracranial compensatory space—the system achieves a leap from phenomenological monitoring to mechanistic judgment. These parameters are endogenous results of model computation, quantifying the real-time status of neural metabolic demand, circulatory pathway pressure load, and physical buffer space, respectively. Risk analysis based on these deep parameters means that the early warning logic no longer relies on exceeding surface numerical limits but directly targets the potential risks of two core pathophysiological processes: brain tissue perfusion imbalance and neural structural compression. The resulting nursing early warning signals can be linked to more specific physiological imbalances, and the recommended physiological regulation pathways are more mechanistically targeted. For example, hemodynamic management is optimized for perfusion risks, and decompression strategies are adjusted for structural compression risks, thus forming a closed loop from deep state perception to precise intervention decisions. Attached Figure Description

[0062] Figure 1 This is a timing diagram of the postoperative care risk warning system for neurosurgical patients described in this invention;

[0063] Figure 2 Flowchart for constructing a dynamic interactive network;

[0064] Figure 3 A flowchart for risk analysis calculations;

[0065] Figure 4 Simulation diagram of the stability boundary of cerebral perfusion pressure under different cerebral blood flow regulation reserve levels;

[0066] Figure 5 Scatter plot showing the correlation between intracranial pressure respiratory wave coupling intensity and intrathoracic pressure at different head-of-bed angles. Detailed Implementation

[0067] 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.

[0068] Please see Figure 1This invention provides a postoperative nursing risk warning system for neurosurgical patients. The system includes: a multi-source stream receiving module that continuously receives real-time physiological monitoring streams from multiple monitoring devices, including electroencephalogram (EEG) waveforms, intracranial pressure waveforms, and surface electromyography (EMG) signals; a dynamic interactive network module that constructs a dynamic interactive network between neurophysiology and intracranial biomechanics, characterizing the coupling relationship between EEG rhythm fluctuations, intracranial pressure transmission, and tissue compliance in the cranial fenestration area; a state parameter parsing module that extracts three core state parameters reflecting cortical excitability, cerebrospinal fluid circulation load, and intracranial compensatory space from the dynamic interactive network; a risk warning analysis module that inputs these three core state parameters into an intracranial steady-state deviation warning unit to calculate the current brain tissue perfusion risk and neural structure compression risk; and a warning path generation module that generates nursing warning signals and corresponding physiological regulation paths based on the brain tissue perfusion risk and neural structure compression risk.

[0069] In one embodiment of the present invention, see [reference] Figure 2 The construction of the dynamic interactive network module includes rhythm decoupling of EEG waveforms, separating slow-wave oscillation energy, fast-wave rhythm synchronization index, and epileptiform discharge density. Simultaneously, morphological decomposition of intracranial pressure waveforms is performed to identify pulse wave transmission delay, respiratory wave coupling strength, and the frequency of abnormal high-amplitude waves. The spatiotemporal distribution characteristics of muscle tone background activity and abnormal myoclonic bursts in surface electromyography (EMG) signals are extracted. A correlation mapping between EEG slow-wave oscillation energy and intracranial pressure pulse wave transmission delay is established, generating a regulatory loop model reflecting the autoregulation function of cerebral blood flow. Cross-analysis of intracranial pressure respiratory wave coupling strength and muscle tone background activity level generates a respiratory mechanics coupling model reflecting the influence of changes in thoracic and abdominal pressure on intracranial pressure. Multi-source event alignment and causal inference are performed on epileptiform discharge density, intracranial pressure abnormal high-amplitude wave frequency, and abnormal myoclonic burst characteristics to generate a pathological event chain model reflecting the transmission of abnormal neural excitability to intracranial mechanical state. The regulatory loop model, respiratory mechanics coupling model, and pathological event chain model are integrated to form a dynamic interactive network between neurophysiology and intracranial mechanical state. The state parameter analysis module calculates the cerebral blood flow regulatory reserve index based on the phase relationship between slow wave oscillation energy and pulse wave delay in the regulatory loop model. This index is used to quantify the metabolic basis of cortical excitability. Based on the coordinated change pattern of respiratory wave coupling strength and muscle tone level in the respiratory mechanics coupling model, an indirect estimate of the central venous pressure in the thoracic cavity is calculated. Combined with the baseline pressure level of the intracranial pressure waveform, the pressure gradient load of cerebrospinal fluid circulation is calculated. By analyzing the sequence and intensity ratio of neural electrical events and intracranial pressure events in the pathological event chain model, the elasticity coefficient of intracranial volumetric pressure response is identified. This coefficient characterizes the remaining capacity of the current intracranial compensatory space.

[0070] To further clarify the specific construction process of the dynamic interactive network, this embodiment uses real-time monitoring data from a patient who has undergone intracranial tumor resection as an example to explain in detail the step-by-step operation method of network construction:

[0071] In data preprocessing, the multi-source stream receiving module acquired the patient's EEG waveform (sampling frequency 256Hz), intracranial pressure waveform (sampling frequency 100Hz), and surface electromyography (EMG) signal (sampling frequency 500Hz). First, data cleaning was performed to remove power frequency interference (50Hz) and spike noise caused by poor electrode contact in the EEG waveform. Low-pass filtering (cutoff frequency 30Hz) was used to retain effective rhythmic components. Baseline drift correction was performed on the intracranial pressure waveform to remove slow drift interference caused by respiration and body movement. Bandpass filtering (10-200Hz) was applied to the surface EMG signal to remove ECG artifacts and environmental noise. The three types of preprocessed signals were then timestamped to ensure data synchronization in the time dimension.

[0072] During feature extraction and quantization, wavelet decomposition was used to decompose the preprocessed EEG waveform into its components for rhythm decoupling. (0.5-4Hz) (4-8Hz) (8-13Hz) Calculate the four frequency bands (13-30Hz). The integral area of ​​the frequency band is used as the slow wave oscillation energy, and the phase synchronization index of the β band is calculated as the fast wave rhythm synchronization index. Epileptiform discharge events are identified using a threshold method, with a threshold of three times the standard deviation of the patient's preoperative baseline EEG. The number of discharges per unit time is counted as the epileptiform discharge density. Intracranial pressure waveform morphological decomposition uses a peak detection algorithm to identify the pulse wave component in the intracranial pressure waveform, and the time difference between adjacent pulse wave peaks is calculated as the pulse wave transmission delay. Frequency components corresponding to the respiratory rate (12-20 breaths / minute) are extracted through spectral analysis, and the amplitude proportion of this component is calculated as the respiratory wave coupling strength. A threshold of 1.5 times the baseline intracranial pressure value is used, and the number of peaks exceeding this threshold per unit time is counted as the frequency of abnormal high-amplitude waves. Feature extraction of surface electromyography (EMG) signals requires calculating the root mean square value of the preprocessed surface EMG signals as the background activity level of muscle tone. Abnormal myoclonic bursts are detected by short-time energy method (window length 200ms, step size 50ms), and the onset time, duration and involved EMG channels of each burst are recorded to form the spatiotemporal distribution characteristics of abnormal myoclonic bursts.

[0073] During the sub-model construction process, the adjustment loop model uses a 5-minute sliding time window to measure the slow-wave oscillation energy within each time window. ) and pulse wave transmission delay ( Perform pairwise analysis to establish an energy-delay correlation table. If within the same time window... and If an inverse change is observed, it is determined to be a valid moderating correlation, and a linear equation is fitted based on this correlation:

[0074]

[0075] in For adjustment coefficients, For the constant term, the goodness of fit of the equation If the correlation within that time window is valid, it is included in the adjustment loop model; if If the segment is abnormal, it is marked as a segment with abnormal regulatory function, and the original data is retained for subsequent analysis. The respiratory mechanics coupling model is based on the patient's real-time respiratory cycle (set as...). ) on respiratory wave coupling strength ( ) and background activity level of muscle tone ( The process is divided into segments, and calculations are performed within each respiratory cycle. and range of change .like and If the signs are consistent, it is determined to be a coordinated change, and the coordinated change rate is calculated. A coupling relationship model is established based on this rate of change:

[0076]

[0077] in, The coupling coefficient is... The constant term covers effective collaborative data for at least 10 consecutive respiratory cycles to complete the model construction. The pathological event chain model timestamps epileptiform discharge density, abnormal high-amplitude wave frequency, and abnormal myoclonic burst characteristics. If an increase in abnormal high-amplitude wave frequency occurs within 3 seconds after an increase in epileptiform discharge density, and is accompanied by an abnormal myoclonic burst characteristic event, a unidirectional conduction chain is formed from epileptiform discharge, abnormal high-amplitude intracranial pressure, to myoclonic burst. If the increase in abnormal high-amplitude wave frequency precedes the increase in epileptiform discharge density, a reverse conduction chain is formed from abnormal intracranial pressure to abnormal nerve excitability. All conduction chains are sorted by frequency of occurrence, and the top 3 core conduction chains with the highest frequency constitute the pathological event chain model.

[0078] The global network integration adopts a hierarchical integration strategy. The first layer is the feature layer, which integrates all the quantized features extracted above. The first layer is the feature matrix, which is classified and stored according to signal type. The second layer is the sub-model layer, which uses the output results of the regulation loop model, respiratory mechanics coupling model and pathological event chain model as intermediate variables to establish the mapping relationship between features and intermediate variables. The third layer is the global network layer, which constructs a visual network structure in the form of nodes and edges. The weight of the edge is determined by the correlation strength or conduction frequency. The higher the correlation strength and the more conduction frequency, the greater the weight. Finally, a dynamically updated interactive network of neurophysiology and intracranial mechanical state is formed.

[0079] In practical implementation, the following description uses a monitoring scenario of a patient undergoing decompressive craniectomy as an example. The multi-source stream receiving module continuously collects real-time physiological monitoring streams output from the patient's bedside monitoring equipment, including electroencephalogram (EEG) waveforms from an EEG machine, intracranial pressure waveforms from an intracranial pressure sensor, and surface electromyography (EMG) signals from an EMG machine. These signals are synchronously transmitted to the system at a specific sampling frequency.

[0080] In some embodiments, the dynamic interactive network module performs rhythm decoupling on the incoming EEG waveform, using wavelet transform and independent component analysis to separate rhythmic components of different frequency bands, and calculates three quantitative indicators: slow-wave oscillation energy, fast-wave rhythm synchronization index, and epileptiform discharge density. For example, within a five-minute time window, the calculated slow-wave oscillation energy is a specific value, and the epileptiform discharge density is a specific number of times per second. Simultaneously, morphological decomposition of the intracranial pressure waveform is performed, and peak detection and spectral analysis algorithms are applied to identify the basic components in the waveform, obtaining pulse wave propagation delay, respiratory wave coupling strength, and the frequency of abnormal high-amplitude waves. For example, the identified pulse wave propagation delay is a specific number of milliseconds, and the respiratory wave coupling strength is a specific value. Temporal and frequency domain features are extracted from the surface electromyography (EMG) signal to obtain the root mean square value of the background muscle tone activity level and the spatiotemporal distribution characteristics of abnormal myoclonic bursts. For example, the background muscle tone activity level is a specific microvolt, and an abnormal myoclonic burst lasts a specific number of milliseconds and involves a specific muscle group.

[0081] A correlation mapping was established between EEG slow-wave oscillation energy and intracranial pressure pulse wave transmission delay. Using a five-minute sliding time window, the coherence and phase difference between the envelope of slow-wave oscillation energy and the pulse wave transmission delay sequence were calculated, generating a regulatory loop model reflecting the autoregulation function of cerebral blood flow. Cross-analysis was performed on the coupling strength of intracranial pressure respiratory waves and the background activity level of muscle tone to calculate the degree of synchronization between muscle tone level and intracranial pressure respiratory wave amplitude during the respiratory cycle, generating a respiratory mechanics coupling model reflecting the influence of changes in thoracic and abdominal pressure on intracranial pressure. Epileptiform discharge density, abnormal high-amplitude wave frequency of intracranial pressure, and abnormal myoclonic burst characteristics were aligned from multiple sources and causal inferences were performed. A time-series event alignment algorithm was used to align the onset times of the three types of events within a specific millisecond-level time tolerance window. The guiding relationship between events was analyzed based on the Granger causality test, generating a pathological event chain model reflecting the transmission of abnormal neural excitability to intracranial mechanical state. By integrating regulatory loop models, respiratory mechanics coupling models, and pathological event chain models, a dynamic interactive network of neuroelectrophysiology and intracranial mechanical state is formed. This network consists of multiple sub-models representing physiological or pathological processes and the interaction relationships between the sub-models.

[0082] It is understandable that the state parameter parsing module parses core state parameters from the constructed dynamic interactive network. Based on the phase relationship between slow wave oscillation energy and pulse wave delay in the regulation loop model, the cerebral blood flow regulatory reserve index is calculated. The calculation process involves assessing the phase lag of the pulse wave delay change relative to the slow wave energy change within a complete slow wave oscillation cycle, and using a normalization function to map the phase lag to a cerebral blood flow regulatory reserve index within a specific range. This index is used to quantify the metabolic basis of cortical excitability. Based on the synergistic change pattern of respiratory wave coupling strength and muscle tone level in the respiratory mechanics coupling model, an indirect estimate of central venous pressure is calculated. The calculation is achieved through a trained multivariate regression model that takes respiratory wave coupling strength, specific muscle group tone level, and respiratory rate as inputs and outputs an estimate of central venous pressure. This indirect estimate is then subtracted from the baseline pressure level of the intracranial pressure waveform and divided by a simulated cerebrospinal fluid circulation pathway resistance parameter to calculate the pressure gradient load of cerebrospinal fluid circulation. The unit of the pressure gradient load is a specific physical unit. This study analyzes the sequence and intensity ratio of neural electrical events and intracranial pressure events in a pathological event chain model to identify the elasticity coefficient of intracranial volumetric pressure response. The identification process involves extracting aligned sequences of neural electrical event intensity and intracranial pressure event intensity, and calculating the ratio of the cumulative sum of intracranial pressure event intensity to the cumulative sum of neural electrical event intensity over a certain time period. This ratio, after being standardized with the patient's individual intracranial volume parameters, is defined as the elasticity coefficient of intracranial volumetric pressure response. The elasticity coefficient characterizes the remaining capacity of the current intracranial compensatory space.

[0083] Optionally, when calculating the cerebral blood flow regulatory reserve index, the quantification of the relationship between slow wave oscillation energy and pulse wave delay phase can be performed using the following synchronicity index formula:

[0084]

[0085] Where: symbol Represents the synchronicity index, symbol Represents the length of the time window used in the calculation, sign Representative moment The instantaneous phase of slow wave oscillations in brainwaves, symbol Representative moment The phase corresponding to the delay in the transmission of intracranial pressure pulse waves.

[0086] In one embodiment of the present invention, see [reference] Figure 3 The risk warning and analysis module inputs the cerebral blood flow regulatory reserve index, the pressure gradient load of cerebrospinal fluid circulation, and the remaining capacity of the intracranial compensatory space into the intracranial steady-state deviation warning unit. This unit inputs the cerebral blood flow regulatory reserve index into the perfusion risk assessment subunit, combining it with the patient's baseline cerebral blood flow velocity data to simulate the stability boundary of cerebral perfusion pressure under different blood pressure fluctuation scenarios, thereby calculating the probability of deviation from the stable boundary as the risk of brain tissue perfusion. The pressure gradient load of cerebrospinal fluid circulation is input into the circulation load assessment subunit, combining it with preoperative ventricular size data to calculate the current balance between cerebrospinal fluid absorption resistance and generation rate, and predict the trend of ventricular volume change. The remaining capacity of the intracranial compensatory space is input into the space compensation assessment subunit, combining it with the degree of brain tissue displacement shown in postoperative imaging to construct an intracranial volume-pressure relationship curve, and locate the current state's position on this curve. By combining the trend of ventricular volume change with the current state's position on the intracranial volume-pressure relationship curve, the probability of irreversible structural displacement or brain herniation of brain tissue is calculated as the risk of neural structure compression.

[0087] In practical implementation, the following description uses a monitoring scenario of a patient after intracranial tumor resection as an example. The state parameter parsing module continuously inputs three types of core state parameters parsed from the dynamic interactive network into the intracranial homeostasis deviation early warning unit: the cerebral blood flow regulatory reserve index, the pressure gradient load of cerebrospinal fluid circulation, and the elasticity coefficient of intracranial volumetric pressure response. The intracranial homeostasis deviation early warning unit includes a perfusion risk assessment subunit, a circulatory load assessment subunit, and a spatial compensation assessment subunit. Each subunit processes the input core state parameters in parallel and performs calculations in conjunction with the patient's specific historical data.

[0088] In some embodiments, the perfusion risk assessment subunit receives real-time input of the cerebral blood flow regulatory reserve index. The subunit retrieves baseline cerebral blood flow velocity data obtained via transcranial Doppler ultrasound measurement at the patient's admission, which is the average value of the blood flow velocity in the middle cerebral artery. The subunit pre-sets a set of blood pressure fluctuation simulation parameters, comprising a continuous blood pressure value sequence from a specific mmHg to a specific mmHg, with a simulation step size of a specific mmHg. The subunit performs simulations based on a physiological model of automatic cerebral blood flow regulation, in which the cerebral blood flow regulatory reserve index is embedded as a core parameter characterizing the vasomotor response. The simulation process involves taking each pre-set simulated blood pressure value as input, combining it with the real-time cerebral blood flow regulatory reserve index and baseline cerebral blood flow velocity data, and calculating the corresponding dynamic trajectory of cerebral perfusion pressure using the physiological model. From all simulated trajectories, the upper and lower limits of the simulated blood pressure values ​​that can maintain cerebral perfusion pressure within a safe range of specific mmHg throughout the entire process are identified. These two limits are defined as the stability boundaries of cerebral perfusion pressure at the current moment. The closest distance between the actual mean arterial pressure value read from the patient's arterial blood pressure monitor at the current moment and the stability boundary is calculated. This distance value is mapped to a probability value between specific values ​​using a preset monotonically increasing function. The mapped probability value is output as the cerebral tissue perfusion risk.

[0089] It is understandable that the circulatory load assessment subunit simultaneously receives real-time input pressure gradient load of cerebrospinal fluid circulation. The subunit retrieves ventricular size data measured from the patient's preoperative cranial computed tomography (CT) images, quantifying this data as initial ventricular volume in milliliters. The subunit incorporates a circulatory model based on the differential equations of cerebrospinal fluid dynamics. The real-time analyzed pressure gradient load values ​​of cerebrospinal fluid circulation are input into the circulatory model. The model uses cerebrospinal fluid absorption resistance and cerebrospinal fluid production rate as internal variables, adjusting these values ​​through an iterative algorithm to minimize the error between the theoretical pressure gradient value output by the model and the actual input pressure gradient load value. This allows the subunit to deduce the most suitable set of cerebrospinal fluid absorption resistance and cerebrospinal fluid production rate values ​​at the current moment. By comparing the derived cerebrospinal fluid (CSF) absorption resistance value with the CSF production rate value, if the difference between the CSF absorption resistance value and the CSF production rate value is greater than a certain threshold, the CSF dynamic equilibrium is determined to be biased towards CSF accumulation; if the difference between the CSF production rate value and the CSF absorption resistance value is greater than a certain threshold, the equilibrium is determined to be biased towards excessive CSF loss; if the absolute value of the difference is less than a certain threshold, it is determined to be basically balanced. Based on the determined equilibrium tendency and the specific values ​​of the CSF absorption resistance value and the CSF production rate value, the circulatory load assessment subunit uses an empirical formula describing the change of ventricular volume over time to calculate the expected change and direction of ventricular volume relative to the initial ventricular volume within a specific future hour. The expected change and direction constitute the predicted trend of ventricular volume change.

[0090] Optionally, the spatial compensation assessment subunit receives the elasticity coefficient of intracranial volumetric pressure response in real time. The subunit retrieves the patient's immediate postoperative cranial computed tomography (CT) image, quantifying the degree of brain tissue displacement as a midline shift in millimeters. Based on the elasticity coefficient of intracranial volumetric pressure response and the degree of brain tissue displacement, the subunit constructs an intracranial volumetric pressure relationship curve characterizing the relationship between intracranial volume and intracranial pressure. The construction process involves starting with the initial intracranial pressure value monitored at the moment, defining the slope of the curve using the elasticity coefficient of intracranial volumetric pressure response, and correcting for non-linear segments of the curve by incorporating the degree of brain tissue displacement. Subsequently, the average value of the actual intracranial pressure continuously monitored by the multi-source receiving module at the current moment, along with the equivalent volume change estimated based on the intracranial pressure waveform and cerebrospinal fluid circulation status, is used as a set of coordinates to locate the current state point on the constructed intracranial volumetric pressure relationship curve, thus determining the specific position of the current state point on the curve.

[0091] In its implementation, the intracranial steady-state deviation early warning unit is equipped with a comprehensive logic processor to aggregate and process the outputs from the three sub-units. The comprehensive logic processor receives the predicted ventricular volume change trend from the cyclic load assessment sub-unit and the position of the current state point on the intracranial volume-pressure relationship curve from the spatial compensation assessment sub-unit. The calculation of the risk of neural structure compression integrates the information of the ventricular volume change trend and the position of the current state point on the intracranial volume-pressure relationship curve. The calculation process assesses whether the ventricular volume change trend indicates a rapid increase in volume and whether the position of the current state point on the intracranial volume-pressure relationship curve has entered a steep segment representing compensation exhaustion. The comprehensive assessment is completed through a weighted decision function, which outputs a probability value as the risk of neural structure compression. The specific implementation of the weighted decision function is as follows: the function receives the quantified result of the predicted ventricular volume change trend from the cyclic load assessment sub-unit and the quantified parameter of the position of the current state point on the intracranial volume-pressure relationship curve from the spatial compensation assessment sub-unit. The function uses preset weighting coefficients to weight and combine the risk components corresponding to the trend of ventricular volume change with the risk components corresponding to the current state point location, where the sum of the weighting coefficients is a fixed value. The combined result is normalized and outputs a probability value between 0 and 1. This probability value represents the risk of neural structure compression, used to quantify the likelihood of irreversible structural displacement or brain herniation. The probability value of neural structure compression risk is calculated using the following formula:

[0092]

[0093] Where: symbol The symbol represents the probability value of the risk of compression of neural structures. and These are pre-defined weighting coefficients, with the following symbols: A quantized vector representing the predicted trend of ventricular volume change, with the symbol... The quantization parameter representing the position of the current state point on the intracranial volume-pressure relationship curve, function It is a transformation function that maps the trend of ventricular volume change to a risk component. It is a transformation function that maps the curve position to risk components. Weighting coefficients. and The sum of these values ​​is a specific numerical value. The two independent probability values—brain tissue perfusion risk and neural structure compression risk—are output in parallel by the intracranial homeostasis deviation early warning unit to the early warning path generation module.

[0094] In one embodiment of the invention, the perfusion risk assessment subunit pre-sets a set of blood pressure fluctuation simulation parameters, which cover a continuous range from hypotension to hypertension. Baseline cerebral blood flow velocity data measured upon patient admission or preoperatively is used as baseline data and coupled with the real-time acquired cerebral blood flow regulatory reserve index. Based on a physiological model of automatic cerebral blood flow regulation, the cerebral blood flow regulatory reserve index is used as the core regulatory capability parameter of the model to simulate the corresponding dynamic change trajectory of cerebral perfusion pressure when system blood pressure varies within the preset fluctuation parameter range. The upper and lower limits of blood pressure that can maintain cerebral perfusion pressure within a safe range are identified from the simulation results and defined as the stability boundary of cerebral perfusion pressure. The closest distance between the patient's actual blood pressure value at the current moment and the stability boundary is calculated, and this distance value is mapped to a probability value through a preset risk transformation function. This probability value is the risk probability of deviating from the stability boundary and is used as the final output of cerebral tissue perfusion risk. The circulatory load assessment subunit retrieves the ventricular size data measured by the patient's preoperative imaging examination and quantifies it as the initial ventricular volume. The pressure gradient load value of cerebrospinal fluid circulation obtained from real-time analysis is substituted into a circulation model based on the differential equation of cerebrospinal fluid dynamics. This circulation model uses cerebrospinal fluid absorption resistance and generation rate as key variables. By iteratively fitting the current pressure gradient load value with the theoretical pressure value calculated by the model, the most suitable cerebrospinal fluid absorption resistance value and generation rate value at the current moment are derived. The relationship between the derived absorption resistance value and generation rate value is compared. If the absorption resistance is significantly greater than the generation rate, the equilibrium state is determined to be biased towards cerebrospinal fluid accumulation; if the generation rate is significantly greater than the absorption resistance, the equilibrium state is determined to be biased towards excessive cerebrospinal fluid loss; if the two are similar, it is determined to be basically balanced. Based on the determined equilibrium state tendency and the specific numerical difference between absorption resistance and generation rate, the expected change in ventricular volume over a preset time period is calculated using an empirical formula for ventricular volume change over time. The expected change in volume and direction together constitute the predicted trend of ventricular volume change.

[0095] In practical implementation, the following explanation uses monitoring data from a patient who underwent intracranial aneurysm clipping surgery. When the perfusion risk assessment subunit initiates the assessment, it loads a set of pre-defined blood pressure fluctuation simulation parameters from the system's internal parameter library. This set of blood pressure fluctuation simulation parameters is an arithmetic sequence containing a specific number of elements, with values ​​ranging from a specific mmHg low blood pressure value to a specific mmHg high blood pressure value, with the interval between adjacent values ​​being a specific mmHg. Simultaneously, the subunit retrieves the baseline cerebral blood flow velocity data recorded by transcranial Doppler ultrasound at the time of the patient's admission from the patient's electronic medical record database. This baseline cerebral blood flow velocity data is the average blood flow velocity value of the middle cerebral artery, serving as the baseline data for the simulation calculation. In practical implementation, the subunit couples the cerebral blood flow regulation reserve index, which is obtained in real-time from the state parameter parsing module, with the retrieved baseline cerebral blood flow velocity data. The coupling process involves introducing the cerebral blood flow regulation reserve index as a gain coefficient into the physiological model of automatic cerebral blood flow regulation. This physiological model describes the dynamic relationship between cerebral perfusion pressure, cerebrovascular resistance, and cerebral blood flow. The cerebral blood flow regulatory reserve index directly corrects the response function of cerebrovascular resistance to changes in blood pressure in the model. After the simulation begins, the perfusion risk assessment subunit sequentially inputs each blood pressure value from the simulated blood pressure fluctuation parameter sequence into the physiological model coupled with the real-time cerebral blood flow regulatory reserve index, calculating the dynamic trajectory of cerebral perfusion pressure over time under each simulated blood pressure input. From all the dynamic trajectories obtained from the simulation, the subunit identifies the upper and lower limits of the input blood pressure values ​​that can ensure that the cerebral perfusion pressure is always maintained within a safe range of a specific mmHg to a specific mmHg. These two limits are defined as the stability boundaries of the cerebral perfusion pressure at the current moment.

[0096] In some embodiments, the perfusion risk assessment subunit acquires the patient's actual arterial blood pressure value in real time through a bedside monitoring system. This actual blood pressure value is typically the mean arterial pressure (MAP). The subunit calculates the closest distance between the current actual MAP value and the previously calculated stability boundary. The distance calculation takes the minimum of the distances from the actual blood pressure value to the upper and lower limits of the stability boundary. This minimum distance value is input into a preset risk transformation function, which is a monotonically decreasing function defined in the real number domain. This function maps the distance value to a probability value within a specific range. The mapped probability value is quantified as the risk probability of deviating from the stability boundary and is output as the final brain tissue perfusion risk value. It can be understood that the execution process of the circulatory load assessment subunit is independent and parallel. When the circulatory load assessment subunit starts, it first retrieves the patient's preoperative cranial MRI images from the medical image archive system and measures the total volume of the ventricular system using a built-in image analysis algorithm, quantifying it as the initial ventricular volume in milliliters. The subunit receives a real-time data stream from the state parameter parsing module, namely the pressure gradient load value of the cerebrospinal fluid circulation. The cyclic model of the cerebrospinal fluid dynamics differential equation built into the cyclic load assessment subunit is activated. The cyclic model uses the cerebrospinal fluid absorption resistance value and the cerebrospinal fluid production rate value as two key variables that need to be solved. The subunit uses the real-time acquired pressure gradient load value as the target value and continuously adjusts the assumed values ​​of the cerebrospinal fluid absorption resistance value and the cerebrospinal fluid production rate value through an iterative fitting algorithm until the theoretical pressure gradient value calculated after substituting the assumed values ​​into the cyclic model infinitely approximates the real-time input pressure gradient load value. In this way, the set of cerebrospinal fluid absorption resistance value and cerebrospinal fluid production rate value that best matches the patient's physiological state at the current moment is deduced.

[0097] Optionally, after obtaining the inversely derived cerebrospinal fluid (CSF) absorption resistance and CSF production rate values, the circulatory load assessment subunit compares their magnitudes. The comparison operation involves calculating the difference between the CSF absorption resistance and CSF production rate values ​​and comparing it to a pre-set positive threshold. If the difference is greater than the positive threshold, the current CSF dynamic equilibrium is considered to favor CSF accumulation; if the absolute value of the difference is less than or equal to the positive threshold, but the difference between the CSF production rate and the CSF absorption resistance is greater than the same positive threshold, the equilibrium is considered to favor excessive CSF loss; if only the absolute value of the difference is less than or equal to the positive threshold, the CSF absorption resistance and CSF production rate are considered to be in a basic equilibrium. Based on this determination and the specific values ​​of the CSF absorption resistance and CSF production rate values, the circulatory load assessment subunit uses an empirical formula for ventricular volume change over time to predict trends. This empirical formula expresses the rate of change of ventricular volume over time under a specific equilibrium state. The formula for predicting the trend of ventricular volume change is:

[0098]

[0099] Where: symbol Represents a specific point in the future Predicted change in ventricular volume, symbol Represents the preset length of the future observation period, symbol This represents the cerebrospinal fluid absorption resistance value derived from reverse calculation, with the symbol... The function represents the cerebrospinal fluid production rate value derived from reverse calculation. It is an empirical function that calculates the rate of volume change based on the cerebrospinal fluid absorption resistance and cerebrospinal fluid production rate. The integration operation starts from the current moment and extends over future time periods. The rate of change in ventricular volume is accumulated to obtain the predicted change in ventricular volume. The sign of the change indicates the direction of the change. The predicted change in ventricular volume and the direction of the change together constitute the predicted trend of ventricular volume change.

[0100] In one embodiment of the present invention, the early warning path generation module compares the risk of cerebral tissue perfusion with a preset perfusion risk threshold. When the threshold is exceeded, a cerebral perfusion maintenance early warning is triggered, and a regulation path aimed at increasing cerebral perfusion pressure is generated. This path includes a recommended target mean arterial pressure range and blood pressure adjustment rate. The risk of neural structure compression is compared with a preset structural compression risk threshold. When the threshold is exceeded, an intracranial pressure control early warning is triggered, and a regulation path aimed at reducing intracranial volume is generated. This path includes suggested head position angles, hyperventilation parameters, and the timing of osmotic dehydration agent activation. When both the risk of cerebral tissue perfusion and the risk of neural structure compression exceed their respective thresholds, an intracranial homeostasis imbalance early warning is triggered, and a composite regulation path that balances the conflict between cerebral perfusion and intracranial volume is generated. The composite regulation path specifies the priority and operating window for implementing measures to reduce intracranial pressure in stages while maintaining the minimum permissible cerebral perfusion pressure. When generating a regulation path aimed at increasing cerebral perfusion pressure, the cerebral blood flow regulatory reserve index is divided into three levels: sufficient reserve, reduced reserve, and depleted reserve, based on the specific value of the index. For levels of sufficient reserve, a plan for gradually increasing mean arterial pressure (MAP) is developed, with simultaneous monitoring of the proportional increase in cerebral blood flow velocity during the adjustment process. For levels of reduced reserve, a plan for gradually increasing MAP in a step-wise manner is developed, with monitoring of the improvement in slow-wave oscillation energy on electroencephalography (EEG) after each step to determine whether to proceed to the next step. For levels of reserve depletion, an emergency volume expansion and vasopressor plan is generated, requiring simultaneous monitoring of central venous pressure and surface electromyography (EMG) signals to prevent heart failure or worsening of myocardial edema caused by rapid volume expansion.

[0101] In practical implementation, the following explanation uses a set of brain tissue perfusion risk and neural structure compression risk values ​​output in real time by the risk warning analysis module. The warning path generation module continuously receives two independent probability values ​​from the risk warning analysis module: brain tissue perfusion risk and neural structure compression risk. The module has preset perfusion risk thresholds and structural compression risk thresholds, which are either fixed values ​​based on clinical consensus or dynamic values ​​adjusted according to the patient's preoperative condition. The warning path generation module compares the real-time brain tissue perfusion risk values ​​with the preset perfusion risk thresholds. When the brain tissue perfusion risk value continuously exceeds the perfusion risk threshold for a specific duration, the module triggers a brain perfusion maintenance warning. The warning signal includes visual alerts, audio prompts, and specific risk description text. Simultaneously with triggering the brain perfusion maintenance warning, the warning path generation module calls its internal algorithm to generate a regulation path aimed at increasing brain perfusion pressure. The generated regulation path is a set of structured nursing intervention suggestions, explicitly including the recommended target mean arterial pressure range and blood pressure adjustment rate. For example, it suggests maintaining the mean arterial pressure between a specific mmHg and a specific range of mmHg, and adjusting it at a rate not exceeding a specific mmHg per hour. The warning path generation module compares the real-time received neural structure compression risk value with the preset structural compression risk threshold. When the neural structure compression risk value continues to exceed the structural compression risk threshold for a certain period of time, the module triggers an intracranial pressure control warning and simultaneously generates an adjustment path aimed at reducing intracranial volume. The generated adjustment path text clearly includes the recommended head position angle, hyperventilation parameters, and the timing of osmotic dehydration agent activation. For example, it recommends raising the head of the bed to a specific degree, controlling the end-tidal carbon dioxide partial pressure within a specific mmHg range, and assessing whether to use a specific dose of mannitol at a specific time point.

[0102] In some embodiments, the warning path generation module simultaneously determines that both the brain tissue perfusion risk and the neural structure compression risk values ​​exceed their respective thresholds. At this point, the module triggers a higher-priority intracranial homeostasis imbalance warning. After triggering the intracranial homeostasis imbalance warning, the warning path generation module does not simply merge the aforementioned two paths, but instead generates a novel composite regulatory path that balances the conflict between brain perfusion and cranial volume. The algorithm for generating the composite regulatory path first reads the minimum permissible brain perfusion pressure value set for the patient from the system parameter library. The core logic of the composite regulatory path is to ensure that the brain perfusion pressure does not fall below this minimum permissible brain perfusion pressure value throughout the intervention process. Based on this, the algorithm prioritizes and plans a series of intracranial pressure reduction measures in stages. The generated specification text clarifies the priority and operational window for implementing intracranial pressure reduction measures in stages. For example, the text specifies that the first stage attempts to reduce intracranial pressure by adjusting head position and mild hyperventilation within a specific time period, while closely monitoring brain perfusion pressure. If the first stage is ineffective and the brain perfusion pressure is stable, the second stage proceeds, using an osmotic dehydrating agent within a specific time window.

[0103] It is understandable that when generating a regulatory pathway aimed at increasing cerebral perfusion pressure, the early warning pathway generation module needs to perform refined grading based on the specific cerebral blood flow regulatory reserve index value. The internal grading logic of the early warning pathway generation module compares the cerebral blood flow regulatory reserve index value with two preset threshold values, thereby classifying it into three levels: sufficient reserve, reduced reserve, and depleted reserve. For the sufficient reserve level, the regulatory pathway text generated by the module sets a plan to gradually increase the mean arterial pressure. The plan clearly specifies the range of gentle increase in the target mean arterial pressure and stipulates that cerebral blood flow velocity measured by transcranial Doppler ultrasound must be monitored simultaneously during the adjustment process. The pathway text requires an assessment of whether the increase in cerebral blood flow velocity is proportional to the increase in mean arterial pressure. For the level of reduced reserve, the module-generated adjustment path text sets a step-by-step, small-amplitude increase in mean arterial pressure. The plan defines several blood pressure increase steps, each with a specific increase in millimeters of mercury. It stipulates that after each successful increase, a specific observation period is required, and the improvement in EEG slow-wave oscillation energy is analyzed to determine whether to proceed to the next step. The improvement is judged based on the percentage increase in EEG slow-wave oscillation energy relative to the pre-adjustment level. For the level of reserve depletion, the module-generated adjustment path text generates an emergency volume expansion and vasopressor plan. This plan directly recommends intravenous infusion of specific types of crystalloid or colloid solutions to achieve rapid volume loading. It explicitly states that this plan must be implemented in conjunction with monitoring central venous pressure and electromyography (EMG) of specific muscle groups. The path text indicates that the purpose of this combined monitoring is to prevent heart failure or worsening of muscle edema caused by rapid volume expansion. See Table 1 for key adjustment parameters corresponding to different reserve levels.

[0104] Table 1: Parameters of the Regulatory Pathway Corresponding to Different Levels of Cerebral Blood Flow Regulatory Reserve

[0105]

[0106] In practical implementation, the calculation of the specific magnitude of blood pressure adjustment within the regulatory pathway involves a simple linear function. For example, at the adequate reserve level, the target mean arterial pressure adjustment value can be calculated based on the baseline mean arterial pressure and the cerebral blood flow regulatory reserve index, using the following formula:

[0107]

[0108] Where: symbol Represents the calculated target mean arterial pressure value, symbol Represents the baseline mean arterial pressure of patients in a stable state before the warning is triggered, symbol The cerebral blood flow regulatory reserve index represents the current moment, with the symbol... It is a preset proportional coefficient used to control the adjustment range. The warning path generation module will calculate the... The value is automatically filled into the recommended target mean arterial pressure range in the regulation pathway text. Optionally, for levels of reduced and depleted reserve, the generation logic of the regulation pathway is more complex, relying not only on the cerebral blood flow regulation reserve index, but also on current central venous pressure, heart rate, and other vital signs parameters. A multi-parameter lookup table or rule engine is used to ultimately determine the recommended fluid type, infusion rate, and vasoactive drug usage recommendations.

[0109] See Figure 4This figure illustrates the stability boundary characteristics of cerebral perfusion pressure as a function of systemic blood pressure when the cerebral blood flow regulatory reserve is at different levels. The figure uses systemic blood pressure (mmHg) as the horizontal axis and cerebral perfusion pressure (mmHg) as the vertical axis. Three curves represent the dynamic trajectory of cerebral perfusion pressure under three states: "sufficient reserve," "decreased reserve," and "reserve depletion." The lowest safe perfusion pressure (60 mmHg) and the highest safe perfusion pressure (100 mmHg) are marked with dashed lines, and the light green area between them represents the safe perfusion zone. Specifically, in the "ample reserve" (green curve) state, cerebral perfusion pressure shows a segmented increase with rising systemic blood pressure. The slope changes as systemic blood pressure crosses thresholds such as 60 mmHg and 90 mmHg, and the overall trajectory remains within the safe perfusion range, demonstrating the stabilizing effect of cerebral blood flow autoregulation on perfusion pressure. In the "decreased reserve" (yellow curve) state, the cerebral perfusion pressure trajectory has a relatively gentle slope, with its upper limit of the safe boundary closer to 100 mmHg, indicating partial impairment of regulatory capacity. In the "reserve depletion" (red curve) state, the cerebral perfusion pressure trajectory is significantly lower than the lower limit of the safe range (60 mmHg), and the fluctuation amplitude with systemic blood pressure changes is smaller, reflecting the pathological state where, after the loss of cerebral blood flow regulation function, cerebral perfusion pressure cannot be maintained within a safe range through self-regulation. The core value of the figure lies in quantifying the influence boundary of systemic blood pressure fluctuations on cerebral perfusion pressure stability under different levels of cerebral blood flow regulatory reserve, providing a visualized physiological model for clinically developing blood pressure regulation pathways based on the cerebral blood flow regulatory reserve index.

[0110] In one embodiment of the invention, when generating a regulatory pathway aimed at reducing intracranial volume, the main sources of pressure gradient load in cerebrospinal fluid circulation are analyzed to distinguish whether the primary source is cerebrospinal fluid absorption impairment or venous return obstruction. For cases where cerebrospinal fluid absorption impairment is the primary source, a pathway centered on regulating the rate of cerebrospinal fluid production is generated. This specifically includes recommendations for the types of drugs used to reduce cerebrospinal fluid production, initial doses, and methods for adjusting the dose based on changes in respiratory wave coupling intensity in the intracranial pressure waveform. For cases where venous return obstruction is the primary source, a pathway centered on optimizing intracranial venous return is generated. This specifically includes specific operational steps and parameter combinations for reducing central venous pressure by adjusting headboard height, neck position, and controlling intrathoracic pressure. The specific operational steps and parameter combinations for reducing central venous pressure by adjusting headboard height, neck position, and controlling intrathoracic pressure include raising the headboard height from its initial position to a first target angle and maintaining this angle for a first preset observation duration, monitoring the decrease in respiratory wave coupling intensity in the intracranial pressure waveform. If the reduction in pressure does not meet expectations, the patient's neck position is further adjusted to a midline position with slight extension, while monitoring the tension of the neck and shoulder muscles in the electromyography (EMG) signal to ensure it does not increase significantly. After positional adjustment, the patient is instructed to adopt a slow ventilation mode primarily using diaphragmatic breathing, and the decrease in intrathoracic pressure is monitored in real time using respiratory monitoring equipment. Based on the feedback data of intracranial pressure-respiratory wave coupling intensity, neck and shoulder EMG tension, and intrathoracic pressure, the head of the bed angle, neck extension angle, and breathing instructions are dynamically fine-tuned to form an optimal combination of position and respiratory parameters that reduces the estimated central venous pressure to the target range.

[0111] In practical implementation, the following scenario illustrates the case of a patient who experienced an increased intracranial pressure alarm after craniocerebral decompression surgery. Upon triggering the intracranial pressure control alarm, the warning path generation module initiates a specialized program to generate a regulatory path aimed at reducing intracranial volume. The program first analyzes the real-time pressure gradient load values ​​and their compositional characteristics of the cerebrospinal fluid circulation input from the state parameter analysis module. The analysis process calls upon historical data and models to perform source analysis of the pressure gradient load. The source analysis algorithm distinguishes whether the main contribution of the pressure gradient load originates from impaired cerebrospinal fluid absorption or obstructed intracranial venous return.

[0112] In some embodiments, when the source resolution algorithm determines that the pressure gradient load of cerebrospinal fluid circulation mainly originates from cerebrospinal fluid absorption impairment, the warning path generation module generates a regulation path centered on adjusting the rate of cerebrospinal fluid production. This path is output in text form, explicitly including the recommended drug type to reduce cerebrospinal fluid production, the initial dose, and the method for adjusting the dose based on changes in the respiratory wave coupling intensity in the intracranial pressure waveform. For example, the regulation path text recommends using acetazolamide, with an initial dose of a specific milligram per dose, administered a specific number of times daily; and specifies that after administration, the respiratory wave coupling intensity in the intracranial pressure waveform should be continuously monitored. If the respiratory wave coupling intensity does not decrease by a specific percentage within a specific hour, it is recommended to increase the acetazolamide dose to the specific milligram per dose. When the source resolution algorithm determines that the pressure gradient load of cerebrospinal fluid circulation mainly comes from venous return obstruction, the early warning path generation module generates a regulation path with the core of optimizing intracranial venous return. This path is also output in text form. The text specifically includes the operation steps and parameter combinations for reducing central venous pressure by adjusting the head of the bed, neck position, and controlling intrathoracic pressure. For example, the path text details "gradually raise the head of the bed from the supine position to 30 degrees, keeping the patient's neck in the neutral position without rotation or flexion."

[0113] Understandably, in cases primarily involving venous return obstruction, the specific operational steps and parameter combinations for adjusting bed head height, neck position, and controlling intrathoracic pressure within the generated adjustment path are executed sequentially and form a closed-loop feedback. The first step involves raising the bed head height from its initial position to a first target angle, a preset initial intervention angle, such as 15 degrees. At this angle, the system specifies maintaining a first preset observation duration, such as five minutes, and continuously monitoring the respiratory wave coupling intensity in the intracranial pressure waveform within this duration, calculating its reduction relative to before adjustment. The second step involves determining whether the reduction has reached the expected threshold. If the reduction in respiratory wave coupling intensity has not reached the expected level, further positional adjustments are performed. These further adjustments include adjusting the patient's neck position to a neutral position with slight extension. Simultaneously, the system initiates monitoring of the electromyographic activity levels of the neck and shoulder muscles in the surface electromyographic signals to ensure that the neck position adjustment does not lead to a significant increase in the electromyographic tension of these muscle groups. The third step in the operation is to guide the patient to adopt a slow ventilation mode that primarily uses abdominal breathing, after adjusting the head of the bed and neck position. This step is done through voice prompts or a breathing guidance screen. The system obtains feedback on the decrease in intrathoracic pressure in real time through the connected respiratory monitoring equipment.

[0114] In practice, the system dynamically fine-tunes the head of the bed, neck extension angle, and rhythm of breathing instructions based on real-time feedback of the decrease in intracranial pressure respiratory wave coupling intensity, changes in electromyographic tension of the neck and shoulder muscles, and the decrease in intrathoracic pressure. The fine-tuning logic is based on an optimization algorithm that aims to find the optimal combination of parameters that lowers the estimated central venous pressure to a target range. The search for the optimal parameter combination can be described by an objective function that minimizes the combined cost function comprised of respiratory wave coupling intensity, electromyographic tension, and intrathoracic pressure. The estimated central venous pressure is... The calculation is derived from this feedback data using a multi-parameter model, and the formula is as follows:

[0115]

[0116] Where: symbol Represents the estimated central venous pressure, symbol The symbol represents the measured intracranial pressure respiratory wave coupling strength. This represents the measured electromyographic tension values ​​of the neck and shoulder muscle groups, with the symbol... Represents the monitored intrathoracic pressure value, symbol , and These are pre-calibrated weighting coefficients based on physiological relationships. The optimization algorithm iteratively fine-tunes the head-of-bed angle, neck extension angle, and breathing pattern parameters, observing and calculating the resulting effects. , and The change ultimately leads to the calculated... If the value falls within the preset target range, the corresponding set of headboard angle, neck extension angle, and breathing guidance parameters are determined as the optimal combination of body position and breathing parameters and locked in the final recommendation of the adjustment path.

[0117] Optionally, for patients unable to actively cooperate with breathing instructions, the portion of the adjustment pathway that controls intrathoracic pressure will be replaced with a parameter adjustment scheme executed by the ventilator. The path text generated by the warning path generation module will suggest adjusting the ventilator mode to a specific mode and setting specific parameters for target tidal volume, respiratory rate, and positive end-expiratory pressure to reduce mean intrathoracic pressure. At this point, the feedback variable intrathoracic pressure in the optimization loop is... The mean airway pressure, which will be directly monitored or estimated by the ventilator, will be replaced, and the system's dynamic fine-tuning will shift to the relevant parameters of the ventilator. The generated final adjustment path text will include clear descriptions of the optimal head-of-bed height angle, neck position, ventilator parameter settings, or key points for spontaneous breathing guidance, for healthcare personnel to follow.

[0118] See Figure 5In the correlation analysis between intracranial pressure respiratory wave coupling intensity and intrathoracic pressure (pressure source analysis stage), the figure uses intracranial pressure respiratory wave coupling intensity as the horizontal axis and intrathoracic pressure (unit: mmHg) as the vertical axis. Different colored scatter points distinguish four body position conditions: 0°, 15°, 30°, and 45° of bed head, and the overall trend is presented by a red dashed line. In the specific analysis, the scatter points corresponding to each bed head angle show obvious stratified distribution characteristics: the scatter points at 45° of bed head are concentrated in the range of low coupling intensity and intrathoracic pressure of 5-6 mmHg; the scatter points at 30° of bed head are distributed in the range of coupling intensity of 0.5-1.25 and intrathoracic pressure of 6-7 mmHg; the scatter points at 15° of bed head correspond to the range of coupling intensity of 1.0-1.75 and intrathoracic pressure of 6.5-8 mmHg; and the scatter points at 0° of bed head are distributed in the range of coupling intensity of 1.5-2.0 and intrathoracic pressure of 7-9 mmHg. The overall trend line shows a significant positive correlation, reflecting that intrathoracic pressure increases with the increase of intracranial pressure-respiratory wave coupling strength. Furthermore, the scatter plots at different head angles correspond to the moderating effect of body position on the relationship between the two. When the head angle increases, the intrathoracic pressure level at the same coupling strength decreases. This feature can provide data support for positional intervention for pressure gradient loading with obstructed venous return.

[0119] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0120] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A postoperative nursing risk early warning system for neurosurgical patients, characterized in that, The following steps are included: The multi-source stream receiving module continuously receives real-time physiological monitoring streams of the patient from multiple monitoring devices, including electroencephalogram (EEG) waveforms, intracranial pressure waveforms, and surface electromyography (EMG) signals. A dynamic interactive network module is constructed to establish a dynamic interactive network between neurophysiology and intracranial biomechanical state. The dynamic interactive network is used to characterize the coupling relationship between brain electrical rhythm fluctuations, intracranial pressure transmission and tissue compliance in the cranial fenestration area. The state parameter parsing module parses three types of core state parameters from the dynamic interactive network, reflecting the excitability of the cerebral cortex, the cerebrospinal fluid circulation load, and the cranial cavity compensation space. The risk warning and analysis module inputs the three types of core state parameters into the intracranial steady-state deviation warning unit to calculate the brain tissue perfusion risk and neural structure compression risk at the current moment. The early warning path generation module generates nursing early warning signals and corresponding physiological regulation paths based on the brain tissue perfusion risk and neural structure compression risk. The construction of the dynamic interactive network between neurophysiology and intracranial biomechanical state includes: The EEG waveform was rhythmically decoupled to separate the slow wave oscillation energy, fast wave rhythm synchronization index, and epileptiform discharge density. Simultaneously, morphological decomposition of the intracranial pressure waveform was performed to identify pulse wave transmission delay, respiratory wave coupling strength, and frequency of abnormal high-amplitude waves. Extract the spatiotemporal distribution characteristics of background muscle tone activity and abnormal myoclonic bursts from the surface electromyography signals; Establish a correlation mapping between the slow wave oscillation energy of EEG and the transmission delay of intracranial pressure pulse waves, and generate a regulatory loop model that reflects the automatic regulation function of cerebral blood flow. Cross-analysis of intracranial pressure respiratory wave coupling strength and muscle tone background activity level generates a respiratory mechanics coupling model that reflects the influence of changes in thoracic and abdominal pressure on intracranial pressure. By aligning epileptiform discharge density, abnormal high-amplitude wave frequency of intracranial pressure, and abnormal myoclonic burst characteristics with multi-source events and making causal inferences, a pathological event chain model reflecting the transmission of abnormal nerve excitability to intracranial mechanical state is generated. By integrating the regulatory loop model, respiratory mechanics coupling model, and pathological event chain model, a dynamic interactive network of neurophysiology and intracranial mechanical state is formed.

2. The postoperative nursing risk early warning system for neurosurgical patients according to claim 1, characterized in that, The three core state parameters, reflecting cortical excitability, cerebrospinal fluid circulation load, and cranial cavity compensatory space, are extracted from the dynamic interactive network, including: Based on the phase relationship between slow wave oscillation energy and pulse wave delay in the aforementioned regulatory loop model, the cerebral blood flow regulatory reserve index is calculated. The cerebral blood flow regulatory reserve index is used to quantify the metabolic basis of cerebral cortex excitability. Based on the coordinated change pattern of respiratory wave coupling intensity and muscle tone level in the respiratory mechanics coupling model, the indirect estimate of central venous pressure in the thoracic cavity is calculated, and combined with the baseline pressure level of intracranial pressure waveform, the pressure gradient load of cerebrospinal fluid circulation is calculated. By analyzing the sequence and intensity ratio of neural electrical events and intracranial pressure events in the pathological event chain model, the elasticity coefficient of intracranial volumetric pressure response is identified. The elasticity coefficient is used to characterize the remaining capacity of the current cranial cavity compensatory space.

3. The postoperative nursing risk early warning system for neurosurgical patients according to claim 2, characterized in that, The three types of core state parameters are input into the intracranial steady-state deviation early warning unit to calculate the current brain tissue perfusion risk and neural structure compression risk, including: The cerebral blood flow regulation reserve index is input into the perfusion risk assessment subunit. Combined with the patient's baseline cerebral blood flow velocity data, the stability boundary of cerebral perfusion pressure under different blood pressure fluctuation scenarios is simulated, thereby calculating the risk probability of deviating from the stability boundary, which is used as the cerebral tissue perfusion risk. The pressure gradient load of the cerebrospinal fluid circulation is input into the circulation load assessment subunit. Combined with the preoperative ventricular size data, the balance between the current cerebrospinal fluid absorption resistance and the generation rate is calculated, and the trend of ventricular volume change is predicted. The remaining capacity of the cranial cavity compensation space is input into the space compensation assessment subunit. Combined with the degree of brain tissue displacement shown in postoperative imaging, a cranial volume pressure relationship curve is constructed, and the current state is located on the cranial volume pressure relationship curve. By combining the trend of changes in ventricular volume with the position of the current state on the intracranial volume-pressure relationship curve, the probability of irreversible structural displacement or brain herniation of brain tissue is calculated as the risk of compression of the neural structure.

4. The postoperative care risk warning system for neurosurgical patients as described in claim 3, characterized in that, The step involves inputting the cerebral blood flow regulatory reserve index into the perfusion risk assessment subunit, combining it with the patient's baseline cerebral blood flow velocity data, simulating the stability boundary of cerebral perfusion pressure under different blood pressure fluctuation scenarios, and thereby calculating the risk probability of deviating from the stability boundary as the brain tissue perfusion risk, including: In the perfusion risk assessment subunit, a set of blood pressure fluctuation simulation parameters are pre-set, which cover a continuous range of variation from low blood pressure to high blood pressure. The baseline cerebral blood flow velocity data measured at the time of admission or before surgery is used as the baseline data and coupled with the real-time acquired cerebral blood flow regulatory reserve index. Based on the physiological model of automatic regulation of cerebral blood flow, the cerebral blood flow regulation reserve index is used as the core regulation capability parameter of the model to simulate the corresponding dynamic change trajectory of cerebral perfusion pressure when the system blood pressure changes within a preset fluctuation parameter range. The upper and lower limits of blood pressure that can maintain cerebral perfusion pressure within a safe range are identified from the simulation results, and these upper and lower limits of blood pressure are defined as the stability boundaries of cerebral perfusion pressure. Calculate the closest distance between the patient's actual blood pressure value at the current moment and the stability boundary, and map the distance value to a probability value through a preset risk transformation function. The probability value is the risk probability of deviating from the stability boundary, and is used as the final brain tissue perfusion risk output.

5. The postoperative nursing risk early warning system for neurosurgical patients as described in claim 4, characterized in that, The step of inputting the pressure gradient load of the cerebrospinal fluid circulation into the circulation load assessment subunit, combining it with preoperative ventricular size data, calculating the current balance between cerebrospinal fluid absorption resistance and generation rate, and predicting the trend of ventricular volume changes includes: In the cyclic load assessment subunit, the ventricular size data measured by the patient's preoperative imaging examination are retrieved and quantified as the initial ventricular volume; The pressure gradient load value of cerebrospinal fluid circulation obtained from real-time analysis is substituted into the circulation model based on the differential equation of cerebrospinal fluid dynamics. The cyclic model uses cerebrospinal fluid absorption resistance and generation rate as key variables. By iteratively fitting the current pressure gradient load value with the theoretical pressure value calculated by the model, the most matching cerebrospinal fluid absorption resistance value and generation rate value at the current moment can be derived. By comparing the relationship between the most suitable cerebrospinal fluid absorption resistance value and the generation rate value at the current moment, if the absorption resistance is significantly greater than the generation rate, the equilibrium state is determined to be inclined towards cerebrospinal fluid accumulation; if the generation rate is significantly greater than the absorption resistance, the equilibrium state is determined to be inclined towards excessive cerebrospinal fluid loss; if the two are similar, it is determined to be basically balanced. Based on the determined equilibrium state tendency and the specific numerical difference between absorption resistance and generation rate, the expected change in ventricular volume and direction of change in a preset time period are calculated using an empirical formula for the change of ventricular volume over time. The expected change and direction of change together constitute the predicted trend of ventricular volume change.

6. The postoperative nursing risk early warning system for neurosurgical patients according to claim 5, characterized in that, The generation of nursing early warning signals and corresponding physiological regulation pathways based on the brain tissue perfusion risk and neural structure compression risk includes: The brain tissue perfusion risk is compared with a preset perfusion risk threshold. When the threshold is exceeded, a brain perfusion maintenance warning is triggered, and an adjustment path is generated with the goal of increasing brain perfusion pressure. The adjustment path includes a recommended target mean arterial pressure range and blood pressure adjustment rate. The risk of neural structure compression is compared with a preset structural compression risk threshold. When the threshold is exceeded, an intracranial pressure control warning is triggered, and an adjustment path aimed at reducing intracranial volume is generated. The adjustment path includes suggested head position angle, hyperventilation parameters, and the timing of activation of osmotic dehydrating agents. When the risk of brain perfusion and the risk of neural structure compression both exceed their respective thresholds, an intracranial homeostasis imbalance warning is triggered, and a composite regulatory pathway that takes into account the contradiction between brain perfusion and intracranial volume is generated. The composite regulatory pathway specifies the priority and operation window for implementing measures to reduce intracranial pressure in stages while maintaining the minimum permissible brain perfusion pressure.

7. The postoperative nursing risk early warning system for neurosurgical patients according to claim 6, characterized in that, The generation of the regulatory pathway aimed at increasing cerebral perfusion pressure includes: Based on the specific value of the cerebral blood flow regulation reserve index, it is divided into three levels: sufficient reserve, reduced reserve, and depleted reserve. For levels of adequate blood reserve, a plan is set to gradually increase the mean arterial pressure, and it is stipulated that the increase in cerebral blood flow velocity should be monitored simultaneously during the adjustment process to ensure that it is proportional. For the level of reduced reserve, a step-by-step plan to gradually increase the mean arterial pressure is set up, and it is stipulated that after each step increase, the improvement of the slow wave oscillation energy of the EEG should be observed to determine whether to proceed to the next step. For each level of reserve depletion, an emergency volume expansion and vasopressor regimen is generated, and it is stipulated that central venous pressure and surface electromyography signals must be monitored in conjunction to prevent heart failure or worsening of myocardial edema caused by rapid volume expansion.

8. The postoperative nursing risk early warning system for neurosurgical patients according to claim 7, characterized in that, The generation of the regulatory pathway aimed at reducing intracranial volume includes: The main sources of the pressure gradient load in the cerebrospinal fluid circulation were analyzed to distinguish whether the main cause was impaired cerebrospinal fluid absorption or obstructed venous return. For cases where the main issue is impaired cerebrospinal fluid absorption, a pathway is developed that focuses on regulating the rate of cerebrospinal fluid production. This includes recommendations on the types of drugs that reduce cerebrospinal fluid production, initial doses, and methods for adjusting the dose based on changes in the coupling strength of respiratory waves in intracranial pressure waveforms. For cases where venous return is primarily obstructed, a pathway is generated with the optimization of intracranial venous return as its core. This includes specific operational steps and parameter combinations for reducing central venous pressure by adjusting the head of the bed, neck position, and controlling intrathoracic pressure.

9. The postoperative nursing risk early warning system for neurosurgical patients according to claim 8, characterized in that, The specific operational steps and parameter combinations for reducing central venous pressure by adjusting the head of the bed, neck position, and controlling intrathoracic pressure include: Raise the head of the bed from its initial position to the first target angle, and maintain this angle for the first preset observation time to monitor the decrease in the intensity of respiratory wave coupling in the intracranial pressure waveform; If the reduction is not as expected, further adjust the patient's neck position to the midline and slightly extend it backward, while monitoring the tension of the neck and shoulder muscles in the electromyography signal on the body surface to ensure that it does not increase significantly. After adjusting the patient's position, instruct them to use a slow ventilation mode that primarily involves abdominal breathing, and use a respiratory monitoring device to provide real-time feedback on the decrease in intrathoracic pressure. Based on feedback data from intracranial pressure respiratory wave coupling intensity, neck and shoulder electromyographic tension, and intrathoracic pressure, the head of the bed angle, neck extension angle, and breathing instructions are dynamically fine-tuned to form an optimal combination of body position and respiratory parameters that reduces the estimated central venous pressure to the target range.

Images