A post-operative patient monitoring system for neuro-interventional therapy

Abstract

This invention provides a postoperative patient monitoring system for neurointerventional therapy, relating to the fields of medical big data and artificial intelligence. It includes a multimodal data acquisition center, an interference feature decoupling unit, a trust capital quantification unit, a game strategy arbitration unit, and a monitoring execution unit. The multimodal data acquisition center is configured to retrieve time-series monitoring data of the monitored subjects. The trust capital quantification unit performs trust loss assessment and analysis on historical interaction feedback data. The game strategy arbitration unit is configured to compare and analyze the current clinical trust capital index with a preset trust threshold. This invention's system performs time-frequency domain matching between residual sequences and behavioral state labeling data, ensuring that the system only predicts the risk evolution of neurogenic hemodynamic changes. This significantly reduces the false alarm rate caused by external interference in complex postoperative monitoring environments and extracts pure pathological feature components.

Description

Technical Field

[0001] This invention relates to the fields of medical big data and artificial intelligence technology, and in particular to a postoperative patient monitoring system for neurointerventional therapy. Background Technology

[0002] With the widespread application of neurointerventional therapy, the postoperative patient monitoring environment is becoming increasingly complex, often facing multi-source noise interference caused by non-pathological factors such as pain, anxiety and body movement; this complexity poses a serious challenge to the false alarm control capability of the monitoring system and the maintenance of human-machine trust.

[0003] Currently, existing postoperative monitoring technologies typically rely on absolute thresholds of physiological parameters to trigger alarms, or rely solely on a single signal filtering algorithm for noise removal. This traditional approach primarily focuses on analyzing the physical characteristics of the physiological signals themselves, lacking effective monitoring of the interaction behavior and psychological trust status of medical staff.

[0004] However, this single-dimensional monitoring strategy has significant limitations. When encountering high-frequency behavioral coupling noise interference, the system often outputs false alarms frequently in order to ensure safety, leading to severe alarm fatigue and delayed response among medical staff. This mechanical alarm mechanism not only easily triggers a crisis of trust, but also causes the system to be gradually abandoned in long-term ineffective interactions, and it is impossible to make a dynamic trade-off between alarm sensitivity and user trust.

[0005] Therefore, how to accurately eliminate behavioral noise in complex and interfering environments, and how to adaptively adjust regulatory strategies based on the trust status of medical staff to prevent the collapse of trust, has become an urgent problem to be solved in this field. Summary of the Invention

[0006] In view of the shortcomings of the prior art, the purpose of this invention is to provide a postoperative patient monitoring system for neurointerventional therapy, which can solve the technical problems existing in the prior art.

[0007] A postoperative patient monitoring system for neurointerventional therapy includes a multimodal data acquisition center, an interference feature decoupling unit, a trust capital quantification unit, a game strategy arbitration unit, and a monitoring execution unit.

[0008] The multimodal data acquisition center is configured to retrieve time-series monitoring data of the monitored object and send the time-series monitoring data to the interference feature decoupling unit for signal source separation analysis to obtain pathological feature components and behavioral coupling noise components. Risk evolution prediction is performed on the pathological feature components to obtain the original risk prediction value.

[0009] The trust capital quantification unit is configured to acquire historical interaction feedback data of the regulatory enforcement unit, perform trust loss assessment and analysis on the historical interaction feedback data, and calculate the current clinical trust capital index, which characterizes the reliability of the system.

[0010] The game strategy arbitration unit is configured to compare and analyze the current clinical trust capital index with a preset trust threshold: if the current clinical trust capital index is greater than or equal to the preset trust threshold, an aggressive regulatory signal is generated; if the current clinical trust capital index is less than the preset trust threshold, a defensive regulatory signal is generated.

[0011] In response to the generation of the aggressive regulatory signal, the regulatory enforcement unit performs a high-sensitivity alarm operation based on the original risk prediction value;

[0012] In response to the generation of the defense monitoring signal, the monitoring execution unit is configured to invoke a preset interpretability verification protocol to perform downgrade verification processing on the original risk prediction value, generate a corrected risk output value, and perform adaptive intervention operations based on the corrected risk output value.

[0013] The signal source separation and analysis process is as follows:

[0014] The time-series monitoring data is collected during the collection period, and the time-series monitoring data is mapped to a preset generative adversarial network model to generate a digital twin simulation waveform corresponding to the collection period. The residual sequence between the time-series monitoring data and the digital twin simulation waveform is calculated.

[0015] The behavioral state label data of the monitored object is obtained, and the residual sequence is matched with the behavioral state label data in the time-frequency domain. If the match is successful, the high-frequency fluctuation part in the residual sequence is extracted and set as the behavioral coupling noise component, and the value obtained by subtracting the behavioral coupling noise component from the time-series monitoring data is set as the pathological feature component.

[0016] The trust loss assessment and analysis process is as follows:

[0017] Obtain the most recent alarm response delay duration from the historical interactive feedback data, and simultaneously obtain a preset effective treatment time window threshold. Compare and analyze the most recent alarm response delay duration with the effective treatment time window threshold.

[0018] If the delay time of the most recent alarm response is less than the effective treatment time window threshold, a preset trust maintenance coefficient is selected; if the delay time of the most recent alarm response is greater than or equal to the effective treatment time window threshold, a preset trust decay coefficient is selected, wherein the value of the trust decay coefficient is greater than the value of the trust maintenance coefficient.

[0019] Obtain the total number of historical false alarms from the historical interaction feedback data, calculate the product of the total number of historical false alarms and the currently selected coefficient to obtain the trust penalty value; calculate the difference between the preset initial trust benchmark value and the trust penalty value, and set the difference as the current clinical trust capital index.

[0020] The downgrade verification process is as follows:

[0021] Calculate the difference between the preset trust threshold and the current clinical trust capital index, and determine the corresponding trust deficit level based on the difference;

[0022] According to the trust deficit level, the corresponding feature filtering window parameters are called from the preset strategy library. The higher the trust deficit level, the longer the time length of the corresponding feature filtering window parameters. The preset strategy library is a mapping table that stores the function mapping relationship between the trust deficit level and the feature filtering window parameters. The mapping table pre-stores discrete trust deficit levels and corresponding numerical feature filtering window parameters for parameter retrieval and matching.

[0023] The original risk prediction value is smoothed over time using the feature filtering window parameter to filter out transient peaks in the original risk prediction value whose duration is less than the feature filtering window parameter, thus obtaining smoothed risk trend data.

[0024] The smoothed risk trend data is set as the corrected risk output value, wherein the alarm trigger sensitivity of the corrected risk output value is lower than the alarm trigger sensitivity of the original risk prediction value.

[0025] When the aggressive regulatory signal is generated, the game strategy arbitration unit is also used to perform the following operations:

[0026] The slope of the change of the pathological feature component is obtained, and the slope of the change is compared with a preset threshold for the slope of the prodromal feature.

[0027] If the slope of change is greater than or equal to the threshold of the precursor feature slope, the original risk prediction value is directly marked as a Level 1 critical state, and the regulatory execution unit is instructed to forcibly output an alarm command while ignoring the behavioral coupling noise component.

[0028] The behavioral coupling noise component represents postoperative sympathetic storm artifacts, and its feature extraction process includes:

[0029] Extract the blood pressure fluctuation amplitude and heart rate variability from the time-series monitoring data to construct a multidimensional physiological feature vector;

[0030] The multidimensional physiological feature vector is input into a preset artifact recognition classifier. If the classification result output by the artifact recognition classifier indicates non-neurogenic fluctuation, then the waveform segment corresponding to the multidimensional physiological feature vector is determined to be the behavioral coupling noise component.

[0031] The adaptive intervention process is as follows:

[0032] When the corrected risk output value exceeds the preset safety baseline, the regulatory execution unit generates a verification request signal containing evidence tracing information;

[0033] The verification request signal is used to simultaneously display the elimination criteria map of the behavior coupling noise component while outputting an alarm prompt, record the user's confirmation operation of the verification request signal, and feed the confirmation operation back to the trust capital quantification unit to update the historical interaction feedback data.

[0034] The trust capital quantification unit is also used to perform the following prediction operations:

[0035] Extract multiple current clinical trust capital indices from historical time periods and construct a trust depletion prediction curve using a time series regression algorithm;

[0036] Calculate the remaining number of alarms required for the value of the trust depletion prediction curve to drop to the preset response failure zero point, and set the remaining number of alarms as the security game budget value;

[0037] The secure game budget value is compared with a preset budget security threshold: if the secure game budget value is less than or equal to the budget security threshold, a trust warning signal is generated and sent to the game strategy arbitration unit;

[0038] The game strategy arbitration unit increases the preset trust threshold in response to receiving the trust warning signal.

[0039] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0040] Firstly, this system utilizes a multimodal data acquisition and interference feature decoupling unit to construct a digital twin simulation waveform of each patient using an improved generative adversarial network, which serves as a physiological benchmark for calculating residual sequences. The system performs time-frequency domain matching between the residual sequences and behavioral state labeling data, accurately identifying and eliminating behaviorally coupled noise components caused by non-pathological factors such as pain, anxiety, or body movement. This mechanism ensures that the system only predicts the risk evolution of neurogenic hemodynamic changes, thereby significantly reducing the false alarm rate caused by external interference in complex postoperative monitoring environments and extracting pure pathological feature components.

[0041] Secondly, this system innovatively introduces a trust capital quantification unit, transforming the abstract psychological state of medical staff into a calculable clinical trust capital index. The system is not limited to the analysis of physiological signals, but dynamically assesses the level of trust medical staff have in the equipment by monitoring historical interaction feedback data, particularly alarm response delay duration and the total number of historical false alarms. This mechanism reveals the dynamic principle of response delay as a sign of trust collapse, enabling the system to keenly perceive a trust crisis before being completely abandoned by medical staff and adjust its strategy according to the trust depletion prediction curve, preventing the system from falling into irreversible trust bankruptcy.

[0042] Third, this system achieves adaptive switching of monitoring modes through a game-theoretic arbitration unit. During periods of high trust capital index (a "bonus period"), the system executes an aggressive monitoring mode, employing highly sensitive alarms to ensure zero missed alarms. During periods of low trust capital index (a "deficit period"), the system automatically switches to a defensive monitoring mode, invoking an interpretability verification protocol and utilizing feature filtering windows for time smoothing to filter out transient peaks. This tactical retreat strategy sacrifices some temporal resolution for extremely high alarm specificity, effectively reducing disruptive alarms and helping to gradually restore human-machine trust when medical staff lack confidence, thus maintaining the system's long-term on-orbit operation.

[0043] Fourth, while pursuing low false alarms, this system also considers safety and transparency in extreme situations. On the one hand, the system has a mandatory alarm circuit breaker mechanism. When the slope of the change in the pathological feature component exceeds the threshold of the prodromal feature slope, the system will determine it as a first-level critical state and directly ignore any noise interference to force an alarm output, ensuring that catastrophic lesions are not missed. On the other hand, in defense mode, the system can generate a verification request signal containing evidence tracing information, which intuitively displays the basis spectrum of noise removal to medical staff. This visualized evidence chain greatly enhances the persuasiveness of the alarm and assists doctors in making rapid clinical decisions. Attached Figure Description

[0044] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0045] Figure 1 This is a system block diagram of a postoperative patient monitoring system for neurointerventional therapy provided in an embodiment of the present invention. Detailed Implementation

[0046] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions 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, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0047] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention.

[0048] Example 1

[0049] Reference Figure 1 As shown, a postoperative patient monitoring system for neurointerventional therapy includes a multimodal data acquisition center, an interference feature decoupling unit, a trust capital quantification unit, a game strategy arbitration unit, and a monitoring execution unit. The multimodal data acquisition center is configured to retrieve time-series monitoring data of the monitored object and send the time-series monitoring data to the interference feature decoupling unit for signal source separation analysis to obtain pathological feature components and behavioral coupling noise components. Risk evolution prediction is then performed on the pathological feature components to obtain the original risk prediction value. The trust capital quantification unit is configured to acquire historical interaction feedback data from the monitoring execution unit, perform trust loss assessment analysis on the historical interaction feedback data, and calculate the current clinical trust value characterizing the system's reliability. The clinical trust capital index is used as the basis for the game strategy arbitration unit to compare and analyze the current clinical trust capital index with a preset trust threshold. If the current clinical trust capital index is greater than or equal to the preset trust threshold, an aggressive regulatory signal is generated; if the current clinical trust capital index is less than the preset trust threshold, a defensive regulatory signal is generated. In response to the generation of the aggressive regulatory signal, the regulatory execution unit performs a high-sensitivity alarm operation based on the original risk prediction value. In response to the generation of the defensive regulatory signal, the regulatory execution unit is configured to call a preset interpretability verification protocol to perform downgrade verification processing on the original risk prediction value, generate a corrected risk output value, and perform an adaptive intervention operation based on the corrected risk output value.

[0050] This embodiment details the overall architecture and data flow logic of a post-neurointerventional therapy monitoring system, aiming to construct an intelligent monitoring closed loop with clinical metacognitive capabilities. First, the system establishes a physical connection through a multimodal data acquisition center, retrieving raw signal streams from bedside monitors and wearable devices in real time. This data stream encompasses high-frequency sampled cerebral blood flow velocity and systemic physiological parameters. Subsequently, an interference feature decoupling unit acts as the system's sensing filter, introducing a dynamic decoupling mechanism in the signal source separation analysis to decompose the mixed raw waveforms into pure pathological features and behavioral noise. Following this, the system targets the pathological feature components... Risk evolution prediction is performed. To capture the uncertainty and dynamic trends of risk evolution over time, the system employs an Ornstein-Uhlenbeck process framework for modeling, a framework commonly used to describe mean-reverting stochastic processes. In this application scenario, the risk state... Modeled as surrounding a pathological stimulus A stochastic process driven by a time-varying mean. The constructed risk state evolution equation is as follows:

[0051]

[0052] in, This represents the original risk forecast value (dimensionless); The decay rate constant (1 / s) characterizes the rate at which the risk state naturally decays. The excitation gain constant (the unit depends on the excitation gain constant). ,like Blood pressure is expressed in units of 1 / (s·mmHg), representing the pathological characteristic component of a unit. The driving force behind changes in risk status; The noise intensity constant (1 / √s) reflects the random fluctuations not captured by the model; These are pathological characteristic components (the units depend on the specific physiological signal, such as blood pressure in mmHg and cerebral blood flow velocity in cm / s); This represents the increment of a standard Wiener process at time step. The inner follows a distribution .

[0053] parameter , , The estimation is based on historical case data. The system uses the Euler-Maruyem discretization method to discretize the above SDE, and constructs a parameter set based on the transition probabilities of the discretized sequence. The likelihood function, obtained by analyzing historical data... and The optimal parameters are obtained by performing maximum likelihood estimation (MLE) on the synchronization sequence.

[0054] The system uses the Euler-Maruye discretization method to solve the stochastic differential equation, with a set time step. The current time is obtained through iterative calculation. Value; where the parameter and Based on The likelihood function was constructed based on the distribution properties and obtained through maximum likelihood estimation of historical case data;

[0055] Specifically, based on the discretized Markov property, the system constructs a parameter... The negative log-likelihood function is as follows:

[0056]

[0057] L represents the loss function. Represented as the total number of samples, It is represented as the conditional mean.

[0058] Wherein, the conditional mean is defined as By minimizing the objective function The system can reverse engineer the optimal dynamic parameters from noisy historical data. Based on this, the trust capital quantification unit works in parallel; it does not directly process physiological signals but instead monitors the feedback logs of the human-computer interaction interface in real time, quantifying the degree of psychological dependence of medical staff on the system, i.e., the current clinical trust capital index. Next, the game strategy arbitration unit, based on... With preset trust threshold Based on the comparison results, the system's operating strategy is dynamically adjusted: when When the system determines that it is in a period of high trust, it activates an aggressive regulatory mode to pursue zero false negatives; conversely, when... When the system determines that it is in a period of trust deficit, it automatically switches to defensive monitoring mode, sacrificing some real-time performance in exchange for high confidence by calling the interpretability verification protocol. Finally, the monitoring execution unit executes high-sensitivity alarms or adaptive intervention operations with evidence chain display according to the arbitration instructions, thereby completing a complete closed loop from signal perception to decision execution. This represents the discretization time step.

[0059] Example 2

[0060] The signal source separation and analysis process is as follows: The acquisition period of the time-series monitoring data is collected, and the time-series monitoring data is mapped to a preset generative adversarial network model to generate a digital twin analog waveform corresponding to the acquisition period. The residual sequence between the time-series monitoring data and the digital twin analog waveform is calculated. Behavioral state labeling data of the monitored object is obtained, and the residual sequence is matched with the behavioral state labeling data in the time-frequency domain. If the match is successful, the high-frequency fluctuation portion in the residual sequence is extracted and set as the behavioral coupling noise component, and the value obtained by subtracting the behavioral coupling noise component from the time-series monitoring data is set as the pathological feature component.

[0061] This embodiment details the core signal source separation and analysis logic in the interference feature decoupling unit. This logic uses an improved generative adversarial network to construct an individualized physiological benchmark. First, the system extracts the current acquisition time period. Time series monitoring data within This is then mapped to a predefined model. To address the issues of standard GANs failing to generate waveforms corresponding to specific inputs and the tendency for latent vectors to wander outside the effective manifold, this model constructs a variational encoder-decoder adversarial architecture. This architecture includes an encoder... Generator and discriminator The system constructs a hybrid objective function that includes reconstruction error, KL divergence regularization, and bidirectional adversarial constraints, as follows:

[0062]

[0063] In the formula, Represents the distribution of real data Find the expected value; Indicates the prior distribution Hidden variables in mid-sample Find the expected value; Represents the natural logarithm function. Represents a mixed objective function. Indicates the reconstructed weights. Indicates the regularization weight. divergence, This represents the prior distribution. The output distribution of the constraint encoder approximates a standard normal distribution. ,and The waveform generated by the explicit constraint reconstructing the path must possess realistic statistical characteristics; based on this model, the system executes the deduction steps: firstly, through... The observed data is encoded into latent feature vectors, and then... Generate digital twin analog waveforms corresponding to the acquisition period. This waveform represents the ideal state of the patient under current physiological benchmarks without external interference, and retains high-frequency details. Next, the difference between the observed data and the digital twin simulated waveform is calculated, yielding the residual sequence formula as follows:

[0064]

[0065] in, Represents the residual sequence. This represents time-series monitoring data. This represents the analog waveform of a digital twin.

[0066] The sequence physically includes all anomalous fluctuations deviating from the baseline; subsequently, external behavioral state marker data from accelerometers or video analytics are introduced. ,Will and Perform cross-correlation analysis in the time and frequency domains; calculate the residual sequence. With acceleration data Maximum cross-correlation number ,like If the power spectral density overlap rate between the two components in the 10Hz-20Hz frequency band is greater than 60%, they are considered a match. The system determines that this fluctuation originates from non-pathological factors and extracts it as a behavioral coupling noise component. Finally, the subtraction operation is performed as follows:

[0067]

[0068] in, Indicates pathological feature components, This represents the behavioral coupled noise component.

[0069] This allows for the acquisition of pure pathological characteristic components. This component only retains the neurogenic hemodynamic changes.

[0070] Example 3

[0071] The trust loss assessment and analysis process is as follows: The most recent alarm response delay duration is obtained from the historical interactive feedback data, and a preset effective treatment time window threshold is obtained simultaneously. The most recent alarm response delay duration is compared and analyzed with the effective treatment time window threshold. If the most recent alarm response delay duration is less than the effective treatment time window threshold, a preset trust maintenance coefficient is selected. If the most recent alarm response delay duration is greater than or equal to the effective treatment time window threshold, a preset trust decay coefficient is selected, wherein the value of the trust decay coefficient is greater than the value of the trust maintenance coefficient. The total number of historical false alarms is obtained from the historical interactive feedback data, and the product of the total number of historical false alarms and the currently selected coefficient is calculated to obtain a trust penalty value. The difference between the preset initial trust benchmark value and the trust penalty value is calculated, and the difference is set as the current clinical trust capital index.

[0072] This embodiment details the mathematical modeling process of the trust capital quantification unit, aiming to transform the abstract psychological state of medical staff into a calculable numerical indicator. First, the system extracts key time-series parameters from the interaction log, namely the delay time of the most recent alarm response. And compare it with the preset effective treatment time window threshold. Perform a comparison; the aforementioned Based on the survival curve of the ischemic penumbra, the value is set as the statistical time threshold for irreversible neurological damage. ,Right now Minutes, minus emergency preparation time Next, the instantaneous loss factor is determined based on the comparison results. The logic is as follows:

[0073] like

[0074] in, Indicates the instantaneous loss factor. Indicates the response delay duration. This represents the effective treatment time window threshold, where: The value is set to 0.05, which is based on the average cognitive wear and tear statistics of normal medical procedures. Set to 0.8, and satisfy... This value is derived from a weighted statistical analysis of delayed responses in historical medical dispute data; , , The settings were derived from a questionnaire survey and behavioral log analysis of 50 medical staff.

[0075] To prevent the trust capital index from collapsing non-linearly due to a single interaction, the system introduces an inertial smoothing mechanism to calculate the currently selected coefficient. The formula is as follows:

[0076]

[0077] in, This indicates the currently selected coefficient. Indicates memory weights, This represents the coefficient at the previous time step.

[0078] in Set as This ensures that trust evaluations have historical memory; subsequently, it is combined with the number of false alarms within the sliding time window. Calculate the trust penalty value generated by this interaction. Finally, using the initial trust baseline value Perform a subtraction update; here, the initial trust baseline value Preset as dimensionless value This symbolizes the complete trust that medical staff have in the device when the system is first started, i.e., 100%; the calculation formula is as follows:

[0079]

[0080] in, This is represented as the current clinical trust capital index. This is represented as the initial trust baseline value. This is represented as a trust penalty value. This embodiment innovatively introduces response delay as a non-linear weighting factor in trust assessment, profoundly revealing the dynamic mechanism of clinical trust collapse; when healthcare workers begin to delay processing alerts due to distrust, the system can keenly capture this change in behavioral pattern and apply a smoothly increasing penalty value. Gradually reduce the trust capital index; this mechanism simulates the vicious cycle in the real world where doubt leads to delay, and delay accelerates failure, enabling the system to perceive the crisis before it is completely abandoned and providing a quantitative basis for subsequent strategy adjustments.

[0081] After calculating the difference, the system adjusts the current clinical trust capital index. Normalization is performed to ensure that it falls within a reasonable range of representation, such as limiting it to an interval. Inside: This process simulates trust capital ranging from a minimum of zero to a maximum of the initial value. The actual situation.

[0082] Example 4

[0083] The downgrade verification process is as follows: Calculate the difference between the preset trust threshold and the current clinical trust capital index, and determine the corresponding trust deficit level based on the difference; Call the corresponding feature filtering window parameter from the preset strategy library according to the trust deficit level, wherein the higher the trust deficit level, the longer the time length of the corresponding feature filtering window parameter; wherein the preset strategy library is a mapping table storing the function mapping relationship between trust deficit levels and feature filtering window parameters. This mapping table pre-stores discrete trust deficit levels and corresponding numerical feature filtering window parameters for parameter retrieval and matching. The preset strategy library establishes a quantitative mapping model between trust deficit levels and feature filtering window parameters to achieve smoothing intensity control under different trust scenarios; Use the feature filtering window parameter to perform time smoothing processing on the original risk prediction value, filtering out transient peaks in the original risk prediction value whose duration is less than the feature filtering window parameter, to obtain smoothed risk trend data; Set the smoothed risk trend data as the corrected risk output value, wherein the alarm trigger sensitivity of the corrected risk output value is lower than the alarm trigger sensitivity of the original risk prediction value.

[0084] This embodiment details the degradation verification processing logic in the defensive monitoring mode, which sacrifices time resolution for alarm specificity; first, the trust deficit is calculated. And based on this, quantify the level of trust deficit. The process of quantifying the trust deficit level follows the principle of rounding down: ,in This indicates the floor function. Next, with... To retrieve the corresponding feature filtering window parameters from the pre-defined strategy library for indexed queries, This parameter is positively correlated with the deficit level; that is, the lower the trust level, the wider the window. The specific function mapping relationship is defined as follows:

[0085]

[0086] in, Represented as feature filtering window parameters, This represents the base window length. Represented as the grade expansion coefficient, Represented as a trust deficit level; the quantitative mapping model uses a linear function This ensures the feature filtering window parameters are implemented. It can expand in stages as the trust deficit increases;

[0087] Set as Second, Set as Seconds / level; subsequently, using For the original risk forecast value The moving average filtering operation is performed, and the specific calculation formula is as follows:

[0088]

[0089] This is represented as the corrected risk output value. This is represented as the original risk forecast value. Represented as feature filtering window parameters, Indicates the current time. It is represented as a discrete lag index.

[0090] This formula aims to filter out those with a duration of less than [a certain value]. The transient spike signal; finally, the smoothed signal... The output is a corrected risk output value, which has a significantly lower alarm trigger sensitivity than the original value and only reflects long-term and stable risk trends.

[0091] In this embodiment, when the system encounters a crisis of trust, a slow but robust monitoring strategy is automatically activated. By dynamically extending the observation window, the system effectively filters out occasional physiological fluctuations and sensor noise. Although this delays the alarm time to some extent, it greatly improves the positive predictive value of the alarm. This tactical retreat strategy is particularly crucial in scenarios where medical staff lack confidence in the equipment. It gradually restores the trust relationship between humans and machines by reducing harassing alarms.

[0092] Example 5

[0093] When the aggressive regulatory signal is generated, the game strategy arbitration unit is also used to perform the following operations: obtain the slope of change of the pathological feature component, compare the slope of change with a preset warning feature slope threshold; if the slope of change is greater than or equal to the warning feature slope threshold, directly mark the original risk prediction value as a first-level critical state, and instruct the regulatory execution unit to forcibly output an alarm command while ignoring the behavioral coupling noise component.

[0094] This embodiment details the mandatory alarm circuit breaker mechanism under aggressive monitoring mode, designed to address catastrophic acute lesions; firstly, the system analyzes pathological feature components. Perform first-order differential operations and calculate its slope. This indicator characterizes the rate of disease progression; then, Compared with the preset precursor feature slope threshold Perform real-time comparison; respond to The system determines that it is currently in an extremely dangerous outbreak period, that is, the moment the aneurysm ruptures. At this time, any denoising algorithm may mistakenly delete the real signal. Therefore, the system immediately bypasses the conventional game logic and marks the original risk prediction value as a level one critical state. Finally, the instruction supervision and execution unit forcibly outputs the highest priority alarm instruction, completely ignoring the existence of behavioral coupling noise components in the process.

[0095] Example 6

[0096] The behavioral coupling noise component represents postoperative sympathetic storm artifacts. Its feature extraction process includes: extracting the blood pressure fluctuation amplitude and heart rate variability from the time-series monitoring data to construct a multidimensional physiological feature vector; inputting the multidimensional physiological feature vector into a preset artifact recognition classifier; if the classification result output by the artifact recognition classifier indicates non-neurogenic fluctuations, then the waveform segment corresponding to the multidimensional physiological feature vector is determined to be the behavioral coupling noise component.

[0097] This embodiment details the feature extraction and identification process for postoperative sympathetic storm artifacts. First, features from two key dimensions are extracted in parallel from time-series monitoring data: blood pressure fluctuation amplitude. With heart rate variability And combine them to construct a multidimensional physiological feature vector. This vector comprehensively reflects the regulatory state of the autonomic nervous system; next, it will... The input is fed into a pre-trained artifact recognition classifier, namely a Support Vector Machine (SVM); this classifier is based on an expert-annotated set of sympathetic storm artifact samples. Trained using a radial basis function kernel:

[0098]

[0099] in, Represents the kernel function. Represents the input vector. Represents the center vector. Indicates kernel parameters.

[0100] Wherein, the kernel parameters The distribution complexity after mapping data to a high-dimensional feature space is controlled, while the penalty coefficient... This is used to balance the classification margin and the number of misclassified samples. In this embodiment, the system uses a combination of cross-validation and grid search to simultaneously... and To perform optimization, set The search scope is , The search scope is The system obtains the parameter combination that corresponds to the highest accuracy of the model on the validation set. If the classification result output by the artifact recognition classifier indicates non-neurogenic fluctuations, i.e. caused by peripheral stimuli such as pain and anxiety, the system determines that the current waveform segment belongs to the behavioral coupling noise component. Finally, the waveform segment is marked for use by the subsequent decoupling unit.

[0101] Example 7

[0102] The adaptive intervention process is as follows: When the corrected risk output value exceeds the preset safety baseline, the regulatory execution unit generates a verification request signal containing evidence tracing information; the verification request signal is used to simultaneously display the elimination basis map of the behavior coupling noise component while outputting an alarm prompt, and record the user's confirmation operation of the verification request signal, and feed the confirmation operation back to the trust capital quantification unit to update the historical interaction feedback data.

[0103] This embodiment details the interaction logic in adaptive intervention, focusing on rebuilding user trust through white-box presentation; firstly, it monitors and corrects risk output values ​​in real time. In response to its value exceeding the preset safety baseline The system triggers a verification request signal. Then, it simultaneously renders two layers of information on the user interface: a standard alarm message and crucial evidence tracing information, namely a graph illustrating the basis for eliminating behaviorally coupled noise components. This graph visually displays the noise segments identified by the system, i.e., motion artifacts, and a waveform comparison before and after their removal. Subsequently, the system records the user's confirmation action on the request, i.e., confirming validity or marking a false alarm, completing a full human-computer interaction. Finally, the result of this action is fed back as a new data point to the trust capital quantification unit for real-time updates of the total number of historical false alarms. and response delay .

[0104] Example 8

[0105] The trust capital quantification unit is also used to perform the following prediction operations:

[0106] Extract multiple current clinical trust capital indices from historical time periods and construct a trust depletion prediction curve using a time series regression algorithm;

[0107] Calculate the remaining number of alarms required for the value of the trust depletion prediction curve to drop to the preset response failure zero point, and set the remaining number of alarms as the security game budget value;

[0108] The secure game budget value is compared with a preset budget security threshold: if the secure game budget value is less than or equal to the budget security threshold, a trust warning signal is generated and sent to the game strategy arbitration unit;

[0109] In response to receiving the trust warning signal, the game strategy arbitration unit increases the preset trust threshold. The preset trust threshold is set based on the false alarm rate and response delay statistics of the first 100 patient interaction data.

[0110] This embodiment details the advanced predictive capabilities of the trust capital quantification unit, aiming to prevent the system from falling into irreversible trust bankruptcy. First, it extracts multiple current clinical trust capital index sequences within a past time window. Next, the least squares method is used to perform a linear fit on the sequence to construct a confidence exhaustion prediction curve, as shown in the following formula:

[0111]

[0112] in, Represented as a prediction curve, Represented as the current index, This is expressed as the average loss rate. The average loss rate is represented by the number of future alarms. The statistical characteristics of time indexes and capital values ​​were clearly distinguished through calculations using the least squares estimation formula.

[0113]

[0114] In the formula, This represents the average loss rate. Represents the sequence index within the historical time window (e.g.) ), This represents the arithmetic mean of the sequence indices; Indicates the first The current clinical trust capital index value corresponding to each time point. This represents the arithmetic mean of all trust capital indices within the historical window. Let there be a total of [number missing] trust capital indices within the historical window. The trust capital index at each point in time is denoted as the sequence. Its corresponding time index is . for The arithmetic mean, for The arithmetic mean.

[0115] Here The definition avoids the index mean The confusion; then, solving the equation The calculated curve drops to the preset response failure zero point. The remaining number of alarms required is defined as the security game budget value. ; This represents a critical state where doctors completely ignore the alarm; finally, Send to the game strategy arbitration unit, in response to If the threshold falls below the security warning line, the arbitration unit will significantly increase the preset trust threshold. This forces the system to enter defensive mode prematurely; the specific adjustment calculations are as follows:

[0116]

[0117] In the formula, This is represented as the adjusted threshold. Represented as the original threshold, This is expressed as the gain magnitude. Represented as sensitivity factor, Represented as the safe game budget value, this nonlinear formula ensures that when the remaining budget... Approaching At this point, the threshold will rise sharply, thereby tightening regulatory policies. In this embodiment, the gain magnitude... Set to 0.5 to limit the maximum threshold fluctuation to no more than 50%; Sensitivity Factor The value is set to 0.2, which is derived from the statistical fitting of the average number of alarms that doctors abandoned in historical cases.

[0118] This embodiment endows the system with a self-protective metacognitive ability; by predicting the critical point of trust depletion, the system no longer passively waits to be abandoned by the user, but actively adjusts its strategy according to the remaining trust budget; when trust is about to collapse, the system can intelligently shrink its defenses and maintain the continuation of the system's vitality during long-term postoperative monitoring.

[0119] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the preferred embodiments, while those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A postoperative patient monitoring system for neurointerventional therapy, characterized in that, It includes a multimodal data acquisition center, an interference feature decoupling unit, a trust capital quantification unit, a game strategy arbitration unit, and a regulatory enforcement unit; The multimodal data acquisition center is configured to retrieve time-series monitoring data of the monitored object and send the time-series monitoring data to the interference feature decoupling unit for signal source separation analysis to obtain pathological feature components and behavioral coupling noise components. Risk evolution prediction is performed on the pathological feature components to obtain the original risk prediction value. The trust capital quantification unit is configured to acquire historical interaction feedback data of the regulatory enforcement unit, perform trust loss assessment and analysis on the historical interaction feedback data, and calculate the current clinical trust capital index, which characterizes the reliability of the system. The game strategy arbitration unit is configured to compare and analyze the current clinical trust capital index with a preset trust threshold: if the current clinical trust capital index is greater than or equal to the preset trust threshold, an aggressive regulatory signal is generated; if the current clinical trust capital index is less than the preset trust threshold, a defensive regulatory signal is generated. In response to the generation of the aggressive regulatory signal, the regulatory enforcement unit performs a high-sensitivity alarm operation based on the original risk prediction value; In response to the generation of the defense monitoring signal, the monitoring execution unit is configured to invoke a preset interpretability verification protocol to perform downgrade verification processing on the original risk prediction value, generate a corrected risk output value, and perform adaptive intervention operations based on the corrected risk output value. The trust loss assessment and analysis process is as follows: Obtain the most recent alarm response delay duration from the historical interactive feedback data, and simultaneously obtain a preset effective treatment time window threshold. Compare and analyze the most recent alarm response delay duration with the effective treatment time window threshold. If the delay time of the most recent alarm response is less than the effective treatment time window threshold, a preset trust maintenance coefficient is selected; if the delay time of the most recent alarm response is greater than or equal to the effective treatment time window threshold, a preset trust decay coefficient is selected, wherein the value of the trust decay coefficient is greater than the value of the trust maintenance coefficient. Obtain the total number of historical false alarms from the historical interaction feedback data, calculate the product of the total number of historical false alarms and the currently selected coefficient, and obtain the trust penalty value; Calculate the difference between the preset initial trust benchmark value and the trust penalty value, and set the difference as the current clinical trust capital index; The trust capital quantification unit is also used to perform the following prediction operations: Extract multiple current clinical trust capital indices from historical time periods and construct a trust depletion prediction curve using a time series regression algorithm; Calculate the remaining number of alarms required for the value of the trust depletion prediction curve to drop to the preset response failure zero point, and set the remaining number of alarms as the security game budget value; The secure game budget value is compared with a preset budget security threshold: if the secure game budget value is less than or equal to the budget security threshold, a trust warning signal is generated and sent to the game strategy arbitration unit; The game strategy arbitration unit increases the preset trust threshold in response to receiving the trust warning signal.

2. The postoperative patient monitoring system for neurointerventional therapy according to claim 1, characterized in that, The signal source separation and analysis process is as follows: The time-series monitoring data is collected during the collection period, and the time-series monitoring data is mapped to a preset generative adversarial network model to generate a digital twin simulation waveform corresponding to the collection period. The residual sequence between the time-series monitoring data and the digital twin simulation waveform is calculated. The behavioral state label data of the monitored object is obtained, and the residual sequence is matched with the behavioral state label data in the time-frequency domain. If the match is successful, the high-frequency fluctuation part in the residual sequence is extracted and set as the behavioral coupling noise component, and the value obtained by subtracting the behavioral coupling noise component from the time-series monitoring data is set as the pathological feature component.

3. The postoperative patient monitoring system for neurointerventional therapy according to claim 1, characterized in that, The downgrade verification process is as follows: Calculate the difference between the preset trust threshold and the current clinical trust capital index, and determine the corresponding trust deficit level based on the difference; According to the trust deficit level, the corresponding feature filtering window parameters are called from the preset strategy library. The higher the trust deficit level, the longer the time length of the corresponding feature filtering window parameters. The preset strategy library is a mapping table that stores the function mapping relationship between the trust deficit level and the feature filtering window parameters. The mapping table pre-stores discrete trust deficit levels and corresponding numerical feature filtering window parameters for parameter retrieval and matching. The original risk prediction value is smoothed over time using the feature filtering window parameter to filter out transient peaks in the original risk prediction value whose duration is less than the feature filtering window parameter, thus obtaining smoothed risk trend data. The smoothed risk trend data is set as the corrected risk output value, wherein the alarm trigger sensitivity of the corrected risk output value is lower than the alarm trigger sensitivity of the original risk prediction value.

4. A postoperative patient monitoring system for neurointerventional therapy according to claim 1, characterized in that, When the aggressive regulatory signal is generated, the game strategy arbitration unit is also used to perform the following operations: The slope of change of the pathological feature component is obtained, and the slope of change is compared with a preset threshold for the slope of the prodromal feature. If the slope of change is greater than or equal to the threshold of the precursor feature slope, the original risk prediction value is directly marked as a Level 1 critical state, and the regulatory execution unit is instructed to forcibly output an alarm command while ignoring the behavioral coupling noise component.

5. A postoperative patient monitoring system for neurointerventional therapy according to claim 2, characterized in that, The behavioral coupling noise component represents postoperative sympathetic storm artifacts, and its feature extraction process includes: Extract the blood pressure fluctuation amplitude and heart rate variability from the time-series monitoring data to construct a multidimensional physiological feature vector; The multidimensional physiological feature vector is input into a preset artifact recognition classifier. If the classification result output by the artifact recognition classifier indicates non-neurogenic fluctuation, then the waveform segment corresponding to the multidimensional physiological feature vector is determined to be the behavioral coupling noise component.

6. A postoperative patient monitoring system for neurointerventional therapy according to claim 1, characterized in that, The adaptive intervention process is as follows: When the corrected risk output value exceeds the preset safety baseline, the regulatory execution unit generates a verification request signal containing evidence tracing information; The verification request signal is used to simultaneously display the elimination criteria map of the behavior coupling noise component while outputting an alarm prompt, record the user's confirmation operation of the verification request signal, and feed the confirmation operation back to the trust capital quantification unit to update the historical interaction feedback data.

Images