A mixed reality-based anesthetic stress visualized digital management system
By employing multi-source data sensing, pharmacokinetic manifold reconstruction, and mixed reality holographic mapping technology, the problem of distinguishing between real physiological responses and environmental noise artifacts in anesthesia monitoring systems has been solved. This has enabled precise stress intensity management and visual monitoring, reduced false positive rates, and improved surgical safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE FIRST AFFILIATED HOSPITAL OF HENAN UNIV OF SCI & TECH
- Filing Date
- 2026-04-09
- Publication Date
- 2026-07-03
AI Technical Summary
Existing anesthesia monitoring systems cannot effectively distinguish between physiological responses caused by real noxious stimuli and environmental noise artifacts, resulting in a high false positive rate. They also cannot provide accurate stress intensity indicators, increasing the cognitive load on physicians and failing to meet the needs for precise control and visualization of anesthesia depth during the perioperative period.
Multi-source data sensing units are used for time-series synchronous processing. Combined with pharmacokinetic manifold reconstruction, active simulation of noxious stimuli, and dual differential verification of spatiotemporal features, true stress determination labels are generated. Real-time visualization management with anti-interference is achieved through mixed reality holographic mapping technology.
It enables accurate identification and noise removal of physiological signals, reduces false positive rates, provides highly reliable stress intensity indicators, reduces the cognitive load on doctors, and improves surgical safety.
Smart Images
Figure CN122337489A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent processing of biomedical signals and mixed reality interaction technology, specifically to a mixed reality-based visualization digital management system for anesthesia stress. Background Technology
[0002] In the current surgical anesthesia monitoring environment, devices such as monitors, anesthesia machines and infusion workstations generate a large amount of multimodal vital signs data and drug administration records in real time. The data sources are scattered and lack a unified time-series benchmark.
[0003] To assess patient stress, existing methods generally employ alarm mechanisms based on static thresholds or historical averages, triggering alarms by monitoring fluctuations in the amplitude of a single indicator. While this approach can capture signal changes, it neglects the dynamic impact of drug metabolism on the patient's physiological baseline and lacks proactive deduction of the causal logic of surgical procedures. This results in the system's inability to effectively distinguish between physiological responses caused by genuine noxious stimuli and artifacts caused by environmental noise such as electrosurgical interference and changes in body position. This passive, non-mechanistic monitoring mode leads to a high false positive rate and fails to provide accurate stress intensity indicators after noise removal. It forces physicians to expend considerable effort to distinguish between genuine and false signals, increasing cognitive load and failing to meet the needs for precise control and intuitive visualization of anesthesia depth during the perioperative period. Therefore, how to construct a mechanism-based closed-loop monitoring model to accurately identify genuine and false signals and achieve interference-resistant real-time visual management has become an urgent technical problem to be solved. Summary of the Invention
[0004] To address the aforementioned technical problems, this invention provides a visual digital management system for anesthesia stress based on mixed reality. Specifically, the technical solution of this invention includes:
[0005] The multi-source data sensing unit is used to collect multimodal vital signs, perioperative medication records, and surgical operation markers in real time, and to perform time-series synchronization processing on the collected data to generate a synchronous panoramic data stream.
[0006] The pharmacokinetic manifold reconstruction unit is used to receive the synchronous panoramic data stream, calculate the in vivo drug effect chamber concentration in real time using a multi-compartment pharmacokinetic model, and input the in vivo drug effect chamber concentration into the autonomic nervous system cybernetics model to generate an ideal sedation state reference vector that dynamically changes with the drug metabolism time axis.
[0007] The active simulation unit for noxious stimuli is used to extract the surgical operation markers from the synchronous panoramic data stream, and based on the surgical action and pain intensity mapping library, convert the surgical action into a noxious input function and inject it into the neural control model to generate a theoretical stimulated state vector.
[0008] The spatiotemporal feature dual difference verification unit is used to receive the multimodal vital signs as real monitoring values, and perform dual difference calculations by combining the ideal sedation state reference vector and the theoretical stress state vector. Based on the morphological similarity between the real difference vector and the theoretical difference vector generated by the difference results, a true stress judgment label or an artifact judgment label is generated, and the purification stress intensity is calculated.
[0009] A mixed reality holographic mapping unit is used to respond to the true stress determination label or the artifact determination label, and drive the mixed reality display device to perform differentiated rendering based on the purification stress intensity.
[0010] Preferably, the multimodal vital signs collected by the multi-source data sensing unit include heart rate, invasive blood pressure, and raw electroencephalogram (EEG) signals;
[0011] The perioperative medication record includes the drug name, dosage, and rate of administration;
[0012] Surgical procedure markers include incision event markers or suture event markers obtained through surgical logs or visual recognition.
[0013] Preferably, the pharmacokinetic manifold reconstruction unit generates an ideal sedation state reference vector, specifically including:
[0014] Based on perioperative medication records and individual patient parameters, the in vivo drug effect compartment concentration was calculated using the multi-compartment pharmacokinetic model.
[0015] Using the autonomic nervous system cybernetics model, based on the in vivo drug effect room concentration, the ideal physiological trajectory of the patient over time is deduced under the assumption of absolute quiet and no pain input, and the ideal physiological trajectory is determined as the reference vector of the ideal sedation state.
[0016] Preferably, the active simulation unit for noxious stimuli generates a theoretical stimulated state vector, specifically including:
[0017] Based on the surgical action and pain intensity mapping library, the surgical operation markers are converted into the nociceptive input function, which includes stimulus intensity, duration, and nerve conduction delay.
[0018] The nociceptive input function is injected as a perturbation parameter into the neural control model to simulate the sympathetic nerve excitation process, generating the theoretical excited state vector representing the theoretical stress response waveform.
[0019] Preferably, the spatiotemporal feature dual-difference verification unit performs dual-difference calculation and generates judgment labels, specifically including:
[0020] The difference between the actual monitored value and the ideal sedation state reference vector is calculated to generate the actual difference vector that mixes the actual physiological stress signal and environmental noise;
[0021] The difference between the theoretical excited state vector and the ideal sedated state reference vector is calculated to generate the theoretical difference vector with physiological characteristics;
[0022] The actual difference vector and the theoretical difference vector are projected and compared in a multi-dimensional feature space.
[0023] If the feature direction of the real difference vector is orthogonal to the theoretical difference vector, then the artifact determination label is generated;
[0024] If the feature direction of the real difference vector is not orthogonal to the theoretical difference vector but has a similar shape, then the true stress determination label is generated.
[0025] Preferably, the spatiotemporal characteristic dual-difference verification unit calculates the purification stress intensity, specifically including:
[0026] After generating the true stress determination label, the real difference vector is projected onto the feature direction where the theoretical difference vector is located;
[0027] Calculate the magnitude of the projected components;
[0028] The magnitude of the projection component is determined as the purification stress intensity after removing orthogonal noise components.
[0029] Preferably, the mixed reality holographic mapping unit performs differentiated rendering, specifically including:
[0030] When the true stress determination label is received, a bright pulsating holographic cloud map is superimposed on the corresponding anatomical site of the patient, and the size and pulsation frequency of the pulsating holographic cloud map are driven according to the intensity of the purification stress.
[0031] When the artifact detection tag is received, the alarm output is suppressed, and a low-salience gray ripple is rendered at the edge of the field of view.
[0032] Compared with the prior art, the present invention has the following beneficial effects:
[0033] 1. This system constructs an ideal sedation state baseline vector that changes in real time with the drug concentration in the body through a pharmacokinetic manifold reconstruction unit. This mechanism breaks through the limitation of traditional monitoring using static average values as the baseline, effectively overcomes the baseline lag and bias caused by ignoring changes in drug metabolism, and ensures that the judgment of stress response is always based on the patient's current real pharmacodynamic state, realizing dynamic adaptive adjustment of physiological baseline.
[0034] 2. This system adopts a dual differential verification mechanism of active simulation of noxious stimuli and spatiotemporal characteristics, and introduces signal processing logic based on causal mechanism. By comparing the consistency of the actual monitoring deviation and the theoretical prediction deviation in morphology, the system can accurately distinguish between real physiological stress and environmental artifacts such as electrosurgical interference, thereby greatly reducing the false positive rate and solving the fundamental problem that existing technologies cannot effectively identify the authenticity of signals.
[0035] 3. This system uses a vector projection algorithm to calculate the purification stress intensity, achieving quantitative purification of physiological signals. By retaining components consistent with theoretical characteristics and removing orthogonal noise components, the system can extract amplitude indicators that represent only effective physiological responses from strong background noise. This quantitative data provides doctors with a highly reliable reference after removing interference, which helps guide precise drug administration and anesthesia depth control during the perioperative period.
[0036] 4. This system applies mixed reality holographic mapping technology to transform complex calculation results into intuitive visual metaphors. By overlaying differentiated pulsating holographic cloud maps or edge ripples on the patient's anatomical sites, the system achieves a WYSIWYG monitoring experience, allowing users to grasp the patient's condition without having to interpret abstract data. This significantly reduces the cognitive load on anesthesiologists and improves surgical safety. Attached Figure Description
[0037] The present invention will be further explained below with reference to the accompanying drawings and embodiments:
[0038] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation
[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.
[0040] Example 1:
[0041] Please see Figure 1 A mixed reality-based visualization digital management system for anesthesia stress includes:
[0042] The multi-source data sensing unit is used to collect multimodal vital signs, perioperative medication records, and surgical operation markers in real time, and to perform time-series synchronization processing on the collected data to generate a synchronous panoramic data stream.
[0043] The pharmacokinetic manifold reconstruction unit is used to receive synchronous panoramic data streams, calculate the in vivo drug effect chamber concentration in real time using a multi-compartment pharmacokinetic model, and input the in vivo drug effect chamber concentration into the autonomic nervous system cybernetics model to generate an ideal sedation state reference vector that dynamically changes with the drug metabolism time axis.
[0044] The active simulation unit for noxious stimuli is used to extract surgical operation markers from the synchronous panoramic data stream, and based on the surgical action and pain intensity mapping library, it transforms the surgical action into a noxious input function and injects it into the neural control model to generate a theoretical stimulated state vector.
[0045] The spatiotemporal feature dual difference verification unit is used to receive multimodal vital signs as real monitoring values, and perform dual difference calculations by combining the ideal sedation state reference vector and the theoretical stress state vector. Based on the morphological similarity between the real difference vector and the theoretical difference vector generated by the difference results, a true stress judgment label or an artifact judgment label is generated, and the purification stress intensity is calculated.
[0046] The mixed reality holographic mapping unit is used to respond to real stress detection tags or artifact detection tags and drive the mixed reality display device to perform differentiated rendering based on the intensity of the purification stress.
[0047] In the system architecture of this embodiment, the multi-source data sensing unit serves as the sensing outpost of the entire system. It is configured to capture raw signals from distributed medical devices and information systems. This unit is not only responsible for data acquisition, but its core task is to perform strict temporal alignment to eliminate timestamp errors between different devices, thereby outputting a unified and synchronized panoramic data stream. This data stream is distributed to two parallel inference engines: the pharmacokinetic manifold reconstruction unit is dedicated to constructing a dynamically changing physiological zero point, that is, the calm state that the patient should exhibit at the current drug concentration; while the nociceptive stimulus active simulation unit is dedicated to constructing a theoretical physiological full scale, that is, assuming that the current surgical action does indeed cause pain, the stress response that the patient should exhibit.
[0048] The spatiotemporal feature dual-difference verification unit, as the decision-making core of the system, introduces a signal processing logic based on causal mechanisms. It no longer relies solely on the magnitude of monitored values to trigger alarms, but rather accurately identifies signal attributes by comparing the consistency of the actual deviation with the theoretically predicted deviation in terms of morphology. The mixed reality holographic mapping unit transforms these abstract calculation results into intuitive visual metaphors, projecting them directly into the doctor's field of vision. In this embodiment, the synchronous panoramic data stream refers to a multi-dimensional data set that has undergone time axis calibration and format standardization. Its function is to provide raw materials without time delay errors for subsequent mechanism deduction. Its source is the aggregation processing of data from monitors, anesthesia machines, and surgical logs.
[0049] By constructing the closed-loop monitoring architecture based on mechanism synthesis analysis, this invention effectively solves the fundamental problem in existing anesthesia monitoring technologies that cannot distinguish between real physiological stress and environmental equipment artifacts. In particular, it provides a management solution with causal explanation capabilities and strong anti-interference capabilities to address the problem of false alarms caused by non-harmful stimuli such as electrosurgical interference.
[0050] Example 2:
[0051] The multimodal vital signs collected by the multi-source data sensing unit include heart rate, invasive blood pressure, and raw electroencephalogram (EEG) signals;
[0052] Perioperative medication records include drug name, dosage, and rate of administration;
[0053] Surgical procedure markers include incision event markers or suture event markers obtained through surgical logs or visual recognition.
[0054] In the specific configuration of this embodiment, the multi-source data sensing unit is designed as a high-throughput data interface. It does not simply receive data, but rather collects key indicators that can reflect the patient's depth of anesthesia. The unit locks the patient's heart rate and invasive blood pressure in real time to monitor fluctuations in the circulatory system, and simultaneously captures raw brain electroencephalogram (EEG) signals to assess the degree of central nervous system inhibition. These three constitute multimodal vital signs. The unit is interconnected with the electronic anesthesia record sheet or infusion pump workstation to read perioperative medication records, including drug name, dosage, and infusion rate, in real time. This provides boundary conditions for subsequent calculation of the cumulative effect of the drug.
[0055] To enable the system to perceive external stimuli, this unit also integrates the function of recognizing surgical operation markers. By parsing log nodes in the surgical anesthesia information system or by connecting to the visual recognition results of the operating room camera, it can accurately capture the time of occurrence of key surgical events such as skin incision and suturing. In this embodiment, surgical operation markers refer to timestamps and type identifiers that can clearly indicate specific physical actions in the surgical process. Their function is to trigger the system's expected simulation of pain stimuli. Their source is usually structured surgical nursing records or intelligent visual analysis modules.
[0056] By comprehensively collecting and accurately identifying key data dimensions, this invention ensures that the monitoring system not only understands the patient's outcomes but also the causes that led to these outcomes, thus laying a solid data foundation for the subsequent leap from simple data statistics to in-depth causal inference.
[0057] Example 3:
[0058] The pharmacokinetic manifold reconstruction unit generates the ideal sedation state baseline vector, specifically including:
[0059] Based on perioperative medication records and individual patient parameters, the in vivo drug effect compartment concentration was calculated using a multi-compartment pharmacokinetic model.
[0060] Using a cybernetics model of the autonomic nervous system, based on the concentration of drugs in the effect room, we reverse-engineer the ideal physiological trajectory of the patient over time under the assumption of absolute quiet and no pain input, and define the ideal physiological trajectory as the reference vector of the ideal sedation state.
[0061] The specific steps of reverse engineering are as follows: First, train a Gaussian process regression model based on a pre-built historical patient database to establish a priori correspondence between drug concentration and manifold coordinates. Second, use Gaussian process regression (GPR) to establish a mapping function between low-dimensional manifold coordinates and physiological parameters. ,in Using manifold coordinates; representing the current in vivo drug effect-site concentration. Mapped to the corresponding coordinates in the manifold space and input the function The corresponding output value is calculated, which is the ideal physiological parameter value under the condition of no external stimuli.
[0062] In this embodiment, the pharmacokinetic manifold reconstruction unit aims to overcome the limitations of traditional monitoring that uses static average values as a baseline. Based on the input dosing records and individual parameters such as the patient's weight and age, the unit drives a built-in multi-compartment pharmacokinetic / pharmacodynamic model to perform calculations. This calculation process not only simulates drug distribution and metabolism but also utilizes topological data analysis technology, specifically employing an isometric mapping manifold learning algorithm. The nearest neighbor number K=12 is set to preserve the geodesic distance structure, mapping the patient's high-dimensional physiological parameters to a low-dimensional phase space, constructing a low-dimensional manifold structure describing the coupling relationship between drug concentration and multimodal physiological parameters. The unit uses the in vivo drug effect compartment concentration as a localization parameter on the manifold. This cybernetics model is constructed as a system of second-order differential equations describing the balance between the sympathetic and parasympathetic nervous systems. Specifically, the system of second-order differential equations uses an improved oscillator model to describe the autonomic nervous system regulation process, and its mathematical expression is:
[0063]
[0064] in, Represents normalized physiological parameter states, such as heart rate variability. This is the nonlinear damping coefficient, with a preset value of 0.5. The natural frequency of the system, used in the equilibrium equations, has a value of 1.0 rad / s. For real-time calculation of in vivo drug effect-site concentration, This is a non-zero local minimum constant, such as 0.01, used to prevent calculation singularities when the drug concentration is zero. This is the drug inhibition gain coefficient, with a value ranging from 1.2 to 1.5. The nonlinearity of drug efficacy is expressed as a value of 1.0; this equation describes the effect of drug concentration... Under the inhibitory effect, the physiological system maintains a homeostatic dynamic trajectory; for example, the simplified state-space equation based on the cardiovascular regulation model, in which drug concentration is coupled as the inhibition coefficient to the gain term of the equation and input into the manifold geometry-based autonomic nervous system cybernetics model;
[0065] This cybernetic model calculates geodesic distances on a manifold surface to simulate the neural regulatory pathways under conditions of no external harmful stimuli and only the inhibitory effect of the current anesthetic drugs. It then extrapolates and reconstructs the set of physiological parameters such as heart rate and blood pressure that the patient should exhibit at this moment. This set of values that evolves continuously over time constitutes the ideal physiological trajectory. In this embodiment, the ideal sedation state reference vector refers to a trajectory that dynamically extends over time in a high-dimensional space. It represents the theoretical physiological performance of the patient under perfect anesthesia and without any pain. Its function is to serve as a dynamically floating reference coordinate system to measure the actual distance of the actual monitoring values from the ideal state.
[0066] Through the above-mentioned dynamic reconstruction mechanism, this invention establishes a physiological benchmark that adapts in real time to drug metabolism, effectively overcoming the benchmark lag and bias problems caused by ignoring changes in drug concentration in traditional technologies, and ensuring that the judgment of stress response is always based on the patient's current real drug efficacy status.
[0067] Example 4:
[0068] The active simulation unit for noxious stimuli generates theoretical stimulated state vectors, specifically including:
[0069] Based on the mapping library between surgical actions and pain intensity, and combined with the preset individual patient pain sensitivity coefficient and local anesthesia blockade range attenuation factor, the surgical operation mark is transformed into a nociceptive input function that includes personalized stimulus intensity, duration and nerve conduction delay.
[0070] The nociceptive input function is injected as a perturbation parameter into the neural control model to simulate the sympathetic nerve excitation process, generating a theoretical excited state vector representing the theoretical stress response waveform.
[0071] The mapping library contains a standardized action-parameter lookup table. The specific data structure includes: skin incision event: corresponding stimulus intensity amplitude. 0-10 scale, duration Delayed nerve conduction Subcutaneous dissection: corresponding to the amplitude of stimulation intensity Duration Delayed nerve conduction Suturing event: corresponding stimulus intensity amplitude Duration Delayed nerve conduction ;
[0072] In this embodiment, the active simulation unit for noxious stimuli endows the system with predictive capabilities. When a specific surgical operation marker is received, the unit immediately queries the built-in surgical action and pain intensity mapping library, which stores data on the pain levels typically caused by different surgical actions. Based on the query results, the unit constructs a standardized noxious input function, which defines in detail the intensity amplitude, duration, and expected delay of the signal transmission from the peripheral nerves to the central nervous system.
[0073] This unit injects this function as a well-defined perturbation signal into the same neural control model as the aforementioned unit; specifically, it uses the aforementioned nociceptive input function as an additive perturbation term. Superimposed on the state variable representing the frequency of sympathetic nerve firing in the model, the model then simulates the excitation cascade response of the sympathetic nervous system after being stimulated in this way, and calculates the theoretical curves of heart rate increase and blood pressure fluctuation; this set of waveforms derived from the mechanism is defined as the theoretical stimulated state vector; in this embodiment, the theoretical stimulated state vector refers to the standard stress response waveform synthesized based on medical prior knowledge, representing a standard stress response purely caused by a specific surgical procedure. Its function is to serve as a standard answer template for comparison with the chaotic signals collected in reality to verify the authenticity of the signal;
[0074] Through this active simulation technique, the present invention transforms abstract surgical procedures into calculable physiological prediction models, enabling the system to proactively anticipate stress response patterns rather than passively waiting for alarms. This provides crucial morphological evidence for distinguishing real physiological responses from environmental noise.
[0075] Example 5:
[0076] The spatiotemporal feature dual-difference verification unit performs dual-difference calculation and generates judgment labels, specifically including:
[0077] The difference between the actual monitoring value and the baseline vector of the ideal sedation state is calculated to generate a reality difference vector that mixes real physiological stress signals and environmental noise.
[0078] The difference between the theoretical excited state vector and the ideal sedated state baseline vector is calculated to generate a theoretical difference vector with physiological characteristics.
[0079] The actual difference vector and the theoretical difference vector are projected and compared in a multi-dimensional feature space.
[0080] If the feature direction of the real difference vector is orthogonal to the theoretical difference vector, then an artifact determination label is generated;
[0081] If the feature direction of the real difference vector is not orthogonal to the theoretical difference vector but has a similar shape, then a true stress determination label is generated.
[0082] The multidimensional feature space is constructed by wavelet packet decomposition of the difference vectors, and the system sampling rate is set to... For example, for 1000Hz, the wavelet basis db4 is selected. The specific extraction is based on the corresponding decomposition layer, such as the low-frequency coefficient vector under the 10th layer. The frequency < 0.1Hz represents the hemodynamic trend. The high-frequency coefficient vector, with a frequency > 50Hz, represents electrophysiological noise. Under this feature space definition, since high-frequency noise such as electrosurgery interference is mainly distributed in the high-frequency subspace, while physiological stress response is distributed in the low-frequency subspace, the feature vectors of the two exhibit mathematical orthogonality in spatial geometry, and the dot product approaches zero.
[0083] If the feature direction of the actual difference vector is not orthogonal to the theoretical difference vector, but the dissimilarity in morphological features exceeds the preset threshold, a non-surgical source abnormality alarm will be triggered, prompting doctors to pay attention to physiological crises caused by non-surgical procedures such as allergic reactions or malignant hyperthermia.
[0084] In this embodiment, the spatiotemporal feature dual-difference verification unit acts as the system's signal discriminator. This unit performs the first difference, which is to subtract the dynamic ideal sedation state reference vector from the actual monitoring value, thereby extracting the patient's current physiological deviation and generating a reality difference vector. This vector not only contains possible real pain responses but also includes environmental noise such as electrocautery interference and changes in body position. This unit performs the second difference, which calculates the difference between the theoretical stress state vector and the ideal sedation state reference vector, generating a theoretical difference vector, which represents a pure, noise-free theoretical stress feature fingerprint.
[0085] The unit projects and compares the two vectors in a multidimensional feature space. If the analysis finds that the feature direction of the actual difference vector is orthogonal to the theoretical difference vector, the system will decisively determine that it is an artifact and generate an artifact determination label. If the two directions are not orthogonal and their shapes are highly similar, the system will confirm that the monitored fluctuations conform to physiological laws and generate a true stress determination label. In this embodiment, the term orthogonality in feature space comparison refers to the extremely low or even zero correlation between the two signal vectors. Its role is to serve as a mathematical criterion to eliminate non-bioelectric signals that, although the amplitude is large, do not conform to physiological laws at all.
[0086] By implementing the above-mentioned dual differential and orthogonal filtering mechanism, this invention achieves a noise reduction effect similar to a lock-in amplifier, which can accurately extract weak real physiological signals from extremely strong environmental background noise, and greatly reduce the false positive rate caused by interference from operating room equipment.
[0087] Example 6:
[0088] The spatiotemporal characteristic dual-difference verification unit calculates the purification stress intensity, specifically including:
[0089] After generating the true stress determination label, the real difference vector is projected onto the feature direction where the theoretical difference vector is located;
[0090] Calculate the magnitude of the projection component; determine the magnitude of the projection component as the purification stress intensity after removing orthogonal noise components.
[0091] In this embodiment, once the stress is confirmed to be real, the spatiotemporal feature dual difference verification unit further performs quantitative analysis to calculate the precise stress level; the unit uses a vector projection algorithm to map and project the real difference vector containing noise onto the theoretical difference vector representing the direction of the standard physiological response.
[0092] The physical significance of this operation is that it retains the components in the real signal that are consistent with the theoretical physiological characteristics, while discarding the noise components that are perpendicular to the theoretical characteristics. The unit calculates the length of this projected component and defines it as the purified stress intensity. In this embodiment, the purified stress intensity refers to the effective physiological response amplitude of the patient caused by the noxious stimulus after mathematical purification. Its function is to provide doctors with a highly reliable quantitative indicator after eliminating environmental interference, which can be used to guide precise drug administration.
[0093] Through this projection-based signal extraction technology, the present invention not only achieves qualitative identification of authenticity, but also quantitative noise removal, ensuring that the stress intensity value presented to the doctor can truly reflect the patient's pain level and avoid misjudgment of the condition due to superimposed noise.
[0094] Example 7:
[0095] The mixed reality holographic mapping unit performs differential rendering, specifically including:
[0096] When a true stress determination label is received, a bright pulsating holographic cloud map is superimposed on the corresponding anatomical site of the patient, and the size and pulsation frequency of the pulsating holographic cloud map are driven according to the intensity of the purification stress.
[0097] When an artifact detection tag is received, the alarm output is suppressed, and a low-salience gray ripple is rendered at the edge of the field of view.
[0098] In this embodiment, the mixed reality holographic mapping unit is responsible for converting complex calculation results into intuitive visual language; when the system determines that there is a true stress, the unit immediately activates the high saliency rendering mode, and displays a bright pulsating holographic cloud map superimposed on the corresponding anatomical position of the patient's body in the mixed reality glasses;
[0099] The cloud map is not a static image. Its volume expansion and contraction amplitude and pulsation frequency are directly driven in real time by the previously calculated purification stress intensity value, so that doctors can visually perceive the intensity of the stress. When the system determines that the detected fluctuation is an artifact, the unit will actively cut off the traditional sound alarm path to avoid noise interfering with the doctor's operation, and only render a low-salience, faint gray ripple in the edge area of the doctor's field of vision.
[0100] In this embodiment, the pulsating holographic cloud map is a data visualization metaphor. Its function is to transform abstract digital signals into biosignature simulations that conform to human intuition, so that users can instantly grasp the patient's condition without having to spend time interpreting the data.
[0101] Through this differentiated holographic rendering strategy, the present invention optimizes the information interaction experience in the operating room, not only achieving WYSIWYG monitoring with accurate data, but also significantly reducing the cognitive load on anesthesiologists, allowing them to focus more on the surgery itself, thereby improving the overall surgical safety.
[0102] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A visual digital management system for anesthesia stress based on mixed reality, characterized in that, include: The multi-source data sensing unit is used to collect multimodal vital signs, perioperative medication records, and surgical operation markers in real time, and to perform time-series synchronization processing on the collected data to generate a synchronous panoramic data stream. The pharmacokinetic manifold reconstruction unit is used to receive the synchronous panoramic data stream, calculate the in vivo drug effect chamber concentration in real time using a multi-compartment pharmacokinetic model, and input the in vivo drug effect chamber concentration into the autonomic nervous system cybernetics model to generate an ideal sedation state reference vector that dynamically changes with the drug metabolism time axis. The active simulation unit for noxious stimuli is used to extract the surgical operation markers from the synchronous panoramic data stream, and based on the surgical action and pain intensity mapping library, convert the surgical action into a noxious input function and inject it into the neural control model to generate a theoretical stimulated state vector. The spatiotemporal feature dual difference verification unit is used to receive the multimodal vital signs as real monitoring values, and perform dual difference calculations by combining the ideal sedation state reference vector and the theoretical stress state vector. Based on the morphological similarity between the real difference vector and the theoretical difference vector generated by the difference results, a true stress judgment label or an artifact judgment label is generated, and the purification stress intensity is calculated. A mixed reality holographic mapping unit is used to respond to the true stress determination label or the artifact determination label, and drive the mixed reality display device to perform differentiated rendering based on the purification stress intensity.
2. The anesthesia stress visualization digital management system based on mixed reality according to claim 1, characterized in that, The multimodal vital signs collected by the multi-source data sensing unit include heart rate, invasive blood pressure, and raw electroencephalogram (EEG) signals. The perioperative medication record includes the drug name, dosage, and rate of administration; Surgical procedure markers include incision event markers or suture event markers obtained through surgical logs or visual recognition.
3. The anesthesia stress visualization digital management system based on mixed reality according to claim 1, characterized in that, The pharmacokinetic manifold reconstruction unit generates an ideal sedation state baseline vector, specifically including: Based on perioperative medication records and individual patient parameters, the in vivo drug effect compartment concentration was calculated using the multi-compartment pharmacokinetic model. Using the autonomic nervous system cybernetics model, based on the in vivo drug effect room concentration, the ideal physiological trajectory of the patient over time is deduced under the assumption of absolute quiet and no pain input, and the ideal physiological trajectory is determined as the reference vector of the ideal sedation state.
4. The anesthesia stress visualization digital management system based on mixed reality according to claim 1, characterized in that, The active simulation unit for noxious stimuli generates a theoretical stimulated state vector, specifically including: Based on the surgical action and pain intensity mapping library, the surgical operation markers are converted into the nociceptive input function, which includes stimulus intensity, duration, and nerve conduction delay. The nociceptive input function is injected as a perturbation parameter into the neural control model to simulate the sympathetic nerve excitation process, generating the theoretical excited state vector representing the theoretical stress response waveform.
5. The anesthesia stress visualization digital management system based on mixed reality according to claim 1, characterized in that, The spatiotemporal feature dual-difference verification unit performs dual-difference calculation and generates judgment labels, specifically including: The difference between the actual monitored value and the ideal sedation state reference vector is calculated to generate the actual difference vector that mixes the actual physiological stress signal and environmental noise; The difference between the theoretical excited state vector and the ideal sedated state reference vector is calculated to generate the theoretical difference vector with physiological characteristics; The actual difference vector and the theoretical difference vector are projected and compared in a multi-dimensional feature space. If the feature direction of the real difference vector is orthogonal to the theoretical difference vector, then the artifact determination label is generated; If the feature direction of the real difference vector is not orthogonal to the theoretical difference vector but has a similar shape, then the true stress determination label is generated.
6. The anesthesia stress visualization digital management system based on mixed reality according to claim 5, characterized in that, The spatiotemporal feature dual-difference verification unit calculates the purification stress intensity, specifically including: After generating the true stress determination label, the real difference vector is projected onto the feature direction where the theoretical difference vector is located; Calculate the magnitude of the projected components; The magnitude of the projection component is determined as the purification stress intensity after removing orthogonal noise components.
7. The anesthesia stress visualization digital management system based on mixed reality according to claim 1, characterized in that, The mixed reality holographic mapping unit performs differentiated rendering, specifically including: When the true stress determination label is received, a bright pulsating holographic cloud map is superimposed on the corresponding anatomical site of the patient, and the size and pulsation frequency of the pulsating holographic cloud map are driven according to the intensity of the purification stress. When the artifact detection tag is received, the alarm output is suppressed, and a low-salience gray ripple is rendered at the edge of the field of view.