A multi-parameter anesthesia depth monitor device and method of use thereof

Abstract

This invention discloses a multi-parameter anesthesia depth monitor and its usage method, including a data acquisition module for simultaneously acquiring electroencephalogram (EEG) signals and body movement response signals; a signal processing module for preprocessing and extracting time-domain, frequency-domain, and nonlinear features; a multi-parameter fusion analysis engine, whose first analysis unit outputs a first anesthesia depth index based on EEG features, and a second analysis unit outputs a second response level quantifying the intensity of noxious stimuli based on body movement features; a data fusion unit fuses the two through a pre-trained machine learning model to generate a comprehensive multi-parameter anesthesia depth index; and a display and alarm module for real-time visualization and threshold alarms. The method includes signal acquisition, index calculation, fusion, and clinical guidance steps. By fusing multi-dimensional physiological signals, the limitations of single EEG monitoring are overcome, enabling a more comprehensive distinction between sedation and analgesia states, providing objective evidence for precise anesthesia management, and effectively reducing the risks of intraoperative awareness and over-anesthesia.

Description

Technical Field

[0001] This invention relates to the field of medical electronic monitoring equipment technology, specifically to a multi-parameter anesthesia depth monitor and its usage method. Background Technology

[0002] The core objective of general anesthesia is to create safe and stable conditions for surgical procedures while ensuring unconsciousness (sedation), painlessness (analgesia), and immobilization (muscle relaxation). Traditional monitoring of anesthetic depth relies primarily on the anesthesiologist's experience, combined with the patient's vital signs (such as blood pressure and heart rate) and limited clinical observations (such as body movement), which has the drawbacks of subjectivity and lag.

[0003] Currently, single-parameter anesthesia depth monitoring devices based on electroencephalography (EEG) analysis (such as BIS index monitors) are widely used in clinical practice. These devices primarily reflect cerebral cortical activity and are relatively accurate in monitoring the effects of sedative-hypnotic drugs (sedative components). However, anesthesia depth is a multi-dimensional concept, and single EEG indicators have significant limitations: 1) They are insensitive to or delayed responses related to analgesic components, such as body movement and hemodynamic fluctuations caused by noxious stimuli (e.g., surgical skin incision); 2) They are easily affected by interference from electrocautery; 3) They cannot effectively distinguish between oversedation and undersedation, potentially leading to intraoperative awareness or analgesic overdose.

[0004] Furthermore, intraoperative body movement is a direct and early manifestation of noxious stimuli, but objective and quantitative monitoring methods have traditionally been lacking. Although some studies have attempted to combine EEG and other parameters such as heart rate variability (HRV), a device that systematically and in real-time intelligently integrates EEG signals reflecting sedation with body movement signals (and other physiological signals) reflecting analgesia / noxious stimuli responses to generate a comprehensive, intuitive, and more instructive monitoring index, and forms a closed-loop clinical decision support system, remains a gap in the market.

[0005] Therefore, this solution proposes a multi-parameter anesthesia depth monitoring device and its usage method to solve the above problems. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the purpose of this invention is to provide a multi-parameter anesthesia depth monitoring device and its usage method.

[0007] To achieve the aforementioned objective, the technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a multi-parameter anesthesia depth monitoring device, characterized in that it comprises: The data acquisition module is configured to be operatively coupled to the patient for the simultaneous acquisition of raw physiological signals from at least two different physiological dimensions. The core raw physiological signals include at least electroencephalogram (EEG) signals reflecting cortical activity (typically acquired via frontal electrodes) and somatokinetic response signals reflecting reflexive movements in response to noxious stimuli (typically acquired via accelerometers or pressure sensors attached to the patient's limbs). This module ensures the temporal synchronization of multimodal signals, laying the foundation for subsequent fusion analysis.

[0008] The signal processing and feature extraction module is communicatively connected to the data acquisition module. This module is used to preprocess (e.g., amplify, filter), and reduce noise (e.g., remove power frequency interference, motion artifacts) of the acquired raw physiological signals, and extract robust feature parameters related to the depth of anesthesia. These feature parameters include, but are not limited to, time-domain features (e.g., signal amplitude statistics), frequency-domain features (e.g., EEG power spectrum, energy ratio of specific frequency bands), and nonlinear dynamic features (e.g., entropy, complexity).

[0009] The multi-parameter fusion analysis engine, which communicates with the signal processing and feature extraction module, is the core processing unit of this invention. It includes: The first analysis unit is specifically designed to process the characteristic parameters of EEG signals. Using mature algorithms (such as those based on power spectrum and dual-spectrum analysis) or trained models, it calculates and outputs a first anesthesia depth index (e.g., an index of 0-100, where a lower value indicates deeper sedation) that characterizes the depth of sedation and hypnosis.

[0010] The second analysis unit is specifically designed to process the feature parameters of the body movement response signal. This unit first filters the signal to distinguish between background spontaneous micromovements and body movements induced by noxious stimuli (such as surgical procedures); then, it extracts features such as the amplitude, duration, and burst pattern complexity of the induced body movements; finally, it maps these features to a second response level representing the intensity of the noxious stimulus response (e.g., divided into 1-5 levels, with higher levels indicating stronger responses and suggesting potentially insufficient analgesia).

[0011] Data fusion unit: This unit receives the first anesthesia depth index and the second response level. It embeds a pre-trained fusion algorithm model that learns the complex nonlinear relationship between two (or more) parameters and its correlation with the actual clinical anesthesia depth. Through this model, the input index and level are fused to generate a comprehensive, dimensionless, multi-parameter anesthesia depth index. This index comprehensively reflects the balance between sedation and analgesia.

[0012] The display and alarm module is communicatively connected to the multi-parameter fusion analysis engine. This module is used to display the first anesthesia depth index, the second reaction level, the multi-parameter anesthesia depth index, and their historical trend curves in real time and clearly. When the multi-parameter anesthesia depth index exceeds the preset safety threshold range (such as indicating "insufficient anesthesia" or "risk of intraoperative awakening"), or when a drastic change occurs within a short period of time, the module will trigger an audible and visual alarm to remind the anesthesiologist to intervene in a timely manner.

[0013] In a preferred embodiment, the functionality of the data acquisition module can be expanded to include a unit for acquiring at least one of the following signals: electrocardiogram (ECG) signals (for extracting heart rate variability (HRV), which is related to pain stress), pulse wave signals (for analyzing pulse wave morphology changes), electromyography (EMG) signals (reflecting muscle tension), skin conductance signals (reflecting sympathetic nerve excitability), or pupil diameter change signals (a potent indicator of opioid effects). Accordingly, the multi-parameter fusion analysis engine is further configured to integrate the feature parameters extracted from these additional signals into the fusion algorithm model, resulting in a more comprehensive final evaluation.

[0014] In a preferred embodiment, the pre-trained fusion algorithm model used by the data fusion unit is a machine learning-based model, such as a random forest, support vector machine, neural network, or gradient boosting decision tree. The model is obtained by training on a large-scale, multimodal physiological signal database labeled with clearly defined clinical anesthesia events (such as skin incision movement, drastic hemodynamic fluctuations, intraoperative awareness events, etc.), thereby enabling it to learn complex patterns from the data.

[0015] Further preferably, the device also includes a continuous learning and adaptive calibration module. This module is configured to anonymously record the patient's physiological signal data, various indices generated by the device, and clinical intervention events manually entered by the anesthesiologist (such as additional analgesics) during routine use of the device. Periodically, this anonymized data can be compared and analyzed with a cloud-based shared database or a local expert knowledge base. Based on the analysis results (such as the discovery of model prediction bias in a specific population), this module can fine-tune or calibrate the fusion algorithm model in the data fusion unit, thereby enabling the device to adapt to the physiological differences of different ethnicities, age groups, or individual patients, achieving personalized monitoring.

[0016] Regarding the processing of body movement signals, in one specific embodiment, the execution flow of the second analysis unit includes: performing bandpass filtering on the raw signals from the accelerometer or pressure sensor to effectively separate spontaneous micro-movements from noxious stimuli-induced body movements; extracting features such as amplitude, duration, and pattern complexity (e.g., approximate entropy) for the induced body movement signal segments; and using a classifier or regression model to map the features into a quantified response intensity level.

[0017] In a preferred embodiment, the display and alarm module employs a layered visualization design to improve information readability and clinical decision-making efficiency. The interface may include: a main view area highlighting the multi-parameter anesthesia depth index with dynamic curves and prominent numerical values; a sub-view area displaying the real-time values ​​and trend graphs of the first anesthesia depth index and the second response level side-by-side; and an event marker area allowing users to manually mark key anesthesia events (such as drug administration, intubation, and skin incision) during monitoring. These markers are displayed synchronously with the timeline of all physiological signals and indices, facilitating post-event review and analysis.

[0018] Secondly, the present invention provides a method for using the multi-parameter anesthesia depth monitoring device as described above, characterized by comprising the following steps: S1: Connect the device's data acquisition module correctly to the patient, including attaching EEG electrodes to the patient's forehead and fixing body movement response sensors to the patient's hands or feet to ensure good signal quality.

[0019] S2: Start the device and the system will automatically perform signal quality detection (such as impedance testing) and baseline calibration (establishing a baseline reference in a calm patient state).

[0020] S3: The device begins to collect physiological signals such as EEG and body movement in real time and synchronously. After processing, it calculates and outputs the first anesthesia depth index and the second reaction level in parallel.

[0021] S4: The multi-parameter fusion analysis engine receives the first and second parameters in real time, performs fusion calculations through the pre-trained model in the data fusion unit, and generates and continuously updates the comprehensive multi-parameter anesthesia depth index.

[0022] S5: All indices and trend graphs are presented together on the display interface. Anesthesiologists can use this to perform precise drug titration: for example, when the second response grade increases, it indicates a strong response to noxious stimuli, and the dosage of analgesics should be adjusted (increased) first; when the first depth of anesthesia index is abnormal (e.g., too high), it indicates insufficient sedation, and the dosage of sedative-hypnotic drugs should be adjusted first; while the overall trend of the multi-parameter depth of anesthesia index is used to assess the patient's overall anesthetic balance.

[0023] S6: When the monitoring system determines that the multi-parameter anesthesia depth index indicates that the anesthesia is too shallow or that there is a risk of intraoperative awakening, the device will immediately issue an early warning to prompt the physician to conduct an examination and intervention.

[0024] S7 (Optional Extended Step): The device records all raw data, processed indices, and user-tagged events throughout the anesthesia process. After anesthesia, a summary report containing key event points, a summary of changes in each index, and data-driven clinical decision recommendations can be automatically generated for medical record archiving and anesthesia quality improvement.

[0025] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0026] The beneficial effects of this invention are reflected in: This invention creatively integrates multiple parameters, such as electroencephalography (EEG) (sedation dimension) and somatokinetic responses (analgesia / nociceptive stimulus response dimension), in real time and intelligently through the combination of hardware integration and intelligent software algorithms. This provides a more comprehensive and reliable anesthesia depth monitoring tool than a single index. Its layered display and logically clear alarms effectively guide clinical decision-making, reduce the risk of intraoperative awareness and over-anesthesia, and improve anesthesia quality and patient safety. Attached Figure Description

[0027] In the attached diagram: Figure 1 This is a system block diagram of a multi-parameter anesthesia depth monitoring device according to an embodiment of the present invention.

[0028] Figure 2 for Figure 1 Internal workflow diagram of the multi-parameter fusion analysis engine.

[0029] Figure 3 for Figure 1 Display and alarm module interface - flowchart (UI structure and alarm logic).

[0030] Figure 4 This is a flowchart of a method for monitoring the depth of anesthesia using the device in one embodiment of the present invention. Detailed Implementation

[0031] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the invention, and not all of them. Unless otherwise specified, the embodiments and features described in this application can be combined with each other. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0032] Furthermore, "multiple" refers to two or more. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of a person skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the invention.

[0033] Example 1: Basic Equipment Composition and Work Process refer to Figure 1 The multi-parameter anesthesia depth monitoring device in this embodiment mainly includes: a data acquisition module, a signal processing and feature extraction module, a multi-parameter fusion analysis engine, a display and alarm module, and a power supply and management module.

[0034] The data acquisition module includes a 4-lead frontal EEG electrode assembly and a triaxial accelerometer. The electrode assembly is attached to the patient's forehead to acquire raw EEG signals. The accelerometer is fixed to the patient's wrist to acquire three-dimensional motion acceleration signals of the limbs. Both signals are transmitted synchronously to the host computer via wired or wireless means.

[0035] The signal processing and feature extraction module receives the raw signal. For EEG signals, a bandpass filter of 0.5-47Hz and a notch filter of 50Hz are applied to remove interference. Then, the power spectral density is calculated, the relative power of delta, theta, alpha, and beta waves is extracted, and relevant features in the Bispectral Index (BIS) algorithm are calculated, outputting feature vector F1. For acceleration signals, a bandpass filter of 0.1-10Hz is applied to focus on the body motion frequency. "Events" with signal amplitude exceeding a threshold are detected, and the peak amplitude, integral area, duration, and sample entropy of the event are calculated, outputting feature vector F2.

[0036] refer to Figure 2 The multi-parameter fusion analysis engine 30 includes a first analysis unit, a second analysis unit, and a data fusion unit. The first analysis unit receives a feature vector F1 and calculates a first anesthesia depth index I1 (0-100) using an empirical formula (or a simple linear model). The second analysis unit receives a feature vector F2 and maps the features to a second response level L2 (1-5) using a preset rule set (e.g., amplitude > X and duration > Y indicates a strong response). The data fusion unit 33 receives I1 and L2. In this embodiment, the fusion algorithm model is a pre-trained 3-layer fully connected neural network. This network takes I1 and L2 as input and outputs a multi-parameter anesthesia depth index I_composite (0-100). During training, the network uses labeled data from retrospective clinical video recordings, with expert-determined states of "appropriate anesthesia depth" or "too shallow / too deep."

[0037] The interface of the display and alarm module 40 is as follows: Figure 3 As shown. The main view area displays the real-time trend of I_composite with a thick curve and shows the current value. The sub-view area displays the value and trend curve of I1 on the left and the grade bar chart and simple trend of L2 on the right. The event marker area displays the timeline, and physicians can mark events "M" via the touchscreen. Alarm indicator lights and speakers activate when I_composite exceeds a preset threshold (e.g., >70).

[0038] Work process as follows Figure 4 As shown: Connect the device (S101), start and calibrate (S102), continuously acquire and process signals (S103), perform parallel calculations of I1 and L2 (S104), fuse and generate I_composite (S105), display and guide medication use (S106), and alarm when necessary (S107).

[0039] Example 2: Extended Devices and Adaptive Functions Based on Example 1, the data acquisition module further integrates a photoplethysmography (PPG) sensor to acquire fingertip pulse wave signals in order to extract features such as the low-frequency to high-frequency power ratio (LF / HF) in heart rate variability (HRV) as the third feature vector F3.

[0040] Accordingly, the input layer of the neural network model of the data fusion unit is expanded to receive I1, L2, and F3. HRV features corresponding to the time period are also added to the training data of the model.

[0041] In addition, the device incorporates a continuous learning and adaptive calibration module. After each anesthesia session, this module packages and stores anonymized I1, L2, F3, and I_composite sequences, along with physician-tagged dosing events (such as the timing of fentanyl administration). Periodically, the module compares the statistical characteristics of these data packets (such as exponential change patterns before and after specific events) with recognized "ideal response patterns" in a cloud-based expert database. If it is found that in elderly patients, the I_composite warning for insufficient analgesia is generally delayed beyond the actual dosing time, the module generates a calibration suggestion and uses locally stored data from this population to iteratively fine-tune the weights of the final layer of the neural network model, making the model more sensitive to warnings from this population.

[0042] Example 3: Clinical Application Method The anesthesiologist completes device connection before induction (S1). During induction, the I_composite in the main view area is observed to decrease rapidly, while I1 also decreases, and L2 remains low (S3-S5). During surgical incision (S6 trigger condition), although I1 may still be at a low level (indicating sufficient sedation), the L2 grade suddenly rises from 1 to 4, and I_composite also shows a spike. An alarm is displayed on the interface (S7). The physician immediately interprets this as: sufficient sedation depth, but insufficient analgesia, resulting in a strong somatokinetic response to the incision stimulus. Therefore, an intravenous supplemental opioid analgesia (such as remifentanil) is administered first (corresponding to the strategy of claim 8). After administration, the L2 grade is observed to rapidly decrease, and I_composite also returns to the target range. Throughout the process, I1 remains stable. After anesthesia, the system automatically generates a report, clearly showing the incision event, L2 surge, intervention, and index recovery process (S9).

[0043] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0044] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0045] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-parameter anaesthesia depth monitor device, characterized by, include: The data acquisition module is configured to be operatively coupled to the patient for simultaneously acquiring raw physiological signals of at least two different physiological dimensions, the raw physiological signals including at least electroencephalogram (EEG) signals and body movement response signals; The signal processing and feature extraction module is communicatively connected to the data acquisition module. It is used to preprocess and denoise the acquired raw physiological signals, and extract time-domain, frequency-domain, and nonlinear dynamic feature parameters related to the depth of anesthesia. A multi-parameter fusion analysis engine, communicatively connected to the signal processing and feature extraction module, includes: The first analysis unit is used to calculate and output the first anesthesia depth index based on the characteristic parameters of the electroencephalogram signal. The second analysis unit is used to identify and quantify the intensity of the noxious stimulus response based on the characteristic parameters of the body motion response signal, and output a second response level. The data fusion unit is used to receive the first anesthesia depth index and the second reaction level, and generate a comprehensive, dimensionless, multi-parameter anesthesia depth index through a pre-trained fusion algorithm model. The display and alarm module is connected to the multi-parameter fusion analysis engine and is used to display the first anesthesia depth index, the second reaction level, the multi-parameter anesthesia depth index and their changing trends in real time, and to trigger an alarm when the multi-parameter anesthesia depth index exceeds a preset safety threshold or undergoes a sudden change.

2. The multi-parameter anesthesia depth monitoring device according to claim 1, characterized in that, The data acquisition module also includes a unit for acquiring at least one of the following signals: electrocardiogram signal, pulse wave signal, electromyography signal, skin conductance signal, or pupil diameter change signal; the multi-parameter fusion analysis engine is further configured to integrate the feature parameters of the acquired additional signals into the fusion algorithm model.

3. A multi-parameter anaesthesia depth monitor device according to claim 1, characterized in that, The data fusion unit uses a pre-trained fusion algorithm model based on machine learning, selected from random forest, support vector machine, neural network or gradient boosting decision tree. The model is trained through a large-scale multimodal physiological signal database labeled with clinical anesthesia events.

4. A multi-parameter anaesthesia depth monitor device according to claim 3, wherein, The device also includes a continuous learning and adaptive calibration module, which is configured to: During device use, patients' physiological signal data, generated indices, and clinical intervention events input by anesthesiologists are anonymously recorded. Regularly compare and analyze anonymous data with cloud databases or local expert knowledge bases; Based on the analysis results, the fusion algorithm model in the data fusion unit is fine-tuned to adapt to different groups of people or individual differences.

5. A multi-parameter anaesthesia depth monitor device according to claim 1, characterized in that, The second analysis unit processes the body motion response signal as follows: Signals from accelerometers or pressure sensors are filtered to separate spontaneous micro-movements from noxious stimulus-induced body movements. Extract the amplitude, duration, and pattern complexity features of induced body movements; The features are mapped to a quantitative level representing the intensity of the response to a noxious stimulus.

6. A multi-parameter anaesthesia depth monitor apparatus as claimed in claim 1, wherein, The display and alarm module's interface adopts a layered visual design, including: Main view area: Highlights the multi-parameter anesthesia depth index in the form of dynamic curves and numerical values; Sub-view area: Displays the real-time values ​​and trend graphs of the first anesthesia depth index and the second reaction level side by side; Event Marking Area: Allows users to mark anesthesia events, which are then displayed synchronously with the physiological signal timeline.

7. A method for using a multi-parameter anesthesia depth monitoring device as described in any one of claims 1-6, characterized in that, Includes the following steps: S1: Connect the device’s data acquisition module to the patient correctly to ensure good signal transmission between the EEG electrodes and the body movement response sensor; S2: Start the device; the system will automatically perform signal quality detection and baseline calibration. S3: The device collects and processes physiological signals in real time and calculates the first anesthesia depth index and the second reaction level. S4: The multi-parameter fusion analysis engine fuses the first anesthesia depth index with the second reaction level to generate and update the multi-parameter anesthesia depth index in real time. S5: Displays all indices and trends on the screen to guide anesthesiologists in titrating sedative and analgesic drugs; S6: When the multi-parameter anesthesia depth index indicates that the anesthesia is too shallow or there is a risk of intraoperative awakening, the device will issue an early warning to prompt the physician to intervene.

8. A method of a multi-parameter anaesthesia depth monitor device according to claim 7, characterized in that, In step S5, the anesthesiologist prioritizes adjusting the analgesic dosage based on the increase in the second response level; prioritizes adjusting the sedative-hypnotic dosage based on the abnormality of the first anesthesia depth index; and assesses the anesthesia balance status based on the overall trend of the multi-parameter anesthesia depth index.

9. The method of claim 1, wherein, The method further includes step S7: the device records all data, indices and marked events throughout the anesthesia process, and automatically generates a summary report containing key event points, index changes and clinical decision recommendations after the anesthesia ends.

10. A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method as described in any one of claims 7-9.

Images