An electroencephalogram warning system for adverse events in acupuncture treatment of central nervous system

Abstract

The application belongs to the technical field of medical monitoring and early warning, and relates to an electroencephalogram early warning system for adverse events of central nervous system in acupuncture courses, which comprises: an instrument integration and time synchronization association module, which is used for synchronously generating an operation marker signal and a skin conductance signal; a reference state vector construction module and a safety stress vector construction module, which are respectively used for establishing an individualized physiological baseline vector and a safety stress limit vector; an instant state vector generation module, which is used for generating a real-time physiological state vector; a dynamic weight adjustment module, which is used for generating a correction vector; an analysis and risk judgment module, which is used for analyzing whether a state evolution trajectory meets a risk condition; a hierarchical early warning execution module and an intervention instruction triggering module, which are used for executing hierarchical early warning and closed-loop physical intervention; and the application solves the technical problem that in the existing acupuncture early warning method, the operation and physiological signal time sequence association is weak, and single-dimensional monitoring is difficult to comprehensively capture the multi-system collaborative risk evolution.

Description

Technical Field

[0001] This invention belongs to the technical field of medical monitoring and early warning, and relates to an electroencephalogram (EEG) early warning system for adverse central nervous system events during acupuncture treatment. Background Technology

[0002] As a widely used traditional physical therapy, the safety of acupuncture has always been a key concern in clinical practice. Central nervous system-related adverse events, such as "acupuncture fainting," occasionally occur during treatment, with patients experiencing sudden dizziness, palpitations, paleness, or even loss of consciousness. This not only interrupts the treatment process but can also cause physical and psychological discomfort for the patient. Currently, prevention of such events mainly relies on the clinical experience of the acupuncturist, judging by observing the patient's complexion and inquiring about their subjective feelings. This method has a certain degree of subjectivity and delay, and the reliability of early warning before adverse events occur needs to be improved.

[0003] To improve the objectivity and foresight of monitoring, the industry has begun exploring the use of bioelectrical signals such as electroencephalography (EEG), heart rate variability, and skin conductance to monitor patients' physiological states. These technologies can provide more quantitative physiological indicators than subjective observation. However, their application in the specific context of acupuncture presents unique challenges. Acupuncture, as an invasive procedure, directly triggers instantaneous and dramatic fluctuations in the patient's physiological signals. These physiological responses directly caused by the procedure may resemble pathological changes predicting adverse events in terms of signal morphology, making it easy for traditional monitoring methods to confuse the signals and potentially leading to a higher false alarm rate.

[0004] Existing physiological monitoring methods for acupuncture early warning still have several areas for improvement. First, most systems rely on post-hoc comparisons to analyze passively collected physiological signals against manually recorded events, lacking the real-time capability to capture the precise temporal correlation between the maneuver and physiological signals. This makes it difficult to effectively distinguish between manipulative artifacts and genuine physiological changes. Second, common warning thresholds often depend on population statistical models, failing to adequately consider individual differences in physiological baselines and stress response patterns, resulting in insufficient individualized accuracy of warnings. Furthermore, many systems only assess thresholds for single-dimensional physiological signals, while adverse events in the central nervous system are often the result of dysregulation across multiple physiological subsystems. Single-dimensional monitoring may struggle to comprehensively capture the complete evolution of the risk. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides an electroencephalogram (EEG) early warning system for adverse central nervous system events during acupuncture treatment.

[0006] An electroencephalogram (EEG) early warning system for adverse central nervous system events during acupuncture treatment includes:

[0007] The device integration and time synchronization module deploys an integrated device monitoring system to establish a time synchronization relationship between operational behavior and physiological signals, and synchronously generates operation marker signals that mark the time of the acupuncturist's operation and skin conduction signals that reflect the patient's autonomic nerve activity;

[0008] The baseline state vector construction module constructs a baseline state vector representing an individual's physiological baseline based on the electroencephalogram (EEG) and skin conductance signals acquired before treatment.

[0009] The safety stress vector construction module synchronously collects EEG signals, skin conductance signals, and operation marker signals through simulated operations, and extracts and constructs a safety stress vector that represents the individual's normal stress limit.

[0010] The real-time state vector generation module generates a real-time state vector representing the current physiological state based on real-time acquired EEG signals, skin conductance signals, and operation marker signals.

[0011] The dynamic weight adjustment module performs dynamic weight adjustment on each component of the instantaneous state vector based on the operation marker signal to generate the corrected state vector.

[0012] The analysis and risk assessment module, in a multi-dimensional state space, analyzes whether the time series trajectory formed by the corrected state vector meets the preset risk evolution conditions based on the physiological state safety region defined by the baseline state vector and the safety stress vector, and performs risk assessment.

[0013] The tiered early warning execution module, based on the risk assessment results, executes tiered early warnings when the state evolution trajectory initially deviates from or is confirmed to enter a risk path.

[0014] The intervention command triggering module triggers a closed-loop intervention command physically integrated with the early warning system when the state evolution trajectory breaks through the boundary of the physiological state safety zone defined by the safety stress vector, as well as a vibration device connected to the intervention command triggering module and used to generate a physical intervention signal in response to the intervention command.

[0015] A further aspect of the present invention includes an instrument integration and time synchronization association module, used to perform the following operations:

[0016] A needle tray is provided, the surface of which is divided into a waiting area and an operation area. The operation area is embedded with a piezoelectric thin film array for sensing needle picking and placing operations to generate operation mark signals.

[0017] The needle tray integrates conductive fabric electrodes for collecting patient skin conductivity signals to generate skin conductivity signals;

[0018] The system uses an internal clock to time-align the operation marker signal with the skin conductance signal, thereby establishing a time-synchronous correlation between the operation event and the physiological signal.

[0019] A further aspect of the present invention includes a baseline state vector construction module, which performs the following operations:

[0020] When the system is in baseline acquisition mode, it simultaneously acquires the patient's electroencephalogram (EEG) signals and skin conductance signals in a relaxed state.

[0021] The energy of the frontal lobe Theta wave and the energy of the occipital lobe Alpha wave were extracted from the electroencephalogram (EEG) signals, and the basic skin conductance level was extracted from the skin conductance signals.

[0022] The baseline state vector is generated by combining the Theta wave energy of the frontal lobe, the Alpha wave energy of the occipital lobe, and the baseline skin conductance level.

[0023] A further embodiment of the present invention includes a blunt-tipped probe, and a safety stress vector construction module is used to perform the following steps:

[0024] Based on the analog operation marker signal detected by the piezoelectric thin film array and triggered by the picking and placing action of the blunt probe;

[0025] Simultaneously acquire EEG signals, skin conductance signals, and operation marker signals during the simulated operation;

[0026] Within the time window marked by the operational label signal, the degree of alpha wave inhibition of the EEG signal, the peak response and recovery time of the skin conductance signal are extracted and quantified to construct a safety stress vector.

[0027] A further aspect of the present invention includes an instantaneous state vector generation module, used to perform the following steps:

[0028] During acupuncture treatment, real-time electroencephalogram (EEG) signals, skin conductance signals, and manipulation marker signals are continuously and synchronously collected.

[0029] Calculate the current values ​​of frontal lobe Theta wave energy and occipital lobe Alpha wave energy from real-time EEG signals;

[0030] Calculate the amplitude of the skin conductance response associated with the operation marker signal and the recovery half-life value after the response from the skin conductance signal;

[0031] The instantaneous state vector is generated by combining the current values ​​of the frontal lobe Theta wave energy, the current values ​​of the occipital lobe Alpha wave energy, the amplitude of the skin conductance response, and the recovery half-life value.

[0032] A further aspect of the present invention includes a dynamic weight adjustment module, which performs the following operations:

[0033] When the operation marker signal indicates the start of the operation, the EEG energy and skin conductance response components caused by the operation in the instantaneous state vector are assigned preset analysis weights.

[0034] When the operation marker signal indicates that the operation is over, the normal analysis weight of the recovery component is restored, and the trend of the recovery half-life value is analyzed.

[0035] Based on the weight coefficients corresponding to the current mode, calculate the weighted values ​​of each component of the instantaneous state vector to generate the corrected state vector.

[0036] A further aspect of this invention includes an analysis and risk assessment module, used to perform the following operations:

[0037] Establish a multidimensional state space with brain electrical energy, skin conductance response, and recovery half-life as coordinate axes;

[0038] The baseline state vector and the safety stress vector are jointly projected onto a multi-dimensional state space to define the physiological state safety region.

[0039] The modified state vector is mapped into a state evolution trajectory in the multidimensional state space;

[0040] The system continuously determines whether the state evolution trajectory meets the conditions of continuously deviating from the safe area, the evolution direction pointing to the preset risk area, and the evolution speed and magnitude exceeding the limits defined by the safety stress vector.

[0041] A further aspect of the present invention includes a tiered early warning execution module, used to perform the following operations:

[0042] When the state evolution trajectory initially shows a deviation from the trend, a level one warning is triggered, which is indicated by visual signals visible only to the acupuncturist.

[0043] When the state evolution trajectory confirms that it has entered a risk evolution path, a level two warning is triggered, and a combined prompt is made through visual signals and signals that can be perceived by the acupuncturist.

[0044] A further aspect of the present invention includes an intervention command triggering module, used to perform the following operations:

[0045] When the evolution trajectory of the determined state breaks through the safety boundary defined by the safety stress vector and reaches the preset emergency threshold, it is determined to be a risk state;

[0046] In response to a risk situation, an intervention command is sent to the vibration device integrated at the bottom of the waiting area of ​​the intelligent needle tray;

[0047] The vibration device responds to intervention commands and outputs mechanical vibration in a preset mode.

[0048] A further aspect of this invention involves determining whether a state evolution trajectory simultaneously satisfies risk conditions, including the following steps:

[0049] By checking whether the most recent consecutive points on the trajectory are all outside the physiological safety zone, it can be determined whether there is a continuous deviation.

[0050] The evolution direction is determined by calculating the angle between the main direction vector of the last segment of the trajectory and the direction vector pointing to the center of the preset risk area.

[0051] Calculate the state evolution rate and deviation magnitude, then calculate the comprehensive impact index, and compare the comprehensive impact index with the individualized threshold derived from the normalized safety stress vector. If the comprehensive impact index is greater than the individualized threshold, it is determined to be out of limit.

[0052] In summary, the present invention has the following beneficial technical effects:

[0053] 1. By integrating a piezoelectric thin film array with conductive fabric electrodes into an intelligent needle placement plate and dynamically adjusting the weight of physiological signal components based on operation marker signals, this system can distinguish between transient physiological fluctuations and potential pathological changes directly caused by acupuncture operations, which helps reduce the interference of operation artifacts on early warning judgment.

[0054] 2. By constructing individualized baseline state vectors and safety stress vectors, and defining individualized physiological state safety regions in a multidimensional state space, the fixed thresholds based on population statistics are replaced, enabling risk assessment to adapt to the physiological differences of different patients and improving the pertinence of early warning.

[0055] 3. By integrating EEG energy characteristics with skin conductance response and recovery dynamics characteristics, a multidimensional state vector is constructed and its spatiotemporal evolution trajectory is analyzed. This can simultaneously reflect the state of the central nervous system and the autonomic nervous system, thus providing a more comprehensive information basis for identifying multi-system dysregulation.

[0056] 4. By connecting risk assessment with vibration intervention devices integrated into the equipment, a closed-loop system from condition monitoring and risk warning to physical intervention is formed, providing a technical basis for achieving non-invasive early proactive intervention. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings are used to provide a further understanding of the present invention.

[0058] Figure 1 This discloses a schematic diagram of the framework in the embodiments of this application.

[0059] Figure 2 This discloses a flowchart of an embodiment of this application. Detailed Implementation

[0060] The following is in conjunction with the appendix Figure 1 - Figure 2 A preferred description of the present invention is provided below.

[0061] See attached document Figure 1 - Figure 2 This invention proposes an electroencephalogram (EEG) early warning system for adverse central nervous system events during acupuncture treatment, comprising the following modules:

[0062] The system comprises several modules: an instrument integration and time synchronization module, a device integration monitoring system, and a dynamic weight adjustment module. The instrument integration module deploys an integrated monitoring system to establish a time-synchronized correlation between operational behavior and physiological signals, synchronously generating operation marker signals to indicate the time of the acupuncturist's operation and skin conductance signals reflecting the patient's autonomic nervous activity. A baseline state vector construction module constructs a baseline state vector representing the individual's physiological baseline based on pre-treatment EEG and skin conductance signals. A safety stress vector construction module extracts and constructs a safety stress vector representing the individual's normal stress limit by synchronously acquiring EEG, skin conductance, and operation marker signals during simulated operation. An instantaneous state vector generation module generates an instantaneous state vector representing the current physiological state based on real-time acquired EEG, skin conductance, and operation marker signals. The operation marker signal performs dynamic weight adjustment on each component of the instantaneous state vector to generate a corrected state vector; the analysis and risk assessment module, in the multi-dimensional state space, analyzes whether the time series trajectory formed by the corrected state vector meets the preset risk evolution conditions based on the physiological state safety region defined by the baseline state vector and the safety stress vector, and performs risk assessment; the graded early warning execution module, based on the risk assessment results, executes graded early warning when the state evolution trajectory initially deviates or is confirmed to enter a risk path; the intervention command triggering module, when the state evolution trajectory breaks through the boundary of the physiological state safety region defined by the safety stress vector, triggers a closed-loop intervention command physically integrated with the early warning system, as well as a vibration device connected to the intervention command triggering module and used to generate physical intervention signals in response to the intervention command.

[0063] In one embodiment of the present invention, the device integration and time synchronization association module is used to perform the following steps:

[0064] A needle placement tray is provided, the surface of which is divided into a waiting area and an operation area. The operation area is embedded with a piezoelectric thin film array for sensing needle picking and placing operations to generate operation marker signals. The needle placement tray is integrated with conductive fabric electrodes for collecting patient skin conductivity signals to generate skin conductivity signals. The operation marker signals and skin conductivity signals are time-aligned by the system's internal clock to establish a time synchronization correlation between operation events and physiological signals.

[0065] Specifically, an intelligent needle placement tray is provided. The tray surface is physically divided or color-coded into two functional areas: a waiting area for neatly placing unused sterile acupuncture needles, and an operating area for temporarily placing needles being used or already used by the practitioner. A piezoelectric thin film array, composed of multiple miniature piezoelectric sensing units arranged in a matrix, is tightly embedded beneath the substrate of the operating area. It should be noted that this intelligent needle placement tray is a physical tray or pad for holding and managing acupuncture needles, serving as both a sensor carrier and an operating platform. The waiting area is a specially reserved physical area on the tray for the orderly storage of unused needles, ensuring separation between needle management and the operating area. The operating area is the physical area on the tray adjacent to the practitioner for needle retrieval actions, defining the sensing range for these actions.

[0066] In one embodiment of the present invention, the piezoelectric sensing unit of the piezoelectric thin film array is made of polyvinylidene fluoride material with a piezoelectric constant of -33 pC / N. The array unit spacing is 5 mm × 5 mm, and the area of ​​the operating area covered is not less than 10 cm × 10 cm, so as to ensure that the minute pressure changes generated during needle placement and removal can be captured.

[0067] When the acupuncturist removes a needle from the operating area or places it back, the pressure change generated by the needle on the disc surface is sensed by the piezoelectric thin film array. The array converts the mechanical pressure into an electrical signal, which is then processed by the signal conditioning circuit to generate a discrete pulse signal with a precise timestamp. This signal is defined as the operation marker signal, with the rising edge marking the start of the operation and the falling edge marking the end of the operation. The signal conditioning circuit includes a charge amplifier, a bandpass filter, and a comparator. When the pressure signal exceeds a preset threshold, a pulse output is triggered to ensure the accuracy and anti-interference capability of the operation marker signal. Simultaneously, an electrode pad made of soft conductive fabric, i.e., a conductive fabric electrode, is integrated into the edge of the intelligent needle placement disc or a preset support area on the patient's wrist. This is a flexible, wearable electrophysiological sensor made using conductive fiber weaving or coating technology. For example, the conductive fabric electrode uses a silver fiber blend material with a surface impedance ≤1 kΩ / cm² and a size of 2 cm × 3 cm. It is attached to the flexor side of the patient's wrist with medical adhesive to ensure stable contact with the skin.

[0068] It should be noted that the piezoelectric film array is configured based on the need to reliably convert needle pick-up and drop pressures as light as 100 mN into measurable electrical signals. The operation marker signal is a discrete, digitized pulse signal output from the piezoelectric film array; its data structure may include a timestamp, operation type (pick-up / drop-off), and signal strength. The conductive fabric electrode is configured based on clinical requirements to ensure long-term stable contact with the patient's skin and to obtain effective skin conductivity signals. Its contact impedance typically needs to be stable below 10 kΩ. In actual use, measured at a constant voltage of 0.5 V, the skin contact impedance should be maintained between 5-15 kΩ. If it exceeds 20 kΩ, it indicates poor contact and triggers a system self-test. The skin conductivity signal is a continuously varying analog voltage signal output from the conductive fabric electrode; it is a physiological data stream reflecting the dynamic changes in skin conductivity levels.

[0069] During treatment, the patient's wrist or forearm skin must maintain stable contact with the electrode pad. The conductive fabric electrode forms a circuit with the human skin through a constant voltage or constant current circuit, continuously measuring the conductivity between two points on the skin and outputting a continuous, time-varying analog voltage signal. This signal is defined as the skin conductivity signal, and its amplitude changes reflect the activity of sweat gland secretion, which in turn relates to the sympathetic nerve excitation level of the autonomic nervous system. Through a high-precision clock within the system, the timestamp of the operation marker signal generated by the piezoelectric film array is strictly aligned with the data stream of the skin conductivity signal collected by the conductive fabric electrode, thereby establishing a precise time synchronization correlation between the specific operation event of the acupuncturist and the corresponding physiological signal period of the patient at the data level.

[0070] For example, the data used to verify the sensing effectiveness of the piezoelectric thin film array is as follows: using a piezoelectric thin film sensing unit based on PVDF material with a piezoelectric constant of -33 pC / N, when a standard acupuncture needle is picked up or placed, the pressure change acting on the sensing unit is about 5 mN. After being amplified by a charge amplifier, the peak-to-peak value of the output voltage signal can reach 150 mV, and the signal-to-noise ratio is greater than 20 dB. This data directly verifies that the piezoelectric thin film array can generate a clear and distinguishable operation mark signal that can be used to accurately mark the moment of needle picking up or placing.

[0071] The data verifying the effectiveness of the conductive fabric electrode for acquisition are as follows: A conductive fabric electrode made of silver fiber blend was attached to the skin on the flexor side of the patient's wrist. A constant voltage of 0.5 V was used for measurement, and the measured baseline skin impedance was 8 kΩ. Furthermore, when the patient was startled by a slight sound, the skin conductance signal showed a transient increase of more than 0.5 μS within 2 seconds. This data directly verifies that the conductive fabric electrode can stably acquire skin conductance signals that reflect the patient's autonomic nerve activity.

[0072] In one embodiment of the present invention, the baseline state vector construction module is configured to perform the following steps:

[0073] When the system is in baseline acquisition mode, it simultaneously acquires EEG signals and skin conductance signals of the patient in a relaxed state; it extracts frontal lobe Theta wave energy and occipital lobe Alpha wave energy from the EEG signals, and extracts baseline skin conductance levels from the skin conductance signals; it combines the frontal lobe Theta wave energy, occipital lobe Alpha wave energy and baseline skin conductance levels to generate a baseline state vector.

[0074] Specifically, before the acupuncture treatment officially begins, the patient is instructed to relax and have specific body parts, such as the inner forearm or wrist, stably contact the conductive fabric electrodes integrated on the deployed intelligent needle placement tray. Two physiological signals are activated and simultaneously acquired: one is acquired using standard EEG acquisition equipment, following the international 10-20 system standard, by attaching electrodes to specific locations on the patient's scalp to collect EEG signals, such as frontal lobe leads Fp1, Fp2, F3, and F4, and occipital lobe leads O1 and O2; the other is acquired via conductive fabric electrodes, reflecting the baseline level of autonomic nervous activity in the patient's relaxed state. A high-precision clock within the system ensures strict alignment of the time sequences of these two signals. Data acquisition continues for a preset duration, such as 120 seconds, to ensure stable baseline data acquisition. This duration can be set according to clinical EEG baseline acquisition guidelines, covering multiple resting-state EEG fluctuation cycles, such as alpha rhythms, and ensuring that skin conductance levels reach a steady state. If the patient cannot remain still, the system supports segmented data acquisition, such as 30 seconds per segment, and automatically removes segments with obvious motion artifacts.

[0075] Subsequently, the acquired synchronization signals were processed: First, the EEG signals were preprocessed, including using bandpass filters from 0.5 Hz to 40 Hz to filter out low-frequency drift and high-frequency noise, and removing EEG artifacts using independent component analysis or regression methods. In the preprocessing, a first-order Butterworth filter could be used for the 0.5 Hz high-pass filter, and a fourth-order Butterworth filter could be used for the 40 Hz low-pass filter to balance the filtering effect with phase delay. EEG artifact removal was performed using extended Infomax independent component analysis (ICA) to remove components with a correlation >0.7 with the EEG channel.

[0076] Then, from the preprocessed EEG signals, the power spectrum of the frontal lobe leads (such as F3 and F4) was calculated using Fast Fourier Transform, and the total power in the frequency range of 4 Hz to 8 Hz was extracted and defined as the frontal lobe Theta wave energy value. Simultaneously, for the EEG data of the occipital lobe leads (such as O1 and O2), their power spectrum was also calculated, and the total power in the frequency range of 8 Hz to 13 Hz was extracted and defined as the occipital lobe alpha wave energy value. Next, the synchronously acquired skin conductance signals are processed. After filtering out high-frequency noise, the average value of the skin conductance level signal during the entire acquisition period is calculated and defined as the baseline skin conductance level value. Finally, the three scalar feature values ​​extracted from the patient in a relaxed state, namely the energy values ​​of the frontal lobe Theta wave, were analyzed. Occipital lobe Alpha wave energy value and baseline skin conductivity level These features are combined in a fixed order to form multidimensional feature points, which are defined as the baseline state vector representing the individualized physiological baseline of the patient in a state of relaxation without stimulation. Its mathematical expression can be regarded as a three-dimensional vector.

[0077] Calculate the bandwidth energy:

[0078]

[0079] in, Energy value representing a specific frequency band (Theta or Alpha), expressed in μV²; It is the power spectral density function obtained through fast Fourier transform, with units of μV² / Hz; and These are the start and cutoff frequencies of the frequency band, respectively, in Hz; This refers to frequency resolution, measured in Hz. The symbols indicate the following: It is the brainwave frequency band energy to be extracted; It is the frequency of brain signals The power spectral density at the point was obtained by FFT after applying the Hanning window to the preprocessed EEG signal segment; and According to the internationally recognized brainwave frequency band division, Theta waves are 4 Hz and 8 Hz, and Alpha waves are 8 Hz and 13 Hz. The number of FFT points and the sampling rate determine the result. For example, if the sampling rate is 250 Hz and the number of FFT points is 1024, then... .

[0080] Baseline state vector:

[0081]

[0082] in, Represents the baseline state vector; The energy of the frontal lobe Theta wave is calculated using Formula 1, and is set based on the fact that the energy of this frequency band is usually low and stable in a relaxed state. The energy of the occipital lobe Alpha wave is calculated using Formula 1. The basis for this setting is that the energy of this frequency band is usually higher in a relaxed state with eyes closed, which is a sign of relaxation. The baseline skin conductance level, measured in μS, is obtained by calculating the arithmetic mean of skin conductance signals over the baseline acquisition period. This mean is chosen because it represents the baseline sympathetic tone of an individual in a stress-free state. This vector combines static features of the central nervous system (EEG) and the autonomic nervous system (skin conductance) as a feature vector for subsequent spatial comparisons of relative positions.

[0083] Electroencephalography (EEG) signals are bioelectrical signals generated by the electrical activity of neurons recorded by electrodes placed on the scalp. They are multi-channel voltage sequence data that varies over time. Frontal lobe Theta wave energy refers to the power integral value of the 4 Hz to 8 Hz frequency components extracted from EEG signals recorded by scalp electrodes located in the frontal lobe region of the brain. It is a scalar value used to quantify the intensity of low-frequency activity in the frontal cortex associated with sedation, drowsiness, or introspective states. Its setting is based on the neurophysiologically accepted correlation between EEG frequency bands and cognitive states. Occipital lobe Alpha wave energy refers to the power integral value of the 8 Hz to 13 Hz frequency components extracted from EEG signals recorded by scalp electrodes located in the occipital lobe region of the brain. It is a scalar value used to quantify the intensity of rhythmic activity specific to the visual cortex in a relaxed state with eyes closed, and is a core indicator for assessing the degree of relaxation. The baseline skin conductivity level refers to the long-term, slowly changing baseline value of sweat gland activity after removing transient fluctuations from the skin conductivity signal. It is a scalar value, with the unit being microsiemens (μS), and reflects the baseline activity level of sympathetic cholinergic fibers. Its setting is based on the fact that the skin conductivity signal can be decomposed into slowly changing skin conductivity level components and rapidly changing skin conductivity response components.

[0084] For example, the data used to verify the effectiveness of simultaneous EEG signal acquisition and feature extraction were as follows: 120 seconds of data were acquired using an EEG device with a sampling rate of 250 Hz while the patient was relaxed. The power spectrum was calculated after applying a 4-8 Hz bandpass filter to the F3 lead signal to obtain the Theta wave energy. The value is 15.2 μV²; the power spectrum is calculated after applying an 8-13 Hz bandpass filter to the O1 lead signal to obtain the Alpha wave energy. The value was 42.7 μV². This data directly validates the ability to extract the two key feature values ​​of frontal lobe Theta wave energy and occipital lobe Alpha wave energy from synchronously acquired EEG signals. The data validating the calculation of baseline skin conductance levels were obtained by removing transient peaks exceeding 0.05 μS from synchronously acquired 120 s skin conductance signals and calculating the average value of the remaining signal. The baseline duration was 2.31 μS, and the standard deviation of this signal was less than 0.1 μS throughout the baseline period. This data directly validates that a stable baseline skin conductance level, representing an individual's relaxation baseline, can be calculated from the skin conductance signal. An example of the final combined baseline state vector is as follows: , (unit: μV², μV², μS), this vector represents the physiological baseline state of the individual patient.

[0085] In one embodiment of the present invention, a blunt-tipped probe is included, and a safety stress vector construction module is used to perform the following steps:

[0086] Based on the simulated operation marker signal detected by the piezoelectric thin film array and triggered by the picking and placing action of the blunt probe; EEG signals, skin conductance signals and operation marker signals are collected simultaneously during the simulated operation; within the time window marked by the operation marker signal, the degree of alpha wave suppression of the EEG signal, the peak response and recovery time of the skin conductance signal are extracted and quantified to construct a safety stress vector.

[0087] Specifically, a trained acupuncturist performs a standardized simulation: using a sterilized, blunt-tipped probe with a spherical or hemispherical tip that will not pierce the skin, a gentle and brief pressure stimulus is applied to a skin area outside the patient's planned acupuncture treatment site, such as a non-acupoint area on the contralateral arm. The pressure is controlled to a level that can elicit a clear tactile sensation but without pain, for example, a pressure value of 50 kPa and a duration of 1 second. The setting of the pressure value of 50 kPa and the duration of 1 second is based on the following: this pressure level is equivalent to the upper limit of the skin's tactile threshold, which can reliably elicit an autonomic nerve response without causing pain or tissue damage, and the duration of 1 second ensures that the stimulation is sufficiently obvious and repeatable.

[0088] Simultaneously, the acupuncturist uses their fingers on the operating area of ​​the intelligent needle placement tray to simulate a complete needle placement and removal action, ensuring that the piezoelectric film array synchronously generates a standard operation marker signal. Throughout the simulation, three signals are simultaneously acquired: the patient's real-time EEG signal, the skin conductance signal, and the operation marker signal triggered by the simulated placement and removal action. Using the precise start and end times marked by the operation marker signal, a precise analysis time window is defined. This window begins at a short baseline before the rising edge of the operation marker signal and ends after the falling edge of the operation marker signal, allowing sufficient time to capture the physiological signal recovery process. For example, the total window duration is 30 seconds, with the analysis time window set from 2 seconds before the start of the operation to 28 seconds after the end, for a total duration of 30 seconds. This setting is based on the typical time characteristics of skin conductance response: latency of 1-3 seconds, rise time of 2-5 seconds, and recovery time of over 20 seconds, ensuring complete capture of the entire stimulus response process. In some embodiments, the total window duration can be adjusted according to actual conditions.

[0089] Within this time window, the synchronously acquired EEG signals are processed: first, the EEG signals undergo the same preprocessing as the baseline state vector construction module; then, for the EEG data in the occipital leads, the average alpha wave energy during the presentation of the simulated stimulus is calculated. and compared it with the obtained individual baseline occipital alpha wave energy. Compare and calculate the degree of alpha wave suppression. The calculation formula is as follows:

[0090]

[0091] This value quantifies the brain's normal alertness-inhibiting response to mild external stimuli; among which, Represents the degree of alpha wave suppression, expressed as a percentage. The baseline value of the occipital lobe alpha wave energy obtained in the individual's quiet state is expressed in μV². This represents the average occipital alpha wave energy calculated within the simulation operation time window, expressed in μV².

[0092] Simultaneously, the synchronously acquired skin conductance signals are processed: firstly, the baseline value of skin conductance during the stable period prior to the simulated stimulation is determined. Then, identify the peak value of the skin conductance signal after the stimulation marked by the operational label signal. Calculate the peak value of skin electrical conductivity response:

[0093]

[0094] This value quantifies the intensity of the autonomic nervous system's transient response to stimuli, where, Represents the peak value of skin electrical conductivity response, measured in μS; This represents the maximum value of the skin conductance signal after stimulation, measured in μS. This represents the average skin conductance during a stable period prior to stimulation, expressed in μS.

[0095] Next, starting from the peak point, the trajectory of the skin conductance signal decline is traced backward to determine the time required for the signal amplitude to recover to half the difference between the peak and the baseline, i.e., the recovery half-life. However, the complete recovery time is not defined in this step. The skin conductance signal was measured from the peak point back to the baseline value. The total duration of the stimulus quantifies the speed at which the physiological system recovers from the stimulus. Ultimately, these three quantitative indicators extracted under standardized safe stimuli—the degree of alpha wave inhibition—are used to analyze the response. Skin conductance peak Recovery time These features are combined in a fixed order to form new multidimensional feature points, which are defined as the safety stress vector characterizing the patient's normal and acceptable physiological response limit to mild and controllable stimuli. Its mathematical expression can be regarded as a three-dimensional vector:

[0096]

[0097] The blunt-tipped probe is a tool with a smooth, rounded tip that won't pierce the skin. It's used to apply standardized, non-invasive mechanical stimulation, designed to reliably induce a mild but safe external stimulus. The non-treatment area refers to skin regions on the patient not planned for the current acupuncture treatment. These are typically chosen on the contralateral limb or away from acupoints; their geographical location serves as a safe application site for stimulation to avoid confusion or presensitization with subsequent treatment. Simulating a single needle removal and release action refers to the practitioner using their fingers to simulate the complete pressure application and release process of a needle removal and release on the intelligent needle placement panel, without using actual needles. This is a standardized procedure that triggers a synchronized operational marker signal with the safety stimulus, precisely defining the time window for physiological signal analysis. Alpha wave inhibition refers to the percentage decrease in occipital lobe alpha wave energy relative to baseline levels caused by external stimuli. It is a scalar value used to quantify the intensity of the central nervous system's normal attentional orientation or alertness response to stimuli. Its definition is based on the common phenomenon in neuroscience related to alpha wave desynchronization and stimulus processing. The peak response of skin conductance refers to the maximum amplitude of the instantaneous change in skin conductance after stimulation. It is a scalar value measured in microsiemens and used to quantify the instantaneous excitation intensity of sympathetic cholinergic fibers in response to stimulation. Recovery time refers to the total time it takes for skin conductance to recover from its peak point after stimulation to its baseline level before stimulation. It is a scalar value measured in seconds and used to quantify the speed at which the autonomic nervous system recovers from an excited state to a calm state.

[0098] For example, data used to verify the effectiveness of the simulated operation and feature extraction includes: during the simulated operation, the operation marker signal clearly marked the start and end times. Within the marked time window, the average alpha wave energy was calculated from the EEG signal in the patient's occipital lobe O1 lead. It is 25.6 μV², relative to its baseline energy. The value was 42.7 μV², and the degree of alpha wave suppression was calculated. The percentage was 40%. Simultaneously, the skin conductance signal decreased from the pre-stimulation baseline value. It reached its peak value at 2.31 μS. The reaction time was 3.15 μS, and the peak value was... The time required for the signal to fall back to the baseline of 2.31 μS from its peak is 0.84 μS. The time frame was 8.5 s. This data directly validated the ability to quantify and extract three key features—alpha wave suppression, peak skin conductance response, and recovery time—from signals synchronously acquired during simulation. The final example of the constructed individualized safety stress vector is as follows: This vector defines the boundary of the patient's normal physiological response to safe stimuli.

[0099] In one embodiment of the present invention, the instantaneous state vector generation module is used to perform the following steps:

[0100] During acupuncture treatment, real-time electroencephalogram (EEG) signals, skin conductance signals, and manipulation marker signals are continuously and synchronously collected.

[0101] The current values ​​of frontal lobe Theta wave energy and occipital lobe Alpha wave energy are calculated from real-time EEG signals; the amplitude of the skin conductance response associated with the operation marker signal and the recovery half-life value after the response are calculated from the skin conductance signal; the current values ​​of frontal lobe Theta wave energy, occipital lobe Alpha wave energy, skin conductance response amplitude and recovery half-life value are combined to generate an instantaneous state vector.

[0102] Specifically, throughout the acupuncture treatment process, a deployed integrated monitoring system continuously and synchronously acquires three real-time signals. The first signal is obtained through an EEG acquisition device, continuously acquiring real-time EEG signals covering the patient's frontal and occipital lobes. The second signal is continuously outputting skin conductance signals reflecting changes in the patient's skin conductance through conductive fabric electrodes integrated on the intelligent needle placement tray. The third signal is continuously monitored through a piezoelectric film array embedded in the operating area of ​​the intelligent needle placement tray, monitoring the needle placement and removal operations and outputting operation marker signals indicating the start and end times of the operation. An internal high-precision clock ensures strict temporal synchronization of these three data streams. Real-time EEG signals are continuously analyzed using a sliding time window approach.

[0103] For the EEG data within the current analysis time window, bandpass filtering and artifact removal preprocessing are first performed. Then, the power of the Theta band from 4 Hz to 8 Hz is extracted from the EEG data of the frontal lobe leads, and its average value within this time window is calculated and defined as the current value of the frontal lobe Theta wave energy. Simultaneously, the power of the 8 Hz to 13 Hz alpha band was extracted from the EEG data of the occipital lobe leads, and its average value within this time window was calculated and defined as the current value of the occipital lobe alpha wave energy. For skin conductance signals, two parallel calculations are performed.

[0104] The first step is to calculate the skin conductance response triggered by transient sympathetic nerve activity. First, the rapid rising edge of the skin conductance signal is detected. The signal is then temporally correlated with the operation marker signal. Within a short time window after each operation begins, the difference between the peak value of the skin conductance signal and the average value of the short baseline period before the operation begins is calculated. This difference represents the amplitude of the skin conductance response corresponding to that stimulus. .

[0105] The second step is to calculate the recovery half-life. Value: After a stimulus event clearly marked by an operational marker signal, the waveform of the skin conductance response triggered by that event is locked. First, the initial baseline value of the response is determined. and peak Calculate the amplitude difference Then, starting from the peak point, the decay process of the skin conductance signal is traced backward to accurately pinpoint the first point at which the signal amplitude drops to ( +0.5 The time corresponding to this level (Δ) is defined as the time difference between this time and the peak time, which is the recovery half-life of the skin conductance level recovering to half of the baseline after the stimulation event. value.

[0106] Finally, the four scalar features calculated within the same time window are: the current value of the frontal lobe Theta wave energy. Current value of occipital lobe Alpha wave energy skin conductance response amplitude and recovery half-life The values ​​are combined in a fixed order to generate an instantaneous state vector that is updated in real time and represents the patient's current multidimensional physiological state. Its mathematical expression can be regarded as a four-dimensional vector:

[0107]

[0108] Calculate the amplitude of skin conductance response:

[0109]

[0110] in, This represents the amplitude of skin electrical conductivity response, measured in μS. This represents the maximum value of the skin conductance signal within the time window after the start of the operation; This represents the average skin conductance signal within a baseline time window prior to the start of the procedure. The setting is based on standard analytical methods for event-related skin conductance responses, and the time window length is determined based on physiological response delay characteristics, such as 1-5 seconds after stimulation.

[0111] Calculate the recovery half-life :

[0112]

[0113] in, Represents the recovery half-life, measured in seconds (s). It is time Skin conductivity at the location; It is the baseline value before stimulation; It is the reaction peak value; This refers to the moment corresponding to the peak value. The setting is based on a commonly used metric for the recovery process of first-order system responses, used to quantify the dynamic characteristics of autonomic neural recovery. This formula defines the moment at which the conditions are met. The operation, the calculation result For time.

[0114] Real-time EEG signals refer to continuously acquired EEG voltage sequences during treatment, without long-term averaging; these are continuous, multi-channel time-series data. The current value refers to a characteristic scalar calculated for the most recent or currently processed time window; it is a value updated over time. The skin conductance response induced by transient sympathetic activity refers to the sweat gland secretion activity triggered by external stimuli and mediated by sympathetic cholinergic fibers. In skin conductance signals, this manifests as a rapid, measurable increase in waveform; it is an event-related transient signal characteristic, with amplitude measured in microsiemens.

[0115] Recovery half-life The value is a time parameter, specifically referring to the time required for the skin conductance signal to decay from its peak to half the difference between the peak and baseline after a single stimulus response. It is a scalar value measured in seconds (s) and is used to quantify the speed at which the physiological system recovers from an excited state. Its setting is based on simplified modeling of autonomic nervous system response dynamics and clinical experimental data. For example, based on the analysis of skin conductance signals during 200 clinical acupuncture sessions, it was found that… Abnormally prolonged values ​​are associated with the risk of adverse events.

[0116] For example, data used to verify the effectiveness of real-time feature calculation is as follows: at some point during the treatment process, an operation start marker is detected. Within the subsequent analysis window, the current value of the frontal lobe Theta wave energy is calculated. The current value of the occipital lobe Alpha wave energy is 28.5 μV². The peak value of the skin conductance signal associated with this operation was 18.3 μV². The time was 3.8 μS, the baseline value before operation. The amplitude of the skin conductance response was calculated to be 2.4 μS. The recovery half-life was 1.4 μS. Starting from the peak point, the skin conductance signal decreased to 2.4 + 0.5 × (3.8 - 2.4) = 3.1 μS after 9.2 s, thus the recovery half-life was [not specified]. The value was 9.2 s. This set of data directly validated the ability to calculate the current values ​​of the frontal lobe Theta wave energy, the occipital lobe Alpha wave energy, the amplitude of the skin conductance response, and the recovery half-life in real time. Value. An example of the final generated instantaneous state vector is as follows: This vector represents the patient's physiological state in real time.

[0117] In one embodiment of the present invention, the dynamic weight adjustment module is used to perform the following steps:

[0118] When the operation marker signal indicates the start of the operation, the EEG energy and skin conductance response components in the instantaneous state vector caused by the operation are assigned preset analysis weights; when the operation marker signal indicates the end of the operation, the normal analysis weights of the restored components are restored, and the trend of the recovery half-life value is analyzed; based on the weight coefficients corresponding to the current mode, the weighted values ​​of each component of the instantaneous state vector are calculated, and the corrected state vector is generated.

[0119] Specifically, dynamic weight adjustments are performed based on the operation marker signal to avoid operational interference. The real-time operation marker signal is continuously monitored. When the rising edge of the operation marker signal is detected, indicating the start of an acupuncture operation, the system logic control unit automatically switches the monitoring and analysis mode to a high-tolerance mode. In this mode, the components in the instantaneous state vector known to fluctuate violently due to instantaneous mechanical stimulation or operator action—that is, the current value of the frontal lobe Theta wave energy—are analyzed. Current value of occipital lobe Alpha wave energy and skin conductance response amplitude Assign a preset low analysis weight This low analysis weight This is a coefficient between 0 and 1, for example, set to 0.2. Its setting is based on an empirical value determined after statistical analysis of numerous physiological signal fluctuations during the operational period, to minimize the impact of transient artifacts on the overall state assessment. For example, a low analysis weight... The value ranges from 0.1 to 0.3, with a default value of 0.2. This value can be obtained by analyzing the electroencephalogram (EEG) and electrodermal transfer function (EDS) signals of 50 healthy subjects during simulated acupuncture procedures. For example, when the weight is 0.2, the contribution of the manipulation artifact to the state vector can be reduced by about 80%, while the retention rate of the true physiological trend is >90%.

[0120] At the same time, the recovery half-life Because the value reflects the recovery dynamics after stimulation, it is relatively less affected by instantaneous operational interference. Therefore, in this mode, its original weight is maintained, typically 1.0. The components of the instantaneous state vector are multiplied by their corresponding weight coefficients to obtain the initially weighted components. When the falling edge of the operation marker signal pulse is detected, indicating the end of a needle insertion operation, the system logic control unit automatically switches the monitoring and analysis mode to the high-sensitivity monitoring mode. In this mode, the... , and The analysis weights of these three components are from Restore to normal standard analysis weights ,For example =1.0.

[0121] Furthermore, this model focuses the analysis on the recovery half-life. The trend of value changes, specifically by calculating the current value. The value after several previous operations The degree of deviation of the moving average of the value; if the deviation exceeds a certain threshold, then... Emphasis coefficients are applied to the value components. To amplify its importance in the state assessment at that moment, the coefficient is set based on whether there is a progressive deterioration in the autonomic nervous system's recovery ability. For example, the trend deviation threshold is set at 20%, meaning that if the current... Value exceeds the last 3 times A value exceeding 20% ​​of the moving average is considered an abnormal trend. It is emphasized that the coefficient K ranges from 1.2 to 2.0, with a default value of 1.5. This coefficient is obtained through training on historical adverse event data. When K=1.5, the sensitivity to recovery delay increases by approximately 30%, and the false alarm rate increases by <5%.

[0122] Finally, the component values, adjusted according to the weights corresponding to the current mode, are recombine into a new vector, which is defined as the corrected state vector after operation interference avoidance processing. Its mathematical expression is, under the high tolerance mode:

[0123]

[0124] After switching to high sensitivity mode:

[0125]

[0126] Among them, whether or not a coefficient is applied depending on Trend analysis results of the value Represents the corrected state vector; This represents a low analytical weight and is a dimensionless coefficient. The representative standard has a high analytical weight and is a dimensionless coefficient. Representative to The emphasis coefficient of the value. The value of 0.2 is based on preliminary experimental data. This weight can effectively suppress more than 80% of transient amplitude fluctuations caused by operational artifacts, while retaining potential real slow-changing trend information. A value of 1.0 represents the standard reference weight; The value of 1.5 is based on clinical experience. When the value shows a continuous upward trend, the weight of its increment in indicating risk needs to be appropriately amplified to trigger more sensitive subsequent judgments.

[0127] Dynamic weight adjustment refers to the process of automatically changing the relative importance coefficients of each physiological characteristic parameter used to calculate the state vector according to different stages indicated by the operation marker signal. It is an event-based (operation start / end) conditional logic control strategy. High tolerance mode is a working state automatically activated by the system when the start of an operation is detected. It is a configuration state in which the system reduces its sensitivity to transient, high-intensity physiological signal fluctuations to filter artifacts caused by the operation itself. Low analysis weights are multiplication coefficients less than 1 applied to specific physiological characteristic components. They are dimensionless scalar parameters whose function is to attenuate the contribution of the corresponding component in subsequent analysis. High sensitivity monitoring mode is a working state automatically activated by the system when the end of an operation is detected. It is a configuration state in which the system restores normal sensitivity to all physiological characteristics and enhances the monitoring and analysis of dynamic characteristics reflecting the recovery process. Recovery half-life. The trend of value change refers to The trend of an autonomic nervous system's recovery ability, whether rising, falling, or remaining stable over a time series, is quantified by calculating its deviation from recent historical mean or fitting its slope. It is a scalar value or a Boolean judgment, used to assess the dynamic changes in autonomic nervous system recovery capacity. The corrected state vector is a new multidimensional state vector obtained after dynamic weight adjustment based on the operator-labeled signal. It is a four-dimensional vector data, where each component value is the product of the original instantaneous state vector component and its corresponding weight, representing the patient's physiological state after excluding transient interference from major operations.

[0128] For example, data used to verify the effectiveness of dynamic weight adjustment is as follows: an operation is detected, and the operation marker signal rises. The instantaneous state vector at this time is: Apply low analysis weights The corrected state vector under the high tolerance mode is obtained. This data directly verifies that during the operation period, low weights were successfully applied to the susceptible EEG energy and skin conductance amplitude components to suppress fluctuations. Subsequently, upon the end-of-operation signal, the system switched to high-sensitivity mode, restoring the standard weights. It is 1.0, and the analysis found that the current The value (10.0 s) is lower than the average of the previous three values. The value (7.0 s) shows a clear upward trend, therefore the emphasis factor is applied. The value is 1.5, resulting in the corrected state vector. This data directly verifies that the weights returned to normal after the operation, and it can also... The abnormal trends in values ​​are highlighted to generate a revised state vector that is more sensitive to potential risks.

[0129] In one embodiment of the present invention, the analysis and risk assessment module is used to perform the following operations:

[0130] A multidimensional state space is established with EEG energy, skin conductance response, and recovery half-life as coordinate axes; the baseline state vector and the safety stress vector are jointly projected onto the multidimensional state space to define the physiological state safety region; the modified state vector is mapped into a state evolution trajectory in the multidimensional state space; and the state evolution trajectory is continuously judged to meet the conditions of continuously deviating from the safety region, the evolution direction pointing to the preset risk region, and the evolution speed and amplitude exceeding the limits defined by the safety stress vector.

[0131] Specifically, firstly, an abstract multidimensional state space is established within the data processing unit. This space uses three key physiological / dynamic parameters as coordinate axes: the first axis represents a comprehensive index of brain electrical energy, specifically derived from the calculated current value of the frontal lobe Theta wave energy. Current value of occipital lobe Alpha wave energy A certain combination, such as ratio The first axis is defined to reflect the balance between alertness and inhibition in the brain; the second axis represents the weighted amplitude of the skin conductance response. The first axis reflects the instantaneous excitation intensity of the autonomic nervous system; the second axis represents the recovery half-life after trend emphasis. These values ​​reflect the recovery ability of the autonomic nervous system. Therefore, any point in this multidimensional state space corresponds to a physiological state determined by these three coordinate values.

[0132] Next, the individualized baseline state vector and individualized safety stress vector Projected onto this space. Since the original vector dimensions may differ, data transformation or mapping is required based on the spatial coordinate axis definitions. For example, ... In and Converted into a comprehensive index of brain electrical energy, such as ,Will As a reference baseline; In and These correspond to the second and third coordinate axes in space, and the mapping rules are as follows:

[0133] X-axis (comprehensive brainwave energy index): (right or (For real-time data)

[0134] Y-axis (skin conductance response): (right Or after weight adjustment );

[0135] Z-axis (recovery time): (right Or after the trend is emphasized value);

[0136] All coordinate values ​​were normalized using the following method: [The text abruptly ends here, likely due to an incomplete sentence or a formatting error.] or The value of the corresponding component is used as a reference, and the dimensionless coordinate is obtained by dividing by the reference value.

[0137] Centered on the projected tranquil state point, and combined with the reaction limits in each direction defined by the safe stress point, a closed region is delineated in space. This region is defined as the physiological state safety zone that characterizes the acceptable changes in the patient's physiological state under the current treatment environment. Its shape can be a multidimensional ellipsoid or a convex hull.

[0138] Subsequently, the continuously generated corrected state vectors In this multidimensional state space, points are sequentially represented as a series of points according to their chronological order of occurrence. Connecting these points with line segments forms a state evolution trajectory that extends from the start of treatment to the current moment. Finally, this state evolution trajectory is logically evaluated in real time to continuously check whether it simultaneously meets three preset risk conditions:

[0139] Condition 1: A series of consecutive points on the trajectory remain outside the safe zone for physiological state. This means that the physiological state has deviated from the individual's normal fluctuation range. For example, in Condition 1, the number N of consecutive points is set to 5, corresponding to a time window of approximately 5 seconds. If the sampling interval is 1 second, this value is based on clinical observation. A continuous deviation exceeding 5 seconds can be considered a non-transient fluctuation, which is highly correlated with precursors to adverse events.

[0140] Condition two: The trajectory's evolution direction over a recent period, determined by calculating the trajectory segment's direction vector or fitting a trend line, points to a pre-defined risk area. This risk area is characterized in space by high EEG energy composite indicators (reflecting high Theta / low Alpha), high skin conductance response amplitude, and a long recovery half-life. The coordinate range of the values ​​is jointly defined;

[0141] Condition 3: The product of the trajectory's evolution rate (i.e., the displacement of the state point in space per unit time) and amplitude (i.e., the distance from the boundary of the safe zone) after deviating from the safe zone, or another comprehensive indicator, exceeds the value determined by the safety stress vector. The defined reference threshold is used to quantify the limits of an individual's normal stress response. Only when the state evolution trajectory simultaneously meets the above three conditions is a high risk determined and an early warning is prepared to be triggered.

[0142] Define the coordinates of the multidimensional state space:

[0143]

[0144] The X-axis integrates the inhibitory (Theta) and alpha (Alpha) activities of the brainwave; an increased ratio generally indicates risk. Y and Z values ​​are directly used after weighting and trend emphasis. This design is based on the induction of physiological characteristics of adverse event precursors, reducing multidimensional information to three orthogonal dimensions with the highest discriminative power.

[0145] Judgment Condition 1: Let the boundary function of the safe physiological state region be F(X, Y, Z) ≤ 0. For the nearest N points on the trajectory... Check if all points satisfy the condition. If so, then condition one is satisfied. Here, N is the length of the time window, such as 30 seconds, set to ensure the persistence of the deviation rather than instantaneous fluctuations. Define the boundaries of the safe zone.

[0146] Judgment condition two: Calculate the principal direction vector of the last segment of the trajectory (e.g., the nearest m points). Define the center of the preset risk area as... .

[0147] The conditions for determining that the direction points to the risk area are: and That is, the angle between the displacement direction and the direction pointing to the center of the risk area is less than 90 degrees. This is a constant threshold set based on the clinical risk model.

[0148] Judgment condition three:

[0149]

[0150] in, To calculate the evolution rate, Where A is the sampling interval. A represents the calculated deviation magnitude. . To define the safe stress limit threshold, speed and amplitude The dynamic characteristics of state deterioration were quantified; threshold Scaling is performed based on the magnitude of the individual's safety stress vector to reflect personalized limits. It is the safety stress vector after each component is normalized to a dimensionless form. This is a proportionality factor (e.g., 1.5). : ,in and ,when , .

[0151] It should be noted that speed The magnitude of the amplitude is obtained by dividing the Euclidean distance between adjacent state points by the time interval. This is obtained by calculating the minimum Euclidean distance from the current state point to the boundary of the physiological state safety zone, reflecting the degree to which the state deviates from the normal range. (Comprehensive Influence Index) Using speed and amplitude The weighted sum, weight coefficients and This can be adjusted according to actual needs, for example, all can be set to 1.0. Individualized risk threshold. The normalized safety stress vector The modulus multiplied by the scaling factor (As shown in 1.5) The normalization method is as follows: divide each component of the safety stress vector by its corresponding reference value to make it a dimensionless quantity. For example, the degree of alpha wave suppression... (Originally a percentage) Divide by 100 to obtain the peak value of skin conductance response. Divided by individual baseline skin conductivity level Recovery time Divide by a reference time (e.g., 10 s) or the average recovery time of the population. When considering the combined impact indicators Exceeding the individualized risk threshold When the condition is met, it is determined that condition three is satisfied.

[0152] In some embodiments, a preset dimensionless threshold may also be used directly. .

[0153] The multidimensional state space is a mathematical model framework used to visualize and quantify changes in physiological states. It is an abstract Euclidean space with three orthogonal coordinate axes, each corresponding to a selected physiological feature dimension. The EEG energy composite index is a derived parameter obtained by fusing frontal lobe Theta wave energy and occipital lobe Alpha wave energy information. It is a dimensionless scalar value used to characterize the functional balance of the brain in a single dimension. Projection refers to the process of converting vectors originally represented by different feature sets into their coordinate values ​​in a new space according to the definition of spatial coordinate axes; it is a data mapping or transformation operation. The physiological state safety region is a continuous spatial region in the multidimensional state space, centered on the individual's calm baseline and bounded by their safe stress limits. It is a set of spatial points, whose boundaries can be described by mathematical inequalities, and its function is to define the "normal" fluctuation range of an individual's physiological state. The state evolution trajectory is a directed path drawn in chronological order in the state space by the modified state vectors; it is an ordered sequence of points. This reflects the continuous change of physiological state over time. The evolution direction refers to the spatial orientation of the state evolution trajectory in the most recent time period, indicated by the tangential or chordal direction. It is approximated by calculating the displacement vector between consecutive state points and is a three-dimensional direction vector. The predefined risk region is a predefined spatial area in the multidimensional state space. The physiological state combinations represented by its points (high Theta / Alpha ratio, high skin conductance response, long recovery time) have been clinically proven to be strongly correlated with a high risk of adverse events. It is a specific set of spatial points, such as a sphere or cube centered at a fixed coordinate point. The evolution speed refers to how quickly a state point changes position in the state space, measured by the arc length or chord length of the trajectory per unit time. It is a scalar value, and the unit can be normalized spatial distance per second. The evolution amplitude refers to the minimum distance the current state point deviates from the boundary of the physiological state safety region. It is a scalar value representing the degree of deviation.

[0154] For example, the data used to verify space construction and trajectory determination is: the patient's individual baseline state vector projected onto the normalized space as points. After projecting the safety stress vector and determining its boundaries, the safety zone is defined as follows: An ellipsoid centered at [center] with radii of [0.8, 1.5, 1.6] along each axis. During a certain period of treatment, five corrected state vectors were continuously obtained, and the resulting trajectory point sequence after projection was: P1=(1.2, 1.6, 1.8), P2=(1.4, 2.0, 2.2), P3=(1.7, 2.5, 2.8), P4=(2.0, 3.0, 3.5), P5=(2.4, 3.5, 4.3). Calculations show that points P1 to P5 are all located outside the safe zone of the aforementioned ellipsoid, satisfying condition one.

[0155] Calculate the displacement vector D = P5 - P3 = (0.7, 1.0, 1.5) from P3 to P5, and preset the center of the risk area. The calculation yields: However, observing the overall trend from P1 to P5 clearly points to... Therefore, a longer trend fitting is needed to determine if condition two is met.

[0156] Assumption and Both are 1. Given a value of 1.5, calculate the evolution rate between the two most recent points. The normalized distance is 0.45 / second, and the amplitude A is the distance from P5 to the boundary of the safe zone, which is 0.8. 1.25. Assuming the patient... If the value is 0.8, then the individual's safe stress limit threshold is... The value is 1.5 × 0.8 = 1.2. Since 1.25 > 1.2, condition three is satisfied.

[0157] In one embodiment of the present invention, the graded early warning execution module is used to perform the following operations:

[0158] When the state evolution trajectory initially shows a deviation from the trend, a level one warning is triggered, which is indicated by visual signals visible only to the acupuncturist; when the state evolution trajectory confirms that it has entered a risk evolution path, a level two warning is triggered, which is indicated by a combination of visual signals and signals perceptible to the acupuncturist.

[0159] Specifically, tiered early warnings are executed based on the risk assessment results. The system continuously receives and analyzes real-time risk assessment results. These results are based on a logical signal indicating whether three conditions are simultaneously met. A pre-defined two-level early warning response mechanism is executed according to the different states of this logical signal. When the judgment logic output indicates that the state evolution trajectory has begun to show a continuous trend of deviating from the safe physiological state (i.e., condition one is met, but conditions two and / or three are not fully met or the trend is weak), the system is determined to be in an early risk or uncertain state, triggering a level one early warning. The level one early warning is executed by generating a visual cue signal visible only to the acupuncturist. This signal is achieved through a dedicated display screen or indicator light provided by the system. For example, the color of a circular indicator icon in the corner of the display screen changes from green to a stable yellow, or a separate yellow LED light is kept constantly lit. The style, color, and position of the visual signal are predefined during system initialization. When the judgment logic output indicates that the state evolution trajectory clearly meets all three defined risk conditions simultaneously, the patient's physiological state is confirmed to have entered a high-risk evolution path, triggering a level two early warning.

[0160] The Level 2 warning system employs a composite alert mechanism, simultaneously generating visual signals and sensory vibration signals perceptible to the practitioner. The visual signal escalates under Level 2 warning; for example, the aforementioned circular indicator icon flashes red, or a separate red LED flashes at a high frequency. Simultaneously, a command is sent wirelessly or via wired connection to a miniature vibration motor within a dedicated wristband or handheld device worn by the practitioner. This motor generates a specific pattern of sensory vibration, such as continuous vibration at 2 Hz for 500 ms, followed by a 300 ms interval before repeating. This vibration pattern aims to alert the practitioner through touch without interfering with delicate procedures. The triggering conditions and response methods for both levels of warnings are pre-programmed and fixed in the system software, ensuring immediate and consistent responses.

[0161] Level 1 warning is a lower-level alert triggered by the system when it detects initial signs of risk. It's a warning level designed to provide early risk alerts to the acupuncturist through a non-invasive visual channel, based on the need to provide opportunities for observation and initial intervention before risk confirmation. The visual signal visible only to the acupuncturist refers to the output device, such as a small display screen or indicator light, whose installation location and display content are designed to ensure that only the operator can clearly perceive it within their field of vision, while the patient is not easily seen directly to avoid unnecessary psychological interference. It is directional optical or image information. Level 2 warning is a higher-level alert triggered by the system when a high-risk state is confirmed. It's a warning level that uses a combination of visual and sensory alerts to attract the acupuncturist's immediate attention with greater prominence and forcefulness. It's designed to ensure that warning information can be reliably and quickly perceived in high-risk situations.

[0162] Composite cues refer to the simultaneous use of two or more different sensory modalities to convey the same warning information, thereby improving the redundancy and reliability of information reception. It is a multimodal signal output strategy. Somatosensory vibration signals refer to a sequence of physical pulses generated by mechanical vibration elements that can be perceived by skin touch. It is a tactile stimulation signal with specific frequency, amplitude, and timing patterns, and its design is based on the sensitive frequency range of human tactile perception and the relatively low noise level of tactile cues in the operating environment.

[0163] For example, the data used to verify the effectiveness of triggering the Level 1 warning is as follows: During treatment, if the trajectory of the state evolution is determined to have 7 consecutive points outside the physiological safety zone (meeting condition 1), but the angle between its evolution direction and the center of the preset risk zone is 120 degrees (not meeting condition 2, which points to the risk zone), and the product of the evolution speed and amplitude is 0.18, which is lower than the individual's safe stress limit threshold of 0.25 (not meeting condition 3), then the Level 1 warning is triggered, and the status indicator icon on the display screen changes from green to a stable yellow.

[0164] The data used to verify the effectiveness of the Level 2 warning trigger was as follows: Subsequently, the new trajectory point sequence was determined to simultaneously meet three conditions: 10 consecutive points were outside the safe zone (condition 1), the average displacement direction of the last 5 points made an angle of 30 degrees with the center of the risk zone (condition 2), and the product of the evolution rate and amplitude increased to 0.38, exceeding the individual threshold of 0.25 (condition 3). Based on this, the Level 2 warning was triggered, the icon on the display screen turned red and flashed 3 times per second, and at the same time, the vibration motor in the acupuncturist's wristband began to vibrate at a frequency of 2 Hz for 500 ms with an interval of 300 ms.

[0165] In one embodiment of the present invention, the intervention command triggering module is used to perform the following steps:

[0166] When the evolution trajectory of the determined state breaks through the safety boundary defined by the safety stress vector and reaches the preset emergency threshold, it is determined to be a risk state; in response to the risk state, an intervention command is sent to the vibration device integrated at the bottom of the waiting area of ​​the intelligent needle tray; the vibration device responds to the intervention command and outputs mechanical vibration in a preset mode.

[0167] Specifically, continuously monitor the real-time risk assessment results. When the assessment logic confirms that the state evolution trajectory not only simultaneously satisfies its three risk conditions, but also that its current state point is relative to the safety stress vector... When the deviation from the defined safety boundary exceeds a preset emergency threshold, the patient is deemed to be at risk of adverse events such as fainting during injection. This emergency threshold is a value calculated proportionally based on the individualized safety stress vector magnitude, and its setting is based on clinical experience boundaries that distinguish between high risk and immediate danger.

[0168] Once an imminent risk is identified, the system's logic control unit immediately generates a digital intervention command. This command is sent directly to the pre-installed and calibrated vibration motor drive circuit integrated into the bottom of the intelligent needle placement tray's waiting area via wired or wireless communication protocols. Upon receiving the intervention command, the vibration motor generates a preset mode of mechanical vibration according to the parameters encoded within the command. The key parameter of this vibration mode is a specific frequency, such as a fixed value within the range of 30 Hz to 50 Hz, like 35 Hz. This setting is based on tactile perception studies and preclinical trials, which show that vibration within this frequency range can effectively stimulate the skin's tactile and proprioceptive afferent nerves without easily causing startle or discomfort. The vibration duration is typically set to last for several seconds, such as 3 seconds. The generated mechanical vibration is transmitted to the entire needle placement tray body through the rigid connection between the vibration motor and the needle placement tray structure.

[0169] Because the patient's body parts, such as the arm or wrist, remain in contact with the surface of the needle placement tray during treatment, this vibration is transmitted to the patient's body through the tray, creating a gentle, non-invasive somatosensory stimulation. This stimulation is intended as a physical arousal input, potentially disrupting an impending vicious cycle of autonomic nervous function, thus attempting to intervene before adverse events occur, forming a closed loop from risk perception to physical intervention.

[0170] Among them, "endangered risk" refers to the system's determination, based on individualized state evolution trajectory analysis, that the patient's physiological state is extremely close to or about to cross the critical point where a clinically visible adverse event will occur, such as fainting during a needle prick. This is the highest level of risk warning state, determined by the extent to which the state point exceeds the individual's safety boundary to an emergency threshold. The intervention command is a digital command containing specific control parameters, such as vibration frequency and duration, generated by the system's logic control unit when assessing the endangered risk. It is a structured data packet sent to the actuator through a predetermined communication interface. The vibration motor is a miniature motor that converts electrical signals into mechanical vibrations, embedded within the intelligent needle placement tray structure. It is the actuator hardware for closed-loop intervention actions, and its model and installation location are set based on the requirements of generating effective vibration intensity, having a small size for integration into the tray body, and reliable operation. Vibration at a specific frequency refers to the core periodic parameter of the mechanical vibration generated by a vibration motor, measured in Hz. It is a key physical characteristic of vibrational stimulation, and its numerical range (e.g., 30-50 Hz) is based on literature reports and preliminary experiments. This frequency band has potential effects on arousing mildly inhibited consciousness and falls within the sensitive range of human touch. Non-invasive arousal stimulation refers to physical energy (in this case, mechanical vibration) transmitted through somatosensory means without penetrating the skin or mucous membranes. It is a safe external intervention method whose function is to attempt to influence a deteriorating autonomic nervous system state through exogenous sensory input. Its effectiveness is hypothesized based on the regulatory effect of sensory stimulation on neural activity.

[0171] For example, the data used to verify the effectiveness of the closed-loop intervention command triggering and execution is as follows: At the end of treatment, the normalized distance of the current state point from the individual's safe stress boundary was calculated to be 2.3, exceeding the preset emergency threshold of 2.0, thus determining an imminent risk. The logic control unit then generates the intervention command "VIB_35 Hz_3000 ms" and sends it via serial port. Upon receiving the command, the miniature linear vibration motor integrated at the bottom of the needle placement tray drives its mass block to resonate at a frequency of 35 Hz for 3 seconds. Using a high-precision accelerometer attached to the surface of the needle placement tray, the effective value of the vibration acceleration was measured to be 0.8g (g is the acceleration due to gravity), and the main peak of the frequency spectrum was 35.1 Hz. This data directly verifies that when an imminent risk is determined, the system can correctly send the intervention command to the integrated vibration motor, and the vibration motor can generate specific vibrations that conform to the preset frequency and duration, thus constituting a complete closed-loop intervention.

[0172] Each of the modules can be implemented in whole or in part through software, hardware, or a combination thereof. It supports hardware embedded in or independent of the processor in the computer device, and also supports software stored in the memory of the computer device, so that the processor can call and execute the operations corresponding to each of the above modules.

[0173] 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 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. Such 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, and should all be included within the protection scope of the present invention.

Claims

1. An electroencephalogram (EEG) early warning system for adverse central nervous system events during acupuncture treatment, characterized in that, include: The device integration and time synchronization module deploys an integrated device monitoring system to establish a time synchronization relationship between operational behavior and physiological signals, and synchronously generates operation marker signals that mark the time of the acupuncturist's operation and skin conduction signals that reflect the patient's autonomic nerve activity; The baseline state vector construction module constructs a baseline state vector representing an individual's physiological baseline based on the electroencephalogram (EEG) and skin conductance signals acquired before treatment. The safety stress vector construction module synchronously collects EEG signals, skin conductance signals, and operation marker signals through simulated operations, and extracts and constructs a safety stress vector that represents the individual's normal stress limit. The real-time state vector generation module generates a real-time state vector representing the current physiological state based on real-time acquired EEG signals, skin conductance signals, and operation marker signals. The dynamic weight adjustment module performs dynamic weight adjustment on each component of the instantaneous state vector based on the operation marker signal to generate the corrected state vector. The analysis and risk assessment module, in a multi-dimensional state space, analyzes whether the time series trajectory formed by the corrected state vector meets the preset risk evolution conditions based on the physiological state safety region defined by the baseline state vector and the safety stress vector, and performs risk assessment. The tiered early warning execution module, based on the risk assessment results, executes tiered early warnings when the state evolution trajectory initially deviates from or is confirmed to enter a risk path. The intervention command triggering module triggers a closed-loop intervention command physically integrated with the early warning system when the state evolution trajectory breaks through the boundary of the physiological state safety zone defined by the safety stress vector, as well as a vibration device connected to the intervention command triggering module and used to generate a physical intervention signal in response to the intervention command.

2. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The appliance integration and time synchronization module is used to perform the following operations: A needle tray is provided, the surface of which is divided into a waiting area and an operation area. The operation area is embedded with a piezoelectric thin film array for sensing needle picking and placing operations to generate operation mark signals. The needle tray integrates conductive fabric electrodes for collecting patient skin conductivity signals to generate skin conductivity signals; The system uses an internal clock to time-align the operation marker signal with the skin conductance signal, thereby establishing a time-synchronous correlation between the operation event and the physiological signal.

3. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The baseline state vector construction module is used to perform the following steps: When the system is in baseline acquisition mode, it simultaneously acquires the patient's electroencephalogram (EEG) signals and skin conductance signals in a relaxed state. The energy of the frontal lobe Theta wave and the energy of the occipital lobe Alpha wave were extracted from the electroencephalogram (EEG) signals, and the basic skin conductance level was extracted from the skin conductance signals. The baseline state vector is generated by combining the Theta wave energy of the frontal lobe, the Alpha wave energy of the occipital lobe, and the baseline skin conductance level.

4. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, Including blunt-tipped probes, the safety stress vector construction module is used to perform the following steps: Based on the analog operation marker signal detected by the piezoelectric thin film array and triggered by the picking and placing action of the blunt probe; Simultaneously acquire EEG signals, skin conductance signals, and operation marker signals during the simulated operation; Within the time window marked by the operational label signal, the degree of alpha wave inhibition of the EEG signal, the peak response and recovery time of the skin conductance signal are extracted and quantified to construct a safety stress vector.

5. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The instantaneous state vector generation module is used to perform the following operations: During acupuncture treatment, real-time electroencephalogram (EEG) signals, skin conductance signals, and manipulation marker signals are continuously and synchronously collected. Calculate the current values ​​of frontal lobe Theta wave energy and occipital lobe Alpha wave energy from real-time EEG signals; Calculate the amplitude of the skin conductance response associated with the operation marker signal and the recovery half-life value after the response from the skin conductance signal; The instantaneous state vector is generated by combining the current values ​​of the frontal lobe Theta wave energy, the current values ​​of the occipital lobe Alpha wave energy, the amplitude of the skin conductance response, and the recovery half-life value.

6. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The dynamic weight adjustment module is used to perform the following steps: When the operation marker signal indicates the start of the operation, the EEG energy and skin conductance response components caused by the operation in the instantaneous state vector are assigned preset analysis weights. When the operation marker signal indicates that the operation is over, the normal analysis weight of the recovery component is restored, and the trend of the recovery half-life value is analyzed. Based on the weight coefficients corresponding to the current mode, calculate the weighted values ​​of each component of the instantaneous state vector to generate the corrected state vector.

7. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The analysis and risk assessment module is used to perform the following steps: Establish a multidimensional state space with brain electrical energy, skin conductance response, and recovery half-life as coordinate axes; The baseline state vector and the safety stress vector are jointly projected onto a multi-dimensional state space to define the physiological state safety region. The modified state vector is mapped into a state evolution trajectory in the multidimensional state space; The system continuously determines whether the state evolution trajectory meets the conditions of continuously deviating from the safe area, the evolution direction pointing to the preset risk area, and the evolution speed and magnitude exceeding the limits defined by the safety stress vector.

8. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The tiered early warning execution module is used to perform the following operations: When the state evolution trajectory initially shows a deviation from the trend, a level one warning is triggered, which is indicated by visual signals visible only to the acupuncturist. When the state evolution trajectory confirms that it has entered a risk evolution path, a level two warning is triggered, and a combined prompt is made through visual signals and signals that can be perceived by the acupuncturist.

9. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 1, characterized in that, The intervention command triggering module is used to perform the following steps: When the evolution trajectory of the determined state breaks through the safety boundary defined by the safety stress vector and reaches the preset emergency threshold, it is determined to be a risk state; In response to a risk situation, an intervention command is sent to the vibration device integrated at the bottom of the waiting area of ​​the intelligent needle tray; The vibration device responds to intervention commands and outputs mechanical vibration in a preset mode.

10. The EEG early warning system for adverse central nervous system events during acupuncture treatment according to claim 7, characterized in that, Determining whether the state evolution trajectory simultaneously satisfies the risk conditions includes the following steps: By checking whether the most recent consecutive points on the trajectory are all outside the physiological safety zone, it can be determined whether there is a continuous deviation. The evolution direction is determined by calculating the angle between the main direction vector of the last segment of the trajectory and the direction vector pointing to the center of the preset risk area. Calculate the state evolution rate and deviation magnitude, then calculate the comprehensive impact index, and compare the comprehensive impact index with the individualized threshold derived from the normalized safety stress vector. If the comprehensive impact index is greater than the individualized threshold, it is determined to be out of limit.

Images