Self-adaptive adjustment nerve electrophysiology monitoring device for orthopedic tumor operation

Abstract

The invention discloses a self-adaptive adjustment nerve electrophysiology monitoring device for an orthopedic tumor operation, and belongs to the technical field of medical equipment. The system comprises a signal acquisition module, a signal feature extraction module, a self-adaptive adjustment control module and an intraoperative feedback early warning module, the signal feature extraction module carries out multi-angle feature extraction on neural signals and accurately identifies changes of neural states in a complex signal environment, and the self-adaptive adjustment control module dynamically adjusts monitoring and stimulation strategies according to changes of real-time neural signals in an operation, compares pre-operation baseline signals with the real-time signals in the operation, and determines whether the neural states are changed or not according to the pre-operation baseline signals and the real-time signals in the operation. The change trend of the nerve state is accurately judged, whether the nerve signal is in the normal physiological fluctuation range or not is evaluated, the intra-operation feedback early warning module converts the nerve signal into visual and auditory feedback, the perception ability of an operator to the nerve state is improved, the nerve injury risk is dynamically evaluated, and the operation safety is improved. And according to the risk level, different levels of early warning mechanisms are triggered, and an operator is suggested to adjust an operation strategy.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and in particular to a self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery. Background Art

[0002] During orthopedic tumor surgery, nerve protection is a key factor in surgical success. However, because the surgical area is often close to neural tissue, even the slightest mistake during surgery can lead to nerve damage, which can cause motor dysfunction, paresthesia, and even permanent nerve damage. Therefore, intraoperative neuroelectrophysiological monitoring technology is widely used in orthopedic tumor surgery to monitor neurological function in real time, assisting the surgeon in identifying and protecting critical neural structures and reducing the risk of nerve injury.

[0003] Existing intraoperative neuroelectrophysiological monitoring devices mainly rely on a single or a few neural signal characteristics, such as electromyographic signals (EMG), somatosensory evoked potentials (SSEP), motor evoked potentials (MEP), etc., and evaluate the neural status through fixed monitoring parameters. However, due to factors such as individual differences, the impact of surgical operations and external interference, fixed monitoring parameters may not be able to adapt to the intraoperative conditions of all patients, resulting in insufficient monitoring sensitivity or a high false alarm rate. In addition, existing monitoring devices usually only provide passive feedback and cannot be intelligently adjusted according to real-time changes in the neural status during surgery, making it difficult for the surgeon to obtain the optimal neuroprotection plan in a timely manner, thereby affecting the safety of the operation and the effect of postoperative recovery. Summary of the Invention

[0004] The object of the present invention is to provide a self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above-mentioned object, the present invention provides the following technical solutions: an adaptively adjustable neuroelectrophysiological monitoring device for orthopedic tumor surgery, comprising: a signal acquisition module, a signal feature extraction module, an adaptive adjustment control module, and an intraoperative feedback warning module;

[0006] A signal acquisition module is configured to acquire intraoperative neural signals in real time; a signal feature extraction module is configured to analyze the neural signals and extract neural signal feature parameters; an adaptive adjustment and control module is configured to adaptively optimize the intraoperative neural monitoring plan based on the signal feature extraction module; and an intraoperative feedback and warning module is configured to provide real-time feedback to the operator based on the adaptive adjustment and control module, and to issue a warning signal when abnormal neural function is detected.

[0007] Furthermore, the signal acquisition module includes multiple electrode assemblies, which are respectively arranged at different nerve distribution positions in the surgical area to obtain nerve signals. The electrode assemblies are electrically connected to an amplification unit, which is used to amplify the collected nerve signals. The amplification unit is connected to a signal feature extraction module.

[0008] Furthermore, the signal feature extraction module includes:

[0009] a signal preprocessing unit configured to receive the neural signal after amplification and processing by the signal acquisition module, perform multi-stage filtering on the original neural signal through an adjustable filter and remove power supply noise, calculate the baseline value of the neural signal, perform baseline drift correction on the neural signal, and normalize the neural signal;

[0010] A feature parameter extraction unit is configured to extract feature parameters related to the neural function state from the preprocessed neural signal, wherein the feature parameters include time domain features, frequency domain features and time-frequency features;

[0011] The time domain features are extracted by calculating the amplitude, duration, mean, standard deviation and coefficient of variation of the neural signal, and the peak potential and peak interval time are counted; the frequency domain features are extracted by analyzing the spectral distribution of the neural signal through fast Fourier transform, and the power spectral density is calculated; the time-frequency features are extracted by short-time Fourier transform, and the instantaneous power change and time-frequency energy distribution parameters are extracted.

[0012] Furthermore, the signal feature extraction module further includes:

[0013] an adaptive feature optimization unit configured to perform baseline recording of the patient's neural signals before surgery to obtain a preoperative baseline signal, obtain individual features, the individual features including amplitude, frequency, and time characteristics of normal neural signals, and establish an initial feature parameter library for the individual patient through statistical analysis. The initial feature parameter library is used for real-time comparison of neural signals during surgery;

[0014] Monitor intraoperative neural signals and compare them with preoperative baseline signals in real time, adjusting feature extraction weights based on real-time neural signal status;

[0015] a signal classification unit configured to set a normal threshold and a damage threshold based on the extracted time domain features, frequency domain features, and time-frequency features, and classify characteristic patterns of the neural signals through pattern recognition, wherein the characteristic patterns include a normal neural state, a mildly damaged state, and a high-risk state;

[0016] The normal neural state refers to a state in which the amplitude, frequency, and time characteristics of the neural signal are within the normal threshold range; the mildly damaged state refers to a state in which the amplitude of the neural signal decreases or the frequency characteristics are abnormal, but does not reach the damage threshold; the high-risk state refers to a state in which the amplitude of the neural signal drops sharply below the damage threshold and the frequency characteristics are disordered;

[0017] The classification results are dynamically revised based on the patient's preoperative baseline signals.

[0018] Furthermore, the adaptive feature optimization unit includes:

[0019] Real-time monitoring of intraoperative neural signals and combining them with preoperative baseline signals to obtain an ordered signal set of neural signals;

[0020] According to different types of neural signals, the ordered signal set is classified to obtain a classified signal set;

[0021] The classified signal set includes an amplitude signal subset, a frequency signal subset, and a time characteristic subset, and the time characteristic subset includes multiple subsets;

[0022] Input the signal data of each subset in the classified signal set into the same coordinate system, and perform curve fitting to obtain the corresponding signal curve, and determine the amplitude signal curve, frequency signal curve and time characteristic curve group;

[0023] Determine the user's reliable signal threshold based on the user's preoperative baseline signal and surgical impact indicators;

[0024] Compare each curve in the amplitude signal curve, the frequency signal curve and the time characteristic curve group to see whether there is a local maximum point exceeding the reliable signal threshold;

[0025] If a local maximum point of a curve in the amplitude signal curve, the frequency signal curve or the time characteristic curve group exceeds the reliable signal threshold, a first weight increase is performed on the corresponding curve;

[0026] If none of the amplitude signal curve, frequency signal curve, and time characteristic curve group has a local maximum point exceeding the reliable signal threshold, the user's individual vital signs are judged to be stable;

[0027] The amplitude signal curve, frequency signal curve and time characteristic curve corresponding to the stable neural signal of the user's individual vital signs are used to comprehensively determine the curve fluctuation degree of each curve based on the maximum curve slope and local maximum point in the current monitoring period combined with the corresponding influence weight;

[0028] Adjust the weight of the corresponding curve based on the strength of the curve fluctuation;

[0029] If the fluctuation degree of the curve increases, the weight of the corresponding curve is increased; if the fluctuation degree of the curve decreases, the weight of the corresponding curve is decreased;

[0030] The initial feature extraction weight of the real-time neural signal is obtained based on the weight increase or weight decrease results;

[0031] Obtain the user's surgical stage to determine the intraoperative environment coefficient, and determine the human correlation coefficient based on the individual influencing factors during the operation;

[0032] The initial feature extraction weight is optimized by combining the intraoperative environment correlation coefficient and the human correlation coefficient to obtain the feature extraction weight, thereby adjusting the feature extraction weight of the neural signal state.

[0033] Furthermore, the adaptive adjustment control module includes:

[0034] An intraoperative signal analysis unit is configured to receive real-time neural signals during surgery from the signal feature extraction module, and analyze neural activity in real time by combining the extracted time domain features, frequency domain features, and time-frequency features, monitoring signal amplitude, frequency, power spectrum density, and instantaneous power changes. Combined with the initial feature parameter library, the intraoperative neural signals are compared in real time and the neural signal change trend is calculated to assess whether the neural signals exceed the normal physiological fluctuation range.

[0035] When abnormal changes in nerve signals are detected, the evaluation results are provided to the intraoperative feedback warning module;

[0036] a stimulation parameter adjustment unit configured to determine a nerve stimulation mode based on a characteristic pattern of an intraoperative nerve signal, wherein the nerve stimulation mode includes a single pulse stimulation mode, a multi-pulse stimulation mode, and a continuous stimulation mode; increase the stimulation frequency when the nerve signal enters a mildly damaged state; and reduce the stimulation intensity or suspend the stimulation when the nerve signal enters a high-risk state;

[0037] Automatically adjusts stimulation current, pulse width, and frequency based on the surgical instruments used during the procedure; determines a comprehensive signal change index based on real-time changes in nerve signals, and compares the comprehensive signal change index with each signal change index in a preset change-intensity database;

[0038] Based on the comparison result, the comprehensive signal change index is determined to correspond to the stimulation intensity adjustment scheme in the preset change-intensity database, thereby automatically adjusting the stimulation intensity.

[0039] Furthermore, the adaptive adjustment control module further includes:

[0040] A dynamic optimization unit is configured to read the stability and change rate of the intraoperative neural signal according to the change trend of the neural signal obtained by the intraoperative signal analysis unit, and adaptively adjust the signal sampling frequency; if the neural signal is stable, the sampling frequency is reduced; if large signal fluctuations are detected, the sampling frequency is increased;

[0041] Combined with the patient's preoperative baseline signal and real-time intraoperative data, the normal threshold and damage threshold of the nerve signal are dynamically set, and the threshold range is updated in real time through dynamic threshold calculation; if the nerve signal changes drastically, the threshold range is narrowed;

[0042] a monitoring intensity adjustment unit configured to dynamically adjust the monitoring mode based on the intraoperative neurological status and risk assessment results, wherein the monitoring mode includes a basic monitoring mode and an enhanced monitoring mode;

[0043] Among them, the basic monitoring mode is to perform low-frequency sampling and standard stimulation under normal conditions; the enhanced monitoring mode is to increase the sampling frequency under mildly damaged conditions and high-risk conditions.

[0044] Furthermore, the intraoperative feedback warning module includes:

[0045] a real-time status feedback unit configured to receive intraoperative neural signal analysis results provided by the adaptive regulation control module, convert the neural status information into visual data, provide real-time feedback during the operation, and construct a multi-level feedback mechanism to provide intraoperative neural status information through visual feedback and auditory feedback;

[0046] When providing visual feedback, the neural signal waveform, characteristic parameters and risk assessment results are displayed in real time on the monitoring interface, and different icons are used to distinguish different states; when providing auditory feedback, sound prompts of different frequencies are provided when neural signal abnormalities are detected.

[0047] Furthermore, the intraoperative feedback warning module further includes:

[0048] The risk assessment and early warning unit is configured to monitor the changing trend of nerve signals during surgery. By combining the preoperative initial characteristic parameter library and real-time signal analysis during surgery, it assesses the degree to which nerve signals deviate from the normal range and sets early warning thresholds to grade the risk of nerve injury. The graded judgment is specifically divided into low risk, medium risk, and high risk. Among them, when the nerve signal fluctuates slightly, it is judged as low risk; when the nerve signal deviates from the normal range, it is judged as medium risk; when the nerve signal changes drastically, it is judged as high risk.

[0049] The early warning mechanism is automatically triggered based on the results of the graded judgment. In low-risk situations, only a visual warning is given; in medium-risk situations, a visual and auditory dual warning is triggered; in high-risk situations, an emergency warning is triggered, and the operator is advised to adjust the operation strategy.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. The signal feature extraction module of the present invention uses time domain, frequency domain and time-frequency analysis methods to perform multi-angle feature extraction on neural signals, which can intuitively reflect the amplitude changes and spectral characteristics of neural signals, and can capture the frequency distribution of signals at different time points, so that changes in neural status can be accurately identified in complex signal environments. It has higher accuracy and stability and can provide richer information support for subsequent neural monitoring and regulation.

[0052] 2. The adaptive adjustment control module of the present invention can dynamically adjust the monitoring and stimulation strategies according to the changes in the real-time neural signals during surgery. By comparing the preoperative baseline signal and the real-time signal during surgery, it can accurately judge the changing trend of the neural state, and combine the various extracted characteristic parameters to evaluate whether the neural signal is within the normal physiological fluctuation range. When the neural signal shows abnormal changes, it can provide the evaluation results to the intraoperative feedback warning module, and can also trigger the stimulation parameter adjustment unit to automatically adjust the neural stimulation mode to reduce the risk of neural damage. The sampling frequency is adjusted according to the change rate of the neural signal, so that the data acquisition burden is reduced when the neural state is stable, and the sampling frequency is increased when the neural state fluctuates violently, so as to capture the signal changes more accurately.

[0053] 3. The intraoperative feedback warning module of the present invention enables the surgeon to understand the changes in the neural status more intuitively and take corresponding measures before risks occur. It provides visual data through the real-time status feedback unit, converts neural signals into intuitive visual and auditory feedback, and improves the surgeon's perception of the neural status. It uses the preoperative initial feature parameter library and the intraoperative real-time signal analysis results to dynamically evaluate the risk of neural injury and make graded judgments according to low risk, medium risk and high risk. According to the risk level, different levels of warning mechanisms are automatically triggered, from visual prompts to auditory warnings, and even emergency warnings are triggered in high-risk situations, suggesting that the surgeon adjust the operation strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a schematic diagram of the module of the neuroelectrophysiological monitoring device for orthopedic tumor surgery of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] See also Figure 1 , the present invention provides the following technical solutions:

[0057] An adaptively adjustable neuroelectrophysiological monitoring device for orthopedic tumor surgery, comprising: a signal acquisition module, a signal feature extraction module, an adaptive adjustment control module, and an intraoperative feedback warning module;

[0058] The signal acquisition module is configured to acquire intraoperative neural signals in real time; the signal feature extraction module is configured to analyze neural signals and extract neural signal feature parameters; the adaptive adjustment and control module is configured to adaptively optimize the intraoperative neural monitoring plan based on the analysis results of the signal feature extraction module; the intraoperative feedback and early warning module is configured to provide real-time feedback to the surgeon based on the output results of the adaptive adjustment and control module, and to issue an early warning signal when abnormal neural function is detected.

[0059] The signal acquisition module includes multiple electrode assemblies, which are arranged at different nerve distribution positions in the surgical area to obtain nerve signals. The electrode assemblies are electrically connected to the amplification unit, which is used to amplify the collected nerve signals. The amplification unit is connected to the signal feature extraction module.

[0060] In the above embodiment, the design of the signal acquisition module fully considers the stability and accuracy of intraoperative neural signal acquisition, and adopts a method of arranging multiple electrode assemblies at different nerve distribution positions in the surgical area to ensure that neural signals can be collected in all directions and at multiple angles, thereby improving the signal coverage range and avoiding monitoring blind spots caused by local electrode failure or signal fluctuations. The signal is amplified by the amplification unit to enhance the detectability of weak neural signals, ensuring that the signal input to the signal feature extraction module is purer, providing a reliable basis for subsequent signal analysis and feature extraction, effectively improving the sensitivity and accuracy of intraoperative neural signal monitoring, and providing the surgeon with more reliable neural status data, thereby reducing the risk of nerve damage during surgery.

[0061] Signal feature extraction module, including:

[0062] a signal preprocessing unit configured to receive the neural signal after amplification and processing by the signal acquisition module, perform multi-stage filtering on the original neural signal through an adjustable filter and remove power supply noise, calculate the baseline value of the neural signal, perform baseline drift correction on the neural signal, and normalize the neural signal;

[0063] A feature parameter extraction unit is configured to extract feature parameters related to the neural function state from the preprocessed neural signal, the feature parameters including time domain features, frequency domain features and time-frequency features;

[0064] The time domain features are extracted by calculating the amplitude, duration, mean, standard deviation and coefficient of variation of the neural signal, and the peak potential and peak interval time are counted; the frequency domain features are extracted by analyzing the spectral distribution of the neural signal through fast Fourier transform, and the power spectral density is calculated; the time-frequency features are extracted by short-time Fourier transform, and the instantaneous power change and time-frequency energy distribution parameters are extracted.

[0065] An adaptive feature optimization unit is configured to perform baseline recording of the patient's neural signals before surgery to obtain a preoperative baseline signal, acquire individual features, including amplitude, frequency, and time characteristics of normal neural signals, and establish an initial feature parameter library for the individual patient through statistical analysis. The initial feature parameter library is used for real-time comparison of neural signals during surgery;

[0066] Monitor intraoperative neural signals and compare them with preoperative baseline signals in real time, adjusting feature extraction weights based on real-time neural signal status;

[0067] a signal classification unit configured to set a normal threshold and a damage threshold based on the extracted time domain features, frequency domain features, and time-frequency features, and classify characteristic patterns of the neural signal through pattern recognition, wherein the characteristic patterns include a normal neural state, a mildly damaged state, and a high-risk state;

[0068] Among them, the normal neural state is when the amplitude, frequency and time characteristics of the neural signal are within the normal threshold range; the mildly damaged state is when the neural signal amplitude decreases or the frequency characteristics are abnormal, but does not reach the damage threshold; the high-risk state is when the neural signal amplitude drops sharply below the damage threshold and the frequency characteristics are disordered;

[0069] The classification results are dynamically revised based on the patient's preoperative baseline signals.

[0070] In the above embodiment, the signal feature extraction module improves the clarity of the signal through multi-stage filtering and denoising processing, and can also calculate the baseline value of the neural signal and perform baseline drift correction to eliminate the baseline drift error caused by factors such as poor electrode contact and changes in tissue conductivity, making the signal more stable. At the same time, the normalization processing eliminates the differences in signal amplitude and characteristic values between individuals, so that the neural signals of different patients can be analyzed under the same standard, improving the versatility and comparability of the device. The time domain, frequency domain and time-frequency analysis methods are used to extract multi-angle features of the neural signal. Among them, the time domain features can intuitively reflect the amplitude changes of the neural signal, the frequency domain features reveal the spectral characteristics of the signal, and the time-frequency features can capture the frequency distribution of the signal at different time points, so that the changes in the neural state can be accurately identified in a complex signal environment, with higher accuracy and stability, and can provide richer information support for subsequent neural monitoring and regulation.

[0071] The adaptive feature optimization unit includes:

[0072] Real-time monitoring of intraoperative neural signals and combining them with preoperative baseline signals to obtain an ordered signal set of neural signals;

[0073] According to different types of neural signals, the ordered signal set is classified to obtain a classified signal set;

[0074] The classified signal set includes an amplitude signal subset, a frequency signal subset, and a time characteristic subset, and the time characteristic subset includes multiple subsets;

[0075] Input the signal data of each subset in the classified signal set into the same coordinate system, and perform curve fitting to obtain the corresponding signal curve, and determine the amplitude signal curve, frequency signal curve and time characteristic curve group;

[0076] Determine the user's reliable signal threshold based on the user's preoperative baseline signal and surgical impact indicators;

[0077] Compare each curve in the amplitude signal curve, the frequency signal curve and the time characteristic curve group to see whether there is a local maximum point exceeding the reliable signal threshold;

[0078] If a local maximum point of a curve in the amplitude signal curve, the frequency signal curve or the time characteristic curve group exceeds the reliable signal threshold, a first weight increase is performed on the corresponding curve;

[0079] If none of the amplitude signal curve, frequency signal curve, and time characteristic curve group has a local maximum point exceeding the reliable signal threshold, the user's individual vital signs are judged to be stable;

[0080] The amplitude signal curve, frequency signal curve and time characteristic curve corresponding to the stable neural signal of the user's individual vital signs are used to comprehensively determine the curve fluctuation degree of each curve based on the maximum curve slope and local maximum point in the current monitoring period combined with the corresponding influence weight;

[0081] Adjust the weight of the corresponding curve based on the strength of the curve fluctuation;

[0082] If the fluctuation degree of the curve increases, the weight of the corresponding curve is increased; if the fluctuation degree of the curve decreases, the weight of the corresponding curve is decreased;

[0083] The initial feature extraction weight of the real-time neural signal is obtained based on the weight increase or weight decrease results;

[0084] Obtain the user's surgical stage to determine the intraoperative environment coefficient, and determine the human correlation coefficient based on the individual influencing factors during the operation;

[0085] The initial feature extraction weight is optimized by combining the intraoperative environmental correlation coefficient and the human correlation coefficient to obtain the feature extraction weight, thereby adjusting the feature extraction weight of the neural signal state.

[0086] In the above embodiments, the preoperative baseline signal refers to the normal signal characteristics obtained by monitoring the patient's neural signals before surgery. The preoperative baseline signal serves as a reference standard for comparison with the real-time neural signal monitored during surgery to determine whether any abnormal changes in the neural signal occur during surgery.

[0087] In the above embodiment, the ordered signal set refers to the signal set obtained by arranging the intraoperative neural signals monitored in real time in chronological order and combining them with the preoperative baseline signals. The signals in the ordered signal set have a certain order and correlation. For example, in orthopedic tumor surgery, it is necessary to monitor neural signals such as heart rate in real time, and arrange the signal data collected at each monitoring moment in the current monitoring cycle in chronological order. At the same time, refer to the cardiac neural baseline signals recorded before the operation to form an ordered signal set. For example, from the beginning to the end of the operation, the cardiac neural signals are recorded every 1 minute, and the signal data are arranged in chronological order and combined with the baseline signals to form an ordered signal set.

[0088] In the above embodiment, the classified signal set is obtained by dividing the ordered signal set according to different types of neural signals, and is obtained by classifying the neural signals in the ordered signal set according to amplitude, frequency and time characteristics.

[0089] In the above embodiment, the amplitude signal subset, the frequency signal subset, and the time characteristic subset are subsets of amplitude, frequency, and time characteristics obtained by classifying neural signals.

[0090] In the above embodiment, a signal curve refers to a curve reflecting the signal variation trend, obtained by inputting the signal data of each subset of the classified signal set into the same coordinate system and using a curve fitting method. For example, by plotting the data points of the amplitude signal subset in the coordinate system and using curve fitting to obtain an amplitude signal curve, the amplitude variation trend over time can be displayed. Similarly, the frequency signal subset and the time characteristic subset data are fitted to obtain a frequency signal curve and a time characteristic curve group, respectively. The time characteristic subset contains multiple subsets, each corresponding to a curve, which together constitute the time characteristic curve group.

[0091] In the above embodiment, the reliable signal threshold is a threshold determined based on the user's preoperative baseline signal and surgical impact indicators, and is used to determine whether the intraoperative nerve signal is abnormal. When certain characteristics of the signal (such as the local maximum point) exceed this threshold, it is considered that the signal may have undergone abnormal changes. For example, based on the amplitude baseline signal of a certain nerve recorded before the operation, combined with various influencing factors that may occur during the operation (such as stimulation from surgical instruments, the influence of anesthetic drugs, etc.), the reliable signal threshold of the nerve amplitude signal is determined to be 3 millivolts. During intraoperative monitoring, if the local maximum point of the amplitude signal curve exceeds 3 millivolts, it is considered that the amplitude signal is abnormal.

[0092] In the above embodiment, a local maximum point refers to a point on the signal curve where the value is greater than the values of all adjacent points. This point is considered a local maximum point. A local maximum point reflects the peak value of a signal within a local range and can be used to determine whether the signal is experiencing abnormal fluctuations. For example, on an amplitude signal curve, if the amplitude value of a point is significantly higher than the amplitude values of its adjacent points, this point is considered a local maximum point of the amplitude signal curve. For example, under normal circumstances, the amplitude signal curve is relatively stable. However, during surgery, if a significant peak appears at a certain moment, the point corresponding to this peak is considered a local maximum point, and the corresponding value is the peak amplitude.

[0093] In the above embodiment, the degree of curve fluctuation is an indicator determined based on the maximum curve slope and local maximum point of the amplitude signal curve, frequency signal curve and time characteristic curve corresponding to the neural signal of the user's individual physical signs being stable in the current monitoring period, combined with the corresponding influence weights, to measure the strength of the curve fluctuation. For example, for an amplitude signal curve, within the current monitoring period, its maximum curve slope (reflecting the speed of curve change) is calculated, while considering the situation of the local maximum point. If the maximum curve slope is large and there are multiple obvious local maximum points, it means that the curve fluctuation degree is strong; conversely, if the maximum curve slope is small and there are fewer local maximum points, it means that the curve fluctuation degree is weak. Combined with the corresponding influence weights, the degree of fluctuation of the curve is comprehensively determined.

[0094] In the above embodiment, the initial feature extraction weight is the weight result obtained after weight adjustment of the corresponding curve based on the strength of the curve fluctuation, reflecting the initial importance of different curves in the neural signal feature extraction. For example, after analyzing the fluctuation degree of the amplitude signal curve, the frequency signal curve and the time characteristic curve, it is found that among the curves, the amplitude signal curve has a stronger fluctuation degree, the frequency signal curve has a moderate fluctuation degree, and the time characteristic curve has a weaker fluctuation degree. According to these fluctuation degrees, different initial weights are assigned to the three curves, such as the amplitude signal curve weight is 0.5, the frequency signal curve weight is 0.3, and the time characteristic curve weight is 0.2. These weights together constitute the initial feature extraction weight.

[0095] In the above embodiment, the intraoperative environment coefficient is a coefficient determined based on the user's surgical stage and is used to reflect the impact of the surgical environment on neural signal feature extraction. The surgical environment (such as the complexity of the surgical operation, the use of surgical instruments, etc.) may vary at different surgical stages, and the impact on neural signals may also vary. For example, in orthopedic tumor surgery, during the tumor cutting stage, the surgical operation is more complex and has a greater impact on neural signals. The intraoperative environment coefficient may be 1.2. During the suturing stage, the surgical operation is relatively simple, and the intraoperative environment coefficient can be adjusted to 0.8.

[0096] In the above embodiment, the human correlation coefficient refers to a coefficient determined based on individual influencing factors during surgery (such as individual differences in patients, anesthesia depth, operating habits of surgeons, etc.), which is used to reflect the influence of human factors on the extraction of neural signal features. For example, for two different patients A and B, due to individual differences (such as age, physical condition, etc.), their tolerance to surgery and anesthesia is different, and their neural signals are affected by human factors differently. Assuming that patient A is in good physical condition, the anesthesia depth is moderate, and the surgeon is more skilled in operation, the human correlation coefficient is set to 0.9; while patient B is in poor physical condition, the anesthesia depth fluctuates greatly, and the surgeon is relatively unfamiliar with the operation, the human correlation coefficient is set to 1.1.

[0097] In the above embodiment, the feature extraction weight refers to the final weight obtained after optimizing the initial feature extraction weight in combination with the intraoperative environment correlation coefficient and the human correlation coefficient, and is used to more accurately extract the features of the neural signal state. For example, it is known that the weight of the amplitude signal curve in the initial feature extraction weight is 0.5, the weight of the frequency signal curve is 0.3, and the weight of the time characteristic curve is 0.2; the intraoperative environment coefficient is 1.0, and the human correlation coefficient is 0.9. Through the weight optimization algorithm (such as multiplying the initial weight by the intraoperative environment coefficient and the human correlation coefficient and then normalizing it), the optimized feature extraction weight is obtained, and the amplitude signal curve weight becomes 0.45, the frequency signal curve weight becomes 0.27, and the time characteristic curve weight becomes 0.28.

[0098] The working principle of the above technical solution is: first, the real-time monitored intraoperative neural signals are combined with the preoperative baseline signals to obtain an ordered signal set, and the ordered signal set is divided into an amplitude signal subset, a frequency signal subset and a time characteristic subset, and each subset is input into the coordinate system for curve fitting to obtain a signal curve. Then, a threshold judgment is made based on the signal curve, and the curve fluctuation degree of the threshold judgment qualified curve is determined based on the slope of the curve and the local maximum point, so as to make corresponding weight adjustments and obtain the initial feature extraction weights. Finally, the weight optimization is performed in combination with the intraoperative environment coefficient and the human correlation coefficient to obtain the feature extraction weights, thereby realizing the feature extraction weight adjustment of the neural signal state.

[0099] The beneficial effects of the above technical solution are as follows: by real-time monitoring of intraoperative neural signals and comparing and classifying them with preoperative baseline signals, judging curve anomalies based on reliable signal thresholds and adjusting weights based on the degree of curve fluctuation, neural signal analysis can not only quickly identify abnormal and prominent key information, but also dynamically reflect the signal stability and reasonably allocate weights. The weights are then optimized by combining the intraoperative environmental coefficient and the human correlation coefficient, fully considering the surgical scenario and individual differences, thereby obtaining more accurate feature extraction weights. This effectively improves the accuracy of the adjustment of neural signal state feature extraction weights, providing more reliable protection for the monitoring and analysis of neural signals during surgery.

[0100] Adaptive regulation control module, including:

[0101] An intraoperative signal analysis unit is configured to receive real-time neural signals during surgery from the signal feature extraction module, and analyze neural activity in real time by combining the extracted time domain features, frequency domain features, and time-frequency features, monitoring signal amplitude, frequency, power spectrum density, and instantaneous power changes. Combined with the initial feature parameter library, the intraoperative neural signals are compared in real time and the neural signal change trend is calculated to assess whether the neural signals exceed the normal physiological fluctuation range.

[0102] When abnormal changes in nerve signals are detected, the evaluation results are provided to the intraoperative feedback warning module;

[0103] a stimulation parameter adjustment unit configured to determine a nerve stimulation mode based on a characteristic pattern of an intraoperative nerve signal, wherein the nerve stimulation mode includes a single pulse stimulation mode, a multi-pulse stimulation mode, and a continuous stimulation mode; increase the stimulation frequency when the nerve signal enters a mildly damaged state; and reduce the stimulation intensity or suspend the stimulation when the nerve signal enters a high-risk state;

[0104] Automatically adjusts stimulation current, pulse width, and frequency based on the surgical instruments used during the procedure; determines a comprehensive signal change index based on real-time changes in nerve signals, and compares the comprehensive signal change index with each signal change index in a preset change-intensity database;

[0105] Based on the comparison result, the comprehensive signal change index is determined to correspond to the stimulation intensity adjustment scheme in the preset change-intensity database, thereby automatically adjusting the stimulation intensity.

[0106] A dynamic optimization unit is configured to read the stability and change rate of the intraoperative neural signal according to the change trend of the neural signal obtained by the intraoperative signal analysis unit, and adaptively adjust the signal sampling frequency; if the neural signal is stable, the sampling frequency is reduced; if large signal fluctuations are detected, the sampling frequency is increased;

[0107] Combined with the patient's preoperative baseline signal and real-time intraoperative data, the normal threshold and damage threshold of the nerve signal are dynamically set, and the threshold range is updated in real time through dynamic threshold calculation; if the nerve signal changes drastically, the threshold range is narrowed;

[0108] A monitoring intensity adjustment unit is configured to dynamically adjust the monitoring mode based on the intraoperative neurological status and risk assessment results. The monitoring modes include basic monitoring mode and enhanced monitoring mode.

[0109] Among them, the basic monitoring mode is to perform low-frequency sampling and standard stimulation under normal conditions; the enhanced monitoring mode is to increase the sampling frequency under mildly damaged conditions and high-risk conditions.

[0110] In the above-described embodiment, the introduction of an adaptive regulation control module greatly enhances the level of intelligence, enabling it to dynamically adjust monitoring and stimulation strategies based on changes in real-time neural signals during surgery. By comparing preoperative baseline signals with real-time signals during surgery, it can accurately determine the changing trend of neural status and, combined with various extracted characteristic parameters, assess whether the neural signals are within the normal physiological fluctuation range. When abnormal changes in neural signals occur, the assessment results can be provided to the intraoperative feedback warning module, and the stimulation parameter adjustment unit can be triggered to automatically adjust the neural stimulation mode to reduce the risk of neural damage. For example, when the neural signal enters a mildly impaired state, the stimulation frequency can be automatically increased to enhance the neural response signal and improve the surgeon's perception of the neural status. When the neural signal enters a high-risk state, stimulation is reduced or suspended to avoid further aggravation of neural damage. In addition, the sampling frequency is adjusted according to the rate of change of the neural signal, reducing the data collection burden when the neural status is stable and increasing the sampling frequency when the neural status fluctuates violently to more accurately capture signal changes.

[0111] In the above embodiment, the calculation formula for determining the comprehensive signal change index S is:

[0112]

[0113] Among them, S is the comprehensive signal change index, f is the instantaneous frequency of the real-time neural signal at the current monitoring moment, and fmin is the minimum frequency of the real-time neural signal in the current monitoring cycle, f max is the maximum frequency of the real-time neural signal in the current monitoring period, a is the peak amplitude of the real-time neural signal in the current monitoring period, T i is the signal characteristic value corresponding to the i-th time characteristic index of the real-time neural signal at the current monitoring moment, T max,i is the maximum signal characteristic value corresponding to the i-th time characteristic index of the real-time neural signal in the current monitoring period, T min,i is the minimum signal characteristic value corresponding to the i-th time characteristic index of the real-time neural signal in the current monitoring period, h1 is the frequency conversion coefficient of the instantaneous frequency, h2 is the amplitude conversion coefficient of the amplitude, and p i is the index conversion coefficient of the i-th time characteristic index of the real-time neural signal, α, β, and γ are the frequency weight, amplitude weight, and time characteristic weight of the real-time neural signal, where the sum of the frequency weight, amplitude weight, and time characteristic weight of the real-time neural signal is 1.

[0114] The beneficial effects of the above technical solution are: by determining the comprehensive signal change index based on the real-time changes of the neural signal, and comparing it one by one with the signal change indicators in the preset change-intensity database, the intensity adjustment plan is determined, and the stimulation intensity is automatically adjusted, which can make the adjustment of the stimulation intensity more accurate and timely.

[0115] Intraoperative feedback and warning module, including:

[0116] a real-time status feedback unit configured to receive intraoperative neural signal analysis results provided by the adaptive regulation control module, convert the neural status information into visual data, provide real-time feedback during the operation, and construct a multi-level feedback mechanism to provide intraoperative neural status information through visual feedback and auditory feedback;

[0117] When providing visual feedback, the neural signal waveform, characteristic parameters and risk assessment results are displayed in real time on the monitoring interface, and different icons are used to distinguish different states; when providing auditory feedback, sound prompts of different frequencies are provided when neural signal abnormalities are detected.

[0118] The risk assessment and early warning unit is configured to monitor the changing trend of nerve signals during surgery. By combining the preoperative initial characteristic parameter library and real-time signal analysis during surgery, it assesses the degree to which nerve signals deviate from the normal range and sets early warning thresholds to grade the risk of nerve injury. The graded judgment is specifically divided into low risk, medium risk, and high risk. Among them, when the nerve signal fluctuates slightly, it is judged as low risk; when the nerve signal deviates from the normal range, it is judged as medium risk; when the nerve signal changes drastically, it is judged as high risk.

[0119] The early warning mechanism is automatically triggered based on the results of the graded judgment. In low-risk situations, only a visual warning is given; in medium-risk situations, a visual and auditory dual warning is triggered; in high-risk situations, an emergency warning is triggered, and the operator is advised to adjust the operation strategy.

[0120] In the above-mentioned embodiment, the intraoperative feedback and warning module enables the surgeon to more intuitively understand changes in neural status and take appropriate measures before risks occur. The real-time status feedback unit provides visual data, converting neural signals into intuitive visual and auditory feedback, thereby enhancing the surgeon's perception of neural status. Regarding visual feedback, the monitoring interface not only displays the waveform, characteristic parameters, and risk assessment results of neural signals in real time, but also uses different colors and icons to distinguish between normal, mildly impaired, and high-risk states, allowing the surgeon to quickly obtain key information. Regarding auditory feedback, when abnormal neural signals are detected, sound prompts of varying frequencies are emitted, ensuring that the surgeon receives timely warning information even when vision is restricted. Furthermore, the risk assessment and warning unit utilizes a preoperative initial characteristic parameter library and real-time intraoperative signal analysis results to dynamically assess the risk of neural injury and categorize it into low, medium, and high risk levels. Based on the risk level, different levels of warning mechanisms are automatically triggered, ranging from visual prompts to auditory alerts, and even emergency warnings in high-risk situations, advising the surgeon to adjust their operating strategy.

[0121] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. An adaptively adjustable neuroelectrophysiological monitoring device for orthopedic tumor surgery, characterized in that: include: Signal acquisition module, signal feature extraction module, adaptive adjustment control module and intraoperative feedback warning module; a signal acquisition module configured to acquire intraoperative neural signals in real time; a signal feature extraction module configured to analyze the neural signal and extract neural signal feature parameters, perform curve fitting on the intraoperative neural signal to obtain a signal curve and obtain initial feature extraction weights based on the signal curve, optimize the weights based on an intraoperative environmental coefficient and a human correlation coefficient, adjust the feature extraction weights of the neural signal state, and perform feature extraction of the neural signal state based on the adjusted feature extraction weights; an adaptive adjustment control module configured to determine a comprehensive signal change index based on the signal feature extraction module and the real-time changes of the neural signal, and compare the comprehensive signal change index with the signal change index in a preset change-intensity database one by one, thereby determining an intensity adjustment plan and adaptively optimizing the intraoperative neural monitoring plan; The intraoperative feedback warning module is configured to provide real-time feedback to the operator based on the adaptive regulation control module and to issue a warning signal when abnormal neurological function is detected.

2. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 1, characterized in that: The signal acquisition module includes an amplification unit and multiple electrode assemblies. The electrode assemblies are respectively arranged at different nerve distribution positions in the surgical area to obtain nerve signals. The electrode assemblies are electrically connected to the amplification unit. The amplification unit is used to amplify the collected nerve signals. The amplification unit is connected to the signal feature extraction module.

3. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 1, characterized in that: The signal feature extraction module, include: an adaptive feature optimization unit configured to perform baseline recording of the patient's neural signals before surgery to obtain a preoperative baseline signal, obtain individual features, the individual features including amplitude, frequency, and time characteristics of normal neural signals, and establish an initial feature parameter library for the individual patient through statistical analysis. The initial feature parameter library is used for real-time comparison of neural signals during surgery; Monitor intraoperative neural signals and compare them with preoperative baseline signals in real time, adjusting feature extraction weights based on real-time neural signal status; a signal classification unit configured to set a normal threshold and a damage threshold based on the extracted time domain features, frequency domain features, and time-frequency features, and classify characteristic patterns of the neural signals through pattern recognition, wherein the characteristic patterns include a normal neural state, a mildly damaged state, and a high-risk state; The normal neural state refers to a state in which the amplitude, frequency, and time characteristics of the neural signal are within the normal threshold range; the mildly damaged state refers to a state in which the amplitude of the neural signal decreases or the frequency characteristics are abnormal, but does not reach the damage threshold; the high-risk state refers to a state in which the amplitude of the neural signal drops sharply below the damage threshold and the frequency characteristics are disordered; The classification results are dynamically revised based on the patient's preoperative baseline signals.

4. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 3, characterized in that: The signal feature extraction module further includes: a signal preprocessing unit configured to receive the neural signal after amplification and processing by the signal acquisition module, perform multi-stage filtering on the original neural signal through an adjustable filter and remove power supply noise, calculate the baseline value of the neural signal, perform baseline drift correction on the neural signal, and normalize the neural signal; A feature parameter extraction unit is configured to extract feature parameters related to the neural function state from the preprocessed neural signal, wherein the feature parameters include time domain features, frequency domain features and time-frequency features; The time domain features are extracted by calculating the amplitude, duration, mean, standard deviation and coefficient of variation of the neural signal, and the peak potential and peak interval time are counted; the frequency domain features are extracted by analyzing the spectral distribution of the neural signal through fast Fourier transform, and the power spectral density is calculated; the time-frequency features are extracted by short-time Fourier transform, and the instantaneous power change and time-frequency energy distribution parameters are extracted.

5. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 4, characterized in that: The adaptive feature optimization unit includes: Real-time monitoring of intraoperative neural signals and combining them with preoperative baseline signals to obtain an ordered signal set of neural signals; According to different types of neural signals, the ordered signal set is classified to obtain a classified signal set; The classified signal set includes an amplitude signal subset, a frequency signal subset, and a time characteristic subset, and the time characteristic subset includes one or more subsets; Input the signal data of each subset in the classified signal set into the same coordinate system, and perform curve fitting to obtain the corresponding signal curve, and determine the amplitude signal curve, frequency signal curve and time characteristic curve group; Determine the user's reliable signal threshold based on the user's preoperative baseline signal and surgical impact indicators; Compare each curve in the amplitude signal curve, the frequency signal curve and the time characteristic curve group to see whether there is a local maximum point exceeding the reliable signal threshold; If a local maximum point of a curve in the amplitude signal curve, the frequency signal curve or the time characteristic curve group exceeds the reliable signal threshold, a first weight increase is performed on the corresponding curve; If none of the amplitude signal curve, frequency signal curve, and time characteristic curve group has a local maximum point exceeding the reliable signal threshold, the user's individual vital signs are judged to be stable; The amplitude signal curve, frequency signal curve and time characteristic curve corresponding to the stable neural signal of the user's individual vital signs are used to comprehensively determine the curve fluctuation degree of each curve based on the maximum curve slope and local maximum point in the current monitoring period combined with the corresponding influence weight; Adjust the weight of the corresponding curve based on the strength of the curve fluctuation; If the fluctuation degree of the curve increases, the weight of the corresponding curve is increased; if the fluctuation degree of the curve decreases, the weight of the corresponding curve is decreased; The initial feature extraction weight of the real-time neural signal is obtained based on the weight increase or weight decrease results; Obtain the user's surgical stage to determine the intraoperative environment coefficient, and determine the human correlation coefficient based on the individual influencing factors during the operation; The initial feature extraction weight is optimized by combining the intraoperative environmental correlation coefficient and the human correlation coefficient to obtain the feature extraction weight, thereby adjusting the feature extraction weight of the neural signal state.

6. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 1, characterized in that: The adaptive adjustment control module includes: An intraoperative signal analysis unit is configured to receive real-time neural signals during surgery from the signal feature extraction module, and analyze neural activity in real time based on the extracted neural signal feature parameters, monitor signal amplitude, frequency, power spectrum density, and instantaneous power changes, compare neural signals during surgery in real time with the initial feature parameter library, calculate neural signal change trends, and assess whether the neural signals exceed the normal physiological fluctuation range; When abnormal changes in nerve signals are detected, the evaluation results are provided to the intraoperative feedback warning module; a stimulation parameter adjustment unit configured to determine a nerve stimulation mode according to a characteristic pattern of an intraoperative nerve signal, wherein the nerve stimulation mode includes a single pulse stimulation mode, a multi-pulse stimulation mode, and a continuous stimulation mode; Automatically adjusts stimulation current, pulse width, and frequency based on the surgical instruments used during the procedure; determines a comprehensive signal change index based on real-time changes in nerve signals, and compares the comprehensive signal change index with each signal change index in a preset change-intensity database; Based on the comparison results, the comprehensive signal change index is determined to correspond to the stimulation intensity adjustment scheme in the preset change-intensity database, thereby automatically adjusting the stimulation intensity; A dynamic optimization unit is configured to read the stability and change rate of the intraoperative neural signal according to the change trend of the neural signal obtained by the intraoperative signal analysis unit, and adaptively adjust the signal sampling frequency; if the neural signal is stable, the sampling frequency is reduced; if large signal fluctuations are detected, the sampling frequency is increased; Combined with the patient's preoperative baseline signal and real-time intraoperative data, the normal threshold and damage threshold of the nerve signal are dynamically set, and the threshold range is updated in real time through dynamic threshold calculation; if the nerve signal changes drastically, the threshold range is narrowed; The monitoring intensity adjustment unit is configured to dynamically adjust the monitoring mode in combination with the intraoperative neurological status and risk assessment results, and the monitoring mode includes a basic monitoring mode and an enhanced monitoring mode.

7. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 6, characterized in that: The basic monitoring mode is to perform low-frequency sampling and standard stimulation under normal conditions; the enhanced monitoring mode is to increase the sampling frequency under mildly impaired conditions and high-risk conditions.

8. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 1, characterized in that: The intraoperative feedback warning module includes: a real-time status feedback unit configured to receive intraoperative neural signal analysis results provided by the adaptive regulation control module, convert the neural status information into visual data, provide real-time feedback during the operation, and construct a multi-level feedback mechanism to provide intraoperative neural status information through visual feedback and auditory feedback; When providing visual feedback, the neural signal waveform, characteristic parameters and risk assessment results are displayed in real time on the monitoring interface, and different icons are used to distinguish different states; when providing auditory feedback, sound prompts of different frequencies are provided when neural signal abnormalities are detected.

9. The self-adaptive neuroelectrophysiological monitoring device for orthopedic tumor surgery according to claim 7, characterized in that: The intraoperative feedback warning module further includes: The risk assessment and early warning unit is configured to monitor the changing trend of nerve signals during surgery. By combining the preoperative initial characteristic parameter library and real-time signal analysis during surgery, it assesses the degree to which nerve signals deviate from the normal range and sets early warning thresholds to grade the risk of nerve injury. The graded judgment is specifically divided into low risk, medium risk, and high risk. Among them, when the nerve signal fluctuates slightly, it is judged as low risk; when the nerve signal deviates from the normal range, it is judged as medium risk; when the nerve signal changes drastically, it is judged as high risk. The early warning mechanism is automatically triggered based on the results of the graded judgment. In low-risk situations, only a visual warning is given; in medium-risk situations, a visual and auditory dual warning is triggered; in high-risk situations, an emergency warning is triggered, and the operator is advised to adjust the operation strategy.

Images