Regional anesthesia block effect evaluation system based on medical big data

Abstract

The present application relates to the technical field of evaluation of blocking effect, and particularly relates to a regional anesthesia blocking effect evaluation system based on medical big data, which extracts the affected side photoplethysmogram of the limb on the side of the patient subjected to nerve block and the healthy side photoplethysmogram of the limb on the side not subjected to nerve block, then synchronously performs period division, baseline removal and normalization, further compares and analyzes the difference between the healthy side and the standardized volume distribution sequence of the patient in the same period, and the discrete features of the difference in the preset breathing observation period, to obtain the affected side vascular impedance characteristic parameters in each period; finally, at each time, according to the change features of the affected side vascular impedance characteristic parameters in the preset historical blocking monitoring period, a nerve block state signal is output. The present application introduces the healthy side reference of the same individual not subjected to block, compares and decouples the impedance features of the affected side vascular denervation in the general anesthesia state, and improves the accuracy of the evaluation of the regional anesthesia blocking effect.

Description

Technical Field

[0001] This invention relates to the field of nerve block effect evaluation technology, specifically to a regional anesthesia nerve block effect evaluation system based on medical big data. Background Technology

[0002] The combined application of nerve blocks and general anesthesia—combined or combined anesthesia—has become an important anesthetic strategy in many complex surgeries. This combined anesthesia strategy, through synergistic analgesia mechanisms, can provide precise surgical area analgesia using nerve blocks, reducing the dosage of general anesthetic drugs and thus lowering the risk of general anesthesia-related complications. Simultaneously, it can utilize general anesthesia to suppress the central nervous system, eliminating patient consciousness and alleviating anxiety, thus meeting the needs of specific surgeries. In clinical practice, accurately assessing the effectiveness of regional anesthesia blocks is crucial for determining the timing of surgical incision and the quality of anesthesia management.

[0003] Under general anesthesia, patients cannot express pain or temperature sensations, rendering traditional assessment methods such as acupuncture tests ineffective. Clinically used indicators such as perfusion index, blood pressure, and heart rate are used to assess the effectiveness of anesthetic blockade. However, general anesthetic drugs relax vascular smooth muscle, causing chemical vasodilation, while nerve blockade blocks sympathetic nerve signal transmission, causing the vascular wall to lose tension and resulting in denervation vasodilation. The "chemical vasodilation" caused by drugs and the "denervation vasodilation" caused by nerve blockade are highly overlapping in conventional time-domain monitoring indicators, making it difficult to distinguish whether the blockade has been completed from background noise, which seriously restricts the accuracy of regional anesthetic blockade effect assessment. Summary of the Invention

[0004] To address the technical problem of low accuracy in assessing the effectiveness of regional anesthesia blockade, the present invention aims to provide a regional anesthesia blockade effectiveness assessment system based on medical big data. The specific technical solution adopted is as follows: A regional anesthesia blockade effect evaluation system based on medical big data, the system comprising: Dual-source index processing module: used to extract the photoplethysmography (PPG) waves of the affected side of the limb with nerve block and the PPG waves of the unaffected side of the limb; simultaneously divide the two PPG waves into periods, remove the baseline and normalize them, and determine the normalized volume distribution sequence of the healthy side and the normalized volume distribution sequence of the affected side in each period. Vascular impedance decoupling module: used to obtain the vascular impedance characteristic parameters of the affected side in each cycle based on the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in the same cycle, and combined with the discrete characteristics of the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in different cycles within a preset respiratory observation period. The nerve block effect evaluation module is used to output a nerve block status signal at each time point based on the changes in the characteristic parameters of the vascular impedance on the affected side during the preset historical block monitoring period.

[0005] Furthermore, the two photoplethysmograms are simultaneously divided into periods, including: The healthy side photoplethysmography (PPG) wave is subjected to trough detection, and the time interval between two adjacent trough points is taken as a period; the healthy side PPG wave and the affected side PPG wave are simultaneously divided, and the healthy side PPG wave sub-segment and the affected side PPG wave sub-segment are determined in each period.

[0006] Furthermore, the method for obtaining the normalized volumetric distribution sequence of the healthy side and the normalized volumetric distribution sequence of the affected side includes: For each photoplethysmography (PPG) wavelet segment on the healthy side, the minimum value is taken as the baseline value on the healthy side. Each value within the PPG wavelet is subtracted from the baseline value on the healthy side to obtain the baseline-removed volume sequence on the healthy side. Each value in the baseline-removed volume sequence on the healthy side is normalized to obtain the standardized volume distribution sequence on the healthy side. The same acquisition method is used to obtain the standardized volume distribution sequence on the affected side.

[0007] Furthermore, the method for obtaining the characteristic parameters of the affected side's vascular impedance includes: Within each cycle, vascular compliance indices are obtained based on the deviation of the standardized volume distribution sequence on the affected side relative to the standardized volume distribution sequence on the healthy side. Within each cycle, the vascular elasticity loss index is obtained based on the deviation of the instantaneous change characteristics of the standardized volume distribution sequence on the affected side relative to the instantaneous change characteristics of the standardized volume distribution sequence on the healthy side. In each cycle, the vascular compliance index and the vascular elasticity loss index within each cycle are fused, and the vascular passive modulation index is obtained based on the discrete characteristics of the fusion results in different cycles within a preset respiratory observation period. The characteristic parameters of vascular impedance on the affected side include vascular compliance indicators, vascular elasticity loss indicators, and vascular passive modulation indicators.

[0008] Furthermore, methods for obtaining vascular compliance indicators include: Within each cycle, the bulldozer distance between the standardized volume distribution sequence on the affected side and the standardized volume distribution sequence on the healthy side is used as an indicator of vascular compliance.

[0009] Furthermore, the method for obtaining the vascular elasticity loss index includes: Within each cycle, the normalized volume distribution sequence of the healthy side and the normalized volume distribution sequence of the affected side are differentiated and normalized to obtain the volume change distribution sequence of the healthy side and the volume change distribution sequence of the affected side; the bulldozer distance between the volume change distribution sequence of the affected side and the volume change distribution sequence of the healthy side is used as an indicator of vascular elasticity loss.

[0010] Furthermore, the method for obtaining the vascular passive modulation index includes: Within each cycle, the vascular compliance index and the vascular elasticity loss index are standardized and then the Euclidean norm is calculated to obtain the vascular fusion characteristic index of the affected side; the standard deviation of the vascular fusion characteristic index of the affected side in different cycles within the preset respiratory observation period is used as the vascular passive modulation index.

[0011] Furthermore, the method for outputting the neural blockade state signal includes: The target indicator is selected from the vascular compliance index, the vascular elasticity loss index, and the vascular passive modulation index. At each time point, the comparison benchmark of the target indicator is determined based on the distribution characteristics of the indicator values ​​in all cycles under the target indicator within the preset historical blockage monitoring period. The benchmark deviation parameter of the target indicator at each time point is evaluated based on the deviation of the indicator value under the target indicator in the cycle to which each time point belongs from the comparison benchmark. At each time point, a neural blockade state signal is output based on the baseline deviation parameter for each metric.

[0012] Furthermore, the output of the neural blockade state signal based on the baseline deviation parameter for each indicator includes: At each time point, when the baseline deviation parameters corresponding to the vascular compliance index, the vascular elasticity loss index, and the vascular passive modulation index are all greater than the corresponding preset thresholds, a valid signal for the nerve block state is output; otherwise, an invalid signal for the nerve block state is output.

[0013] Furthermore, the preset respiratory observation period includes a preset number of respiratory cycles.

[0014] This invention has the following beneficial effects: First, it extracts the photoplethysmography (PPG) waves from the affected side of the limb with nerve block and the unaffected side of the limb without nerve block; simultaneously, it divides the two PPG waves into periods, facilitating independent decoupling of subsequent vascular microimpedance characteristics and eliminating waveform interference caused by heart rate changes. Then, it simultaneously removes the baseline and normalizes the PPG waves to eliminate background DC signal interference unrelated to nerve block and removes linear amplitude interference caused by general anesthetic drugs. This allows for the extraction of pure geometric features from the chaotic signal, thereby determining the standardized volume distribution sequence of the unaffected side and the standardized volume distribution sequence of the affected side within each period, providing a standardized analytical basis for the quantitative evaluation of nerve block effects; further, based on the same period... The difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side is used to offset the common vascular deformation caused by general anesthetic drugs, thereby decoupling the aliasing effect of general anesthetic drugs and nerve block. Combined with the discrete characteristics of the difference between the standardized volume distribution sequences of the healthy side and the affected side in different cycles within a preset respiratory observation period, the severity of the passive compression state of the vascular wall caused by respiration is quantified, thus obtaining the affected side vascular impedance characteristic parameters in each cycle. Finally, at each time point, based on the changes in the affected side vascular impedance characteristic parameters within a preset historical block monitoring period, the change in the affected side vascular impedance characteristic parameters is analyzed to assess the block onset status, thereby outputting a nerve block status signal. This invention introduces a reference from the unblocked healthy side of the same individual, comparing and decoupling the impedance characteristics of the denervated blood vessels on the affected side under general anesthesia, improving the accuracy of regional anesthesia block effect assessment. Attached Figure Description

[0015] Figure 1 A system module diagram of a regional anesthesia blockade effect evaluation system based on medical big data provided by the present invention; Figure 2 The flowchart illustrates a method for obtaining characteristic parameters of vascular impedance on the affected side, as provided by this invention. Detailed Implementation

[0016] like Figure 1 The diagram shown is a system module diagram of a regional anesthesia block effect evaluation system based on medical big data provided in an embodiment of the present invention, including a dual-source index processing module 101, a vascular impedance decoupling module 102, and a block effect evaluation module 103.

[0017] It should be noted that the embodiments of the present invention are applicable to the implementation scenario of regional nerve block on one side of the limb and the contralateral limb has no serious peripheral vascular disease. The unblocked limb on the contralateral side of the patient can be used as a physiological reference.

[0018] It should be noted that the medical big data in this plan specifically refers to the physiological characteristic stream data of individual patients during continuous cardiac cycles, which are accumulated in real time during the induction and maintenance of general anesthesia.

[0019] Dual-source index processing module 101: used to extract the photoplethysmography (PPG) wave of the affected side of the limb with nerve block and the PPG wave of the unaffected side of the limb; simultaneously divide the two PPG waves into periods, remove the baseline and normalize them, and determine the normalized volume distribution sequence of the healthy side and the normalized volume distribution sequence of the affected side in each period.

[0020] Photoplethysmography (PPG) characterizes the dynamic changes in blood volume in the microvascular bed of peripheral tissues during the cardiac cycle. It mainly consists of a direct current (DC) component and an alternating current (AC) component. The DC component reflects the basic light absorption of tissues, bones, and non-pulsatile blood, representing the basic volume. The AC component reflects the filling and emptying of arterial blood flow caused by cardiac pulsation, and its waveform profile directly reflects the mechanical characteristics (elasticity, compliance, tension) of the blood vessel wall. The essence of nerve blockade is the disappearance of sympathetic nerve tension, which leads to a "pathophysiological phase transition" in the blood vessels of the innervation area. After the sympathetic nerves are blocked, the blood vessel wall degenerates from an "active elastic tube" to a "passive compliant tube." This change in physical properties directly causes a deformation in the energy distribution of photoplethysmography (PPG). Furthermore, PPG can achieve continuous sampling at the millisecond level, capturing the weak chest pressure modulation signal generated with the respiratory cycle. PPG determines whether nerve blockade is effective by capturing the micromechanical differences before and after this phase transition. Because general anesthesia induction drugs (such as propofol) have a systemic vasodilatory effect, if the condition of the patient's limb affected by the nerve block (hereinafter referred to as the affected side) is monitored, it is impossible to distinguish whether the waveform change is caused by the "general anesthesia drug" or by the "nerve block"; while the patient's limb not affected by the nerve block (hereinafter referred to as the healthy side) is not affected by the block, and its pulse wave arrival time is stable, which can provide a physiological reference benchmark for the affected side.

[0021] Based on this, in one embodiment of the present invention, two photoplethysmography (PPG) sensors are first configured. One sensor is placed at the fingertip or toe of the patient's unblocked limb, and its collected data stream is defined as the healthy side control channel C. The other sensor is placed at the fingertip or toe of the patient's blocked limb, and its collected data stream is defined as the affected side monitoring channel D. The data stream is collected via an analog-to-digital converter at a sampling frequency. (Set to 125Hz or higher to preserve waveform details) Simultaneously acquire and record signals from two channels to obtain the photoplethysmography (PPG) waves of the affected side and the healthy side.

[0022] During general anesthesia, the patient's heart rate fluctuates continuously. The speed of the heart rate stretches or compresses the time axis of the waveform, which may overlap with the deformation caused by nerve block, making it difficult to identify whether the nerve block is effective. Furthermore, the pulse wave is essentially a pressure wave transmission generated by the heart pumping blood, and each heartbeat is a complete and independent event. To transform the chaotic, heart rate-disturbed continuous signal into a standardized, comparable physical energy distribution unit, one embodiment of the present invention further performs periodic division on two photoplethysmography pulse waves simultaneously, thereby facilitating the independent decoupling of subsequent vascular micro-impedance characteristics and eliminating waveform interference caused by heart rate changes.

[0023] Preferably, in one embodiment of the present invention, considering that in photoplethysmography (PPG), the trough corresponds to the end of diastole, at which point the peripheral vascular volume is at its lowest and the vascular wall pressure is minimal; compared to the peak, the trough is least affected by microscopic morphological fluctuations such as reflected waves and dicrotic notches, and its timing positioning is the most accurate; while the healthy side (non-blocked side) retains normal vascular tension and clear waveform characteristics, its trough positioning accuracy is relatively higher than that of the affected side; therefore, the (heartbeat) cycle can be determined by using the trough of the healthy side's PPG for synchronous division; thus, the synchronous division of the two PPG cycles includes: Valley detection was performed on the photoplethysmography (PPG) wave of the healthy side, and the time interval between the corresponding times of two adjacent valley points was taken as a cycle; the PPG wave of the healthy side and the PPG wave of the affected side were simultaneously divided, and the PPG wave sub-segments of the healthy side and the affected side were determined in each cycle.

[0024] It should be noted that the trough detection method is a well-known technology, which is briefly described here: First, the photoplethysmography pulse wave of the healthy side is filtered by moving average to eliminate high-frequency noise. Then, the local minimum point where the signal changes from falling to rising is identified by the first-order differential sign method. Finally, the local minimum point is screened by combining the preset minimum beat amplitude threshold and physiological heart rate zone constraints, thereby eliminating false troughs caused by baseline drift or high-frequency spikes. Specifically, a valid trough is only identified when the difference between the detected local minimum and the previous local maximum (peak) exceeds a preset minimum pulsation threshold, such as 10% of the current signal dynamic range. The time interval between adjacent troughs is set according to human physiological limits. If the interval between two troughs is less than Tmin, such as 0.4s, corresponding to a heart rate of 150bpm, it is considered high-frequency interference and the point is automatically discarded.

[0025] As an example, after identifying all the troughs in the photoplethysmography (PPG) wave of the healthy side, the time interval between the corresponding times of two adjacent troughs is taken as a period. For example, the time interval between the corresponding times of a certain trough is... The time corresponding to the next adjacent trough point The period between them is represented as the kth period: Therefore, it can be determined The healthy side and the affected side photoplethysmography (PPG) pulse wavelet segments within each cycle; the corresponding lengths of the healthy side and the affected side PPG pulse wavelet segments and the number of data acquisition points within each cycle. same.

[0026] Since photoplethysmography (PPG) is mainly composed of a DC component (tissue and venous baseline absorption) and an AC component (arterial blood pulsation), the degree of vasodilation varies at different stages of general anesthesia, which may cause the baseline of the DC component to be different at different stages. Furthermore, slight changes in sensor contact pressure or respiratory movements can also cause the baseline to drift overall. If the baseline is not removed, the large fluctuations in the DC component may mask the small deformations in the AC component. Baseline removal can align the signal to the physical zero point (trough alignment) and eliminate background DC signal interference unrelated to nerve block. Furthermore, since the amplitude of photoplethysmography (PPG) is related to vascular filling, vasodilation after general anesthesia induction leads to a dramatic increase in absolute amplitude. However, this increase in amplitude is systemic and not a specific characteristic of nerve block. Normalization can eliminate linear amplitude interference caused by general anesthetic drugs, thereby extracting pure geometric features from the chaotic signal and providing a standardized analytical basis for the quantitative evaluation of nerve block effect. Therefore, in this embodiment of the invention, after division, each independent unit (healthy side photoplethysmography pulse wavelet segment or affected side photoplethysmography pulse wavelet segment) is further subjected to baseline removal and normalization to determine the healthy side standardized volume distribution sequence and the affected side standardized volume distribution sequence in each cycle. This prepares for the subsequent precise decoupling of the nerve block features superimposed on the affected side standardized volume signal under the background of general anesthesia, using the healthy side standardized volume signal as a reference, and then outputting the nerve block status signal.

[0027] Preferably, in one embodiment of the present invention, considering that in each independent unit (healthy side photoplethysmography pulse wavelet segment or affected side photoplethysmography pulse wavelet segment), the minimum value can characterize the DC baseline generated within the corresponding (cardiac) cycle due to tissue baseline absorption, venous blood baseline light attenuation, etc., baseline removal processing can be performed; and normalization processing can be performed by dividing each value in the baseline-removed volume sequence of the healthy side by the sum of all its values; therefore, the method for obtaining the healthy side standardized volume distribution sequence and the affected side standardized volume distribution sequence includes: For each photoplethysmography (PPG) wavelet segment on the healthy side, the minimum value is taken as the baseline value on the healthy side. Each value within the PPG wavelet is subtracted from the baseline value on the healthy side to obtain the baseline-removed volume sequence on the healthy side. Each value in the baseline-removed volume sequence on the healthy side is normalized to obtain the standardized volume distribution sequence on the healthy side. The same acquisition method is used to obtain the standardized volume distribution sequence on the affected side.

[0028] As an example, taking any healthy side photoplethysmography (PPG) pulse wavelet segment as an example, its minimum value is taken as the healthy side baseline value. Subtracting the healthy side baseline value from each value (sampled value) within it can flatten the trough baseline of the pulse wave to zero potential, eliminating baseline error interference. Furthermore, dividing each value in the healthy side baseline-free volume sequence by the sum of all its values ​​(where, to avoid the sum of values ​​being 0, a very small zero-prevention positive parameter can be added to the sum of values, such as...) The probability is normalized to eliminate the interference of increased absolute amplitude caused by vasodilation induced by general anesthetic drugs, and the normalized volume distribution sequence of the healthy side is obtained. Similarly, the baseline value of the affected side in the photoplethysmography wavelet segment can be obtained, and the baseline can be de-normalized to obtain the standardized volumetric distribution sequence of the affected side.

[0029] Vascular impedance decoupling module 102: used to obtain the vascular impedance characteristic parameters of the affected side in each cycle based on the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in the same cycle, and combined with the discrete characteristics of the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in different cycles within a preset respiratory observation period.

[0030] Considering that vasodilation is systemic and synchronous under general anesthesia, and that nerve block causes the loss of sympathetic tone, which also leads to vasodilation, by analyzing the differences between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side within the same (cardiac) cycle, the common vascular deformation caused by general anesthetic drugs can be offset, thereby decoupling the aliasing effect of general anesthetic drugs and nerve block. Furthermore, considering that blood vessels without nerve innervation (after the blockade takes effect) are in a state of paralysis, their pulse waves fluctuate violently and periodically with respiratory pressure; while blood vessels with preserved nerve innervation (only during general anesthesia) can resist the fluctuations in chest pressure during mechanical ventilation through their own tension, and their pulse wave characteristic values ​​are relatively stable; by observing the discrete characteristics (fluctuations) of the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side during respiration, it is possible to observe the dynamic response of the blood vessel wall to the periodic (chest pressure) pressure test of respiration. When the blockade takes effect, this dynamic response (fluctuation) will be significantly amplified, thereby helping to distinguish between active vasodilators innervated by nerves and passive paralysis without nerve innervation, and achieving high specificity verification of the state of nerve blockade effect; Based on the above analysis, the embodiments of the present invention can further obtain the vascular impedance characteristic parameters of the affected side. The vascular impedance characteristic parameters of the affected side are a quantitative description of the changes in the physical properties of the vascular wall on the affected side, characterizing the vascular denervation characteristics caused by nerve block, and preparing for the subsequent output of nerve block state signals.

[0031] Preferably, in one embodiment of the present invention, please refer to Figure 2 The flowchart illustrates a method for obtaining characteristic parameters of vascular impedance on the affected side according to an embodiment of the present invention, including: Step S201: In each cycle, vascular compliance indices are obtained based on the deviation of the standardized volume distribution sequence on the affected side relative to the standardized volume distribution sequence on the healthy side.

[0032] Considering that after the nerve block takes effect, the sympathetic nerves lose their tension control over the blood vessel walls, and the vascular smooth muscle is completely relaxed, the pulse wave's energy is largely absorbed and buffered by the soft vessel walls when passing through such a relaxed vascular bed, resulting in a significant decrease in conduction velocity; by comparing the deviation of the standardized volume distribution sequence of the affected side relative to the standardized volume distribution sequence of the healthy side in each cycle, the overall phase shift of the affected side pulse wave relative to the healthy side on the time axis can be captured, and the vascular compliance index can be determined; the vascular compliance index characterizes the conduction delay characteristics of the affected side pulse wave and is the first physical indicator for determining the effectiveness of the nerve block.

[0033] In a preferred embodiment of the present invention, considering that traditional cross-correlation or feature point comparison (such as peak alignment) often fails to accurately find the alignment point when the waveform is severely distorted (circular passivation after hysteresis), leading to the failure of delay measurement; while the optimal transmission algorithm does not rely on specific feature points and can accurately calculate the optimal displacement cost to move the entire energy distribution from the source position to the target position, thereby stably outputting the quantified value of the conduction delay through the shift of the distribution's center of gravity, the optimal transmission solution process can be regarded as the solution process of the Earth Mover's Distance (EMD); furthermore, the Earth Mover's Distance can be used to measure the delay deviation and determine the vascular compliance index; therefore, the method for obtaining the vascular compliance index includes: Within each cycle, the bulldozer distance between the standardized volume distribution sequence on the affected side and the standardized volume distribution sequence on the healthy side was used as an indicator of vascular compliance.

[0034] It should be noted that both the standardized volumetric distribution sequence of the affected side and the standardized volumetric distribution sequence of the healthy side have been probability normalized, and the bulldozer distance can be calculated using the optimal transmission algorithm; the calculation of the bulldozer distance is a well-known technical method and will not be elaborated further.

[0035] In another embodiment of the invention, the implementer may also use a quadratic time difference as the ground cost matrix of the bulldozer when calculating the bulldozer distance, with a dimension of The ground cost matrix of the th Line number Column elements ,in, The i-th sampling point in the healthy side's standardized volumetric distribution sequence is moved to the i-th sampling point in the affected side's standardized volumetric distribution sequence. The physical cost consumed by each sampling point; The calculation formula uses squaring to nonlinearly amplify the lag penalty over long distances; then the optimal transmission algorithm is used to calculate the bulldozer distance; these are well-known technical methods and will not be elaborated further.

[0036] Step S202: Within each cycle, the vascular elasticity loss index is obtained based on the deviation of the instantaneous change characteristics of the standardized volume distribution sequence on the affected side relative to the instantaneous change characteristics of the standardized volume distribution sequence on the healthy side.

[0037] Normal blood vessel walls have elastic recoil force under nerve innervation, which makes the pulse wave produce a distinct dicrotic notch in the descending limb and remain steep in the ascending limb. However, after nerve block takes effect, the blood vessel wall loses the dynamic tension maintained by the nerve and degenerates into an inelastic passive conduit. This loss of elasticity leads to the wear and tear of the high-frequency characteristics of the pulse wave waveform, which manifests as a slower ascending limb, blurred or even disappearing dicrotic notch, and an overall rounded waveform. By comparing the instantaneous change characteristics of the standardized volume distribution sequence on the affected side with those on the healthy side within each cycle, the microscopic morphological changes of the pulse wave can be captured, thereby helping to capture the characteristics of vascular elasticity loss and determine the vascular elasticity loss index. The vascular elasticity loss index characterizes the morphological collapse characteristics of the pulse wave on the affected side and is the second physical indicator for determining the effectiveness of nerve block.

[0038] In a preferred embodiment of the present invention, considering that differential analysis can help extract the instantaneous change features of the waveform, further normalization is performed to facilitate subsequent calculation of the bulldozer distance, in order to measure the deviation of the instantaneous change features of the normalized volume distribution sequence on the affected side relative to the instantaneous change features of the normalized volume distribution sequence on the healthy side, thereby assessing the high-frequency characteristic wear of the pulse wave waveform and determining the vascular elasticity loss index; then the method for obtaining the vascular elasticity loss index includes: Within each cycle, the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side are differentiated and normalized to obtain the volume change distribution sequence of the healthy side and the volume change distribution sequence of the affected side. The bulldozer distance between the volume change distribution sequence of the affected side and the volume change distribution sequence of the healthy side is used as an indicator of vascular elasticity loss.

[0039] As an example, taking the healthy side normalized capacity distribution sequence within any period as an example, a first-order forward differencing operation is performed on the healthy side normalized capacity distribution sequence, and the absolute value is taken to extract the instantaneous change characteristics of the corresponding waveform of the healthy side normalized capacity distribution sequence; further, each difference absolute value is divided by the sum of all difference absolute values ​​(where, to avoid the sum of difference absolute values ​​being 0, a very small zero-prevention positive parameter can be added to the sum of difference absolute values, such as...). The probability is normalized, and the absolute values ​​of all normalized differences are sorted in chronological order to construct the volume change distribution sequence of the healthy side; similarly, the volume change distribution sequence of the affected side can be obtained, and the bulldozer distance can be further calculated to obtain the vascular elasticity loss index.

[0040] In another embodiment of the invention, the implementer may also use linear time difference as the ground cost matrix of the bulldozer when calculating the bulldozer distance, with dimension [missing information]. The ground cost matrix of the th Line number Column elements ,in, The absolute value of the difference in the i-th volume change distribution sequence of the healthy side is moved to the i-th volume change distribution sequence of the affected side. The physical cost consumed by the absolute value of each difference; Indicates taking the absolute value; The calculation formula uses linear distance to allow for small-range morphological jitter without incurring excessive penalties; then, the optimal transmission algorithm is used to calculate the bulldozer distance, which is a well-known technical method and will not be elaborated further.

[0041] In step S203, in each cycle, the vascular compliance index and vascular elasticity loss index within each cycle are fused, and the vascular passive modulation index is obtained based on the discrete characteristics of the fusion results in different cycles within the preset respiratory observation period.

[0042] Normal blood vessels can regulate tension through neural feedback to resist fluctuations in chest pressure caused by respiration, and the pulse wave characteristics are relatively stable. After nerve block, the blood vessels are in a state of near paralysis and cannot regulate themselves, resulting in drastic fluctuations in pulse wave characteristics with the respiratory cycle. The vascular compliance index and vascular elasticity loss index in different cycles will also oscillate violently with the respiratory cycle. By measuring the discrete characteristics, we can help quantify the severity of the passive pressure on the blood vessel wall, and thus determine the passive modulation index of the blood vessel. The passive modulation index of the blood vessel can characterize whether the blood vessel wall has the ability to regulate itself against external mechanical pressure interference, and is the first physical indicator for determining the effectiveness of nerve block.

[0043] It should be noted that the method for obtaining the passive vascular modulation indexes is the same for each cycle, and the analysis and description here will only take any one cycle as the target cycle.

[0044] In a preferred embodiment of the present invention, the preset respiratory observation period includes a preset number of respiratory cycles; wherein the preset number ranges from 2 to 6; in this example, it is 4. If the preset number is too small, it is insufficient to observe the effect of chest pressure fluctuations caused by breathing on the pulse wave; if the preset number is too large, it may lead to a delay in the determination of the onset of the blockade.

[0045] It should be noted that the respiratory cycle is determined by at least one of the following methods: synchronously acquiring the airway pressure cycle signal from the mechanical ventilation equipment, or performing low-pass filtering on the photoplethysmography sampling voltage sequence to extract the respiratory fluctuation component; these are well-known technical means and will not be elaborated further.

[0046] In another embodiment of the present invention, the (heartbeat) cycle corresponding to 4 respiratory cycles can also be estimated to determine the preset respiratory observation period; for example, 20 (heartbeat) cycles can be directly used as the preset respiratory observation period to prepare for subsequent analysis.

[0047] It should be noted that the preset respiratory observation period includes the target cycle, and the target cycle is the last cycle within the preset respiratory observation period. If the respiratory cycles within the preset respiratory observation period corresponding to the target cycle are less than the preset number, the target cycle is still in the observation accumulation stage, and the calculation of vascular passive modulation index and subsequent nerve block status signal output are not performed. Calculation and subsequent evaluation output will only begin when the respiratory cycles within the preset respiratory observation period corresponding to the target cycle are greater than or equal to the preset number.

[0048] In a preferred embodiment of the present invention, considering that the fluctuations of vascular compliance indicators and vascular elasticity loss indicators often occur synchronously during the respiratory cycle, the two features can be fused by calculating the Euclidean norm. However, since the vascular compliance indicators and vascular elasticity loss indicators differ significantly in order of magnitude and dimension, standardization is required to measure the discrete characteristics of the fused features during the respiratory observation period, thereby evaluating the vascular passive modulation index. Therefore, the method for obtaining the vascular passive modulation index includes: Within each cycle, the vascular compliance index and vascular elasticity loss index were standardized and then the Euclidean norm was calculated to obtain the vascular fusion characteristic index of the affected side. The standard deviation of the vascular fusion characteristic index of the affected side in different cycles within the preset respiratory observation period was used as the vascular passive modulation index.

[0049] As an example, the standardization method used is Z-score standardization. When using Z-score standardization, the sample used to calculate the mean and standard deviation is the data of all cycles within the preset respiratory observation period. This is a well-known technique and will not be elaborated further. The Euclidean norm of the standardized vascular compliance index and the standardized vascular elasticity loss index within the target cycle is further calculated, and the Euclidean norm is used as the vascular fusion characteristic index of the affected side. Furthermore, within the preset respiratory observation period (including the target cycle and several cycles within its historical periods), the standard deviation of the Euclidean norm corresponding to all cycles is calculated, and the standard deviation is used as the vascular passive modulation index.

[0050] Step S204: The characteristic parameters of vascular impedance on the affected side include vascular compliance index, vascular elasticity loss index, and vascular passive modulation index.

[0051] As an example, within each cycle, a characteristic triplet consisting of vascular compliance index, vascular elasticity loss index, and vascular passive modulation index is used as the characteristic parameter of vascular impedance on the affected side.

[0052] The nerve block effect evaluation module 103 is used to output a nerve block status signal at each time point based on the changes in the characteristic parameters of the vascular impedance on the affected side during the preset historical block monitoring period.

[0053] It should be noted that in one embodiment of the present invention, the nerve block effect determination is not performed in the early stage of anesthesia block, such as within 120 (heartbeat) cycles after the start of the anesthesia block operation, and the output signal is defaulted to "not effective"; the monitoring and determination stage is only started after 120 (heartbeat) cycles, and the nerve block status signal is output; taking any moment in the monitoring and determination stage, such as the current moment, as an example, the analysis and description are carried out to determine the nerve block status signal and output it.

[0054] It should be noted that the preset historical block monitoring period is 120 heartbeat cycles. For the 120th heartbeat cycle and all subsequent cycles, the preset historical block monitoring period is fixed within the stable baseline window before the anesthesia block procedure. That is, the preset historical block monitoring period is fixed from the 1st to the 120th heartbeat cycle in time sequence. For example: if the current time is the 120th heartbeat cycle, the preset historical block monitoring period is the 1st to 120th heartbeat cycles; if the current time is the 121st heartbeat cycle, the preset historical block monitoring period is also the 1st to 120th heartbeat cycles; if the current time is the 125th heartbeat cycle, the preset historical block monitoring period is also the 1st to 120th heartbeat cycles.

[0055] To measure the deviation characteristics of the affected vascular impedance characteristic parameters at each time point, this embodiment of the invention first assesses the individual baseline of the affected vascular impedance during the "general anesthesia but no block" stage based on the changes in the affected vascular impedance characteristic parameters during the preset initial block monitoring period. This prepares for subsequent analysis of whether there are significant abrupt deviations in the affected vascular impedance characteristic parameters at the current time point to assess the block effectiveness and output a nerve block status signal.

[0056] Based on this, in a preferred embodiment of the present invention, the method for outputting a nerve block state signal includes: The target indicator is selected from the following indicators: vascular compliance, vascular elasticity loss, and vascular passive modulation. At each time point, the benchmark for the target indicator is determined based on the distribution characteristics of the indicator values ​​in all cycles within the preset historical blockage monitoring period. The benchmark deviation parameter for each time point is evaluated based on the deviation of the indicator value under the target indicator from the benchmark within the cycle to which the target indicator belongs. At each time point, the nerve block status signal is output based on the benchmark deviation parameter of each indicator.

[0057] As an example, we first analyze three indicators in the characteristic parameters of vascular impedance on the affected side separately, taking the vascular compliance indicator as an example: At the current moment, the mean value of the vascular compliance indicator in all cycles of the preset historical block monitoring period is used as the comparison benchmark for the vascular compliance indicator; the vascular compliance indicator in the current cycle is used as the numerator, the comparison benchmark is used as the denominator, and the ratio is used as the benchmark deviation parameter for the vascular compliance indicator; the dimensionless benchmark deviation ratio can be obtained through the ratio method; similarly, the benchmark deviation parameters for the vascular elasticity loss indicator and the vascular passive modulation indicator can be obtained; further, the nerve block status signal is output based on the benchmark deviation parameters of each indicator.

[0058] In a preferred embodiment of the present invention, at each time point, when the baseline deviation parameters corresponding to the vascular compliance index, the vascular elasticity loss index, and the vascular passive modulation index are all greater than the corresponding preset thresholds, a valid signal of the nerve block state is output; otherwise, an invalid signal of the nerve block state is output.

[0059] As an example, the baseline deviation parameters of the vascular compliance index, vascular elasticity loss index, and vascular passive modulation index correspond to preset thresholds of 1.5, 2.0, and 2.5, respectively. These preset thresholds are determined based on the statistical characteristics of historical clinical data. Specifically, a set of confirmed cases with known nerve block efficacy and a set of control cases without efficacy are selected, and their characteristic deviation fold distributions are calculated. The optimal thresholds are determined using receiver operating characteristic (ROC) curve analysis, a well-known technique that will not be elaborated further. In this example, statistical verification shows that the system achieves optimal specificity in identifying nerve block efficacy when the preset thresholds are 1.5, 2.0, and 2.5, respectively.

[0060] The essence of nerve block is the severing of sympathetic nerve tension, which causes the overall phase shift of the pulse wave on the affected side relative to the healthy side on the time axis, resulting in an increase in vascular compliance indicators; it also causes the high-frequency characteristics of the pulse wave waveform to wear down, resulting in an increase in vascular elasticity loss indicators; and it also causes the pulse wave characteristics to fluctuate violently with the respiratory cycle, resulting in an increase in vascular passive modulation indicators. Through "AND" logic, when the baseline deviation parameters of the three indicators are all greater than the corresponding preset thresholds, a valid nerve block state signal is output and an early warning is issued; otherwise, an invalid nerve block state signal is output.

[0061] In summary, this invention first extracts the photoplethysmography (PPG) waves from the affected side of the patient's limb with nerve block and the unaffected side of the patient's limb with nerve block. Simultaneously, the two PPG waves are periodically divided, baseline-removed, and normalized to determine the standardized volume distribution sequence of the unaffected and affected sides within each period. Further, based on the differences between the standardized volume distribution sequences of the unaffected and affected sides within the same period, and combined with the discrete characteristics of the differences between the standardized volume distribution sequences of the unaffected and affected sides within different periods of a preset respiratory observation timeframe, the vascular impedance characteristic parameters of the affected side within each period are obtained. Finally, at each moment, based on the changes in the vascular impedance characteristic parameters of the affected side during a preset historical block monitoring period, a nerve block status signal is output. This invention introduces a reference from the unaffected side of the same individual, comparing and decoupling the impedance characteristics of the denervated blood vessels on the affected side under general anesthesia, thus improving the accuracy of regional anesthesia block effect assessment.

Claims

1. A regional anesthesia blockade effect evaluation system based on medical big data, characterized in that, The system includes: Dual-source index processing module: used to extract the photoplethysmography (PPG) waves of the affected side of the limb with nerve block and the PPG waves of the unaffected side of the limb; simultaneously divide the two PPG waves into periods, remove the baseline and normalize them, and determine the normalized volume distribution sequence of the healthy side and the normalized volume distribution sequence of the affected side in each period. Vascular impedance decoupling module: used to obtain the vascular impedance characteristic parameters of the affected side in each cycle based on the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in the same cycle, and combined with the discrete characteristics of the difference between the standardized volume distribution sequence of the healthy side and the standardized volume distribution sequence of the affected side in different cycles within a preset respiratory observation period. The nerve block effect evaluation module is used to output a nerve block status signal at each time point based on the changes in the characteristic parameters of the vascular impedance on the affected side during the preset historical block monitoring period.

2. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 1, characterized in that, The two photoplethysmograms are periodically divided simultaneously, including: The healthy side photoplethysmography (PPG) wave is subjected to trough detection, and the time interval between two adjacent trough points is taken as a period; the healthy side PPG wave and the affected side PPG wave are simultaneously divided, and the healthy side PPG wave sub-segment and the affected side PPG wave sub-segment are determined in each period.

3. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 1 or 2, characterized in that, The methods for obtaining the healthy side standardized volumetric distribution sequence and the affected side standardized volumetric distribution sequence include: For each photoplethysmography (PPG) wavelet segment on the healthy side, the minimum value is taken as the baseline value on the healthy side. Each value within the PPG wavelet is subtracted from the baseline value on the healthy side to obtain the baseline-removed volume sequence on the healthy side. Each value in the baseline-removed volume sequence on the healthy side is normalized to obtain the standardized volume distribution sequence on the healthy side. The same acquisition method is used to obtain the standardized volume distribution sequence on the affected side.

4. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 1, characterized in that, The method for obtaining the characteristic parameters of the affected side's vascular impedance includes: Within each cycle, vascular compliance indices are obtained based on the deviation of the standardized volume distribution sequence on the affected side relative to the standardized volume distribution sequence on the healthy side. Within each cycle, the vascular elasticity loss index is obtained based on the deviation of the instantaneous change characteristics of the standardized volume distribution sequence on the affected side relative to the instantaneous change characteristics of the standardized volume distribution sequence on the healthy side. In each cycle, the vascular compliance index and the vascular elasticity loss index within each cycle are fused, and the vascular passive modulation index is obtained based on the discrete characteristics of the fusion results in different cycles within a preset respiratory observation period. The characteristic parameters of vascular impedance on the affected side include vascular compliance indicators, vascular elasticity loss indicators, and vascular passive modulation indicators.

5. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 4, characterized in that, Methods for obtaining vascular compliance indicators include: Within each cycle, the bulldozer distance between the standardized volume distribution sequence on the affected side and the standardized volume distribution sequence on the healthy side is used as an indicator of vascular compliance.

6. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 4, characterized in that, The method for obtaining the vascular elasticity loss index includes: Within each cycle, the normalized volume distribution sequence of the healthy side and the normalized volume distribution sequence of the affected side are differentiated and normalized to obtain the volume change distribution sequence of the healthy side and the volume change distribution sequence of the affected side; the bulldozer distance between the volume change distribution sequence of the affected side and the volume change distribution sequence of the healthy side is used as an indicator of vascular elasticity loss.

7. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 4, characterized in that, The method for obtaining the vascular passive modulation index includes: Within each cycle, the vascular compliance index and the vascular elasticity loss index are standardized and then the Euclidean norm is calculated to obtain the vascular fusion characteristic index of the affected side; the standard deviation of the vascular fusion characteristic index of the affected side in different cycles within the preset respiratory observation period is used as the vascular passive modulation index.

8. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 4, characterized in that, The method for outputting the nerve block state signal includes: The target indicator is selected from the vascular compliance index, the vascular elasticity loss index, and the vascular passive modulation index. At each time point, the comparison benchmark of the target indicator is determined based on the distribution characteristics of the indicator values ​​in all cycles under the target indicator within the preset historical blockage monitoring period. The benchmark deviation parameter of the target indicator at each time point is evaluated based on the deviation of the indicator value under the target indicator in the cycle to which each time point belongs from the comparison benchmark. At each time point, a neural blockade state signal is output based on the baseline deviation parameter for each metric.

9. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 8, characterized in that, The baseline deviation parameter for each indicator outputs a neural blockade state signal, including: At each time point, when the baseline deviation parameters corresponding to the vascular compliance index, the vascular elasticity loss index, and the vascular passive modulation index are all greater than the corresponding preset thresholds, a valid signal for the nerve block state is output; otherwise, an invalid signal for the nerve block state is output.

10. The regional anesthesia blockade effect evaluation system based on medical big data according to claim 1, characterized in that, The preset respiratory observation period includes a preset number of respiratory cycles.

Images