Acupuncture sensation data collection and control system based on artificial intelligence

Abstract

The present invention relates to the field of data processing technology, and in particular to an acupuncture needle sensation data acquisition and control system based on artificial intelligence, the system comprising a data processing module, used to divide the collected skin electrical signal data sequence into a preset number of subsequences; a data analysis module, used to set different sliding windows for each subsequence respectively, and obtain a target subsequence and a target sliding window; a filtering processing module, used to use a non-target sliding window to obtain a filtered subsequence corresponding to each subsequence except the target subsequence, and adaptively adjust the target sliding window of the target subsequence to obtain the corresponding filtered subsequence; a parameter control module, used to form a target smooth data sequence from all the filtered subsequences, thereby ensuring the denoising effect of filtering the skin electrical signal data and improving the accuracy of subsequent acupuncture parameter adjustment based on the target smooth data sequence.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an artificial intelligence-based acupuncture sensation data acquisition and control system. Background Art

[0002] Acupuncture is an important component of traditional Chinese medicine. It involves acupuncture or moxibustion at specific acupuncture points on the human body to regulate qi and blood, unclog the meridians, and prevent and treat disease. Acupuncture involves inserting fine needles into the patient's body at specific acupuncture points, using techniques such as twisting and lifting to treat disease. Acupuncture techniques are the core technology in acupuncture treatment. They involve a series of operations using specially designed needles to stimulate specific acupuncture points on the body. Different acupuncture techniques can produce different biophysical and biochemical changes. Acupuncture needle sensation, also known as deqi (qi sensation), refers to the sensation of soreness, numbness, swelling, heaviness, etc. at the acupuncture site during acupuncture, sometimes accompanied by a sense of electric current or other special sensations. The occurrence of needle sensation indicates that the acupuncture stimulation has triggered bioelectrical activity and qi and blood flow within the body, which has positive implications for the treatment of disease.

[0003] In the existing technology, smart acupuncture equipment is used to automatically identify the patient's acupoints, and a personalized acupuncture treatment plan is formulated according to the patient's symptoms and historical treatment responses. The doctor controls the smart acupuncture equipment through a tablet system, so that it can accurately perform the processes of needle insertion, needle movement and needle removal, thereby achieving acupuncture treatment for the patient. Specifically, the patient wears wearable acupuncture equipment such as smart bracelets and patches, and the doctor uses the tablet system to control the smart acupuncture equipment to perform acupuncture on the patient. Taking the skin electrical signal data as an example, the skin electrical signal data generated by the patient during the acupuncture process is collected. According to the skin electrical signal data generated by the patient, the acupuncture parameters such as acupuncture angle, acupuncture depth, and acupuncture speed in the smart acupuncture equipment are updated and adjusted, so that it can provide more accurate auxiliary acupuncture for the patient, and automatically record the acupuncture data, thereby continuously improving the system. Since the generated skin electrical signal data is easily affected by machine interference, electromyographic interference, electromagnetic interference, etc., direct analysis of the skin electrical signal data will be inaccurate, which will affect the subsequent adjustment of acupuncture parameters. Therefore, the collected skin electrical signal data is processed by Savitzky-Golay filtering (SG filtering). However, when using a fixed-size sliding window for SG filtering, if the sliding window is set too large, the skin electrical signal data in the window will be over-smoothed, resulting in loss of details in the skin electrical signal data. If the sliding window is set too small, the smoothing effect of the skin electrical signal data will be reduced, making the denoised skin electrical signal data less effective, thereby affecting the subsequent adjustment of the acupuncture parameters in the intelligent acupuncture device.

[0004] Therefore, how to ensure the denoising effect of SG filtering on skin electrical signal data to improve the accuracy of subsequent acupuncture parameter adjustment has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, an embodiment of the present invention provides an artificial intelligence-based acupuncture needle sensation data acquisition and control system to solve the problem of how to ensure the denoising effect of SG filtering processing on skin electrical signal data to improve the accuracy of subsequent acupuncture parameter adjustment.

[0006] The present invention provides an artificial intelligence-based acupuncture sensation data acquisition and control system, which includes the following modules:

[0007] a data processing module for obtaining a skin electrical signal data sequence within a preset time period during acupuncture treatment on a patient using the intelligent acupuncture device, and dividing the skin electrical signal data sequence into a preset number of subsequences based on data differences in the skin electrical signal data sequence;

[0008] a data analysis module configured to set a sliding window for filtering each subsequence based on data fluctuations in each subsequence, and to obtain a subsequence and sliding window corresponding to a maximum length as a target subsequence and a target sliding window based on the length of the sliding window corresponding to each subsequence;

[0009] a filtering processing module, configured to perform filtering processing on each subsequence other than the target subsequence using a non-target sliding window to obtain a corresponding filtered subsequence; and in the process of filtering the target subsequence using the target sliding window, adaptively adjusting the target sliding window according to data features in the target sliding window to obtain a filtered subsequence;

[0010] The parameter control module is used to combine all filtered subsequences into a target smoothed data sequence, and to perform real-time control on the preset acupuncture parameters in the intelligent acupuncture device according to the target smoothed data sequence.

[0011] Preferably, the data processing module divides the skin electrical signal data sequence into a preset number of subsequences according to data differences in the skin electrical signal data sequence, including:

[0012] For any data except the last data in the skin electrical signal data sequence, obtaining an absolute value of a difference between the data and the next data after the data, and using the absolute value of the difference as a first change degree of the data;

[0013] obtaining a preset target number of data after the data, forming a target data sequence with the data, and obtaining a standard deviation of the target data sequence as a second degree of change of the data;

[0014] respectively acquiring acupuncture start data and acupuncture end data in the skin electrical signal data sequence according to a first change degree and a second change degree of each data except the last data in the skin electrical signal data sequence;

[0015] The acupuncture start data and the acupuncture end data are used as truncation points to divide the skin electrical signal data sequence into a preset number of subsequences.

[0016] Preferably, the data processing module obtains the acupuncture start data and the acupuncture end data in the skin electrical signal data sequence according to the first change degree and the second change degree of each data except the last data in the skin electrical signal data sequence, including:

[0017] According to the first change degree of each data, at least one maximum first change degree is obtained to form a maximum data sequence, and data corresponding to the first maximum first change degree in the maximum data sequence is used as acupuncture start data in the skin electrical signal data sequence;

[0018] Perform weighted summation on the first degree of change and the second degree of change of the data to obtain the overall degree of change of the data. According to the overall degree of change of each data, obtain at least one minimum overall degree of change to form a minimum data sequence. The data corresponding to the first minimum overall degree of change in the minimum data sequence is used as the acupuncture end data in the skin electrical signal data sequence.

[0019] Preferably, in the data analysis module, a sliding window for filtering is set for each subsequence according to data fluctuations in each subsequence, including:

[0020] The variance of each subsequence is obtained respectively, and each variance is used as the fluctuation degree of the corresponding subsequence. According to the fluctuation degree of each subsequence, the length of the sliding window corresponding to the filtering process is set for each subsequence, wherein the greater the fluctuation degree of the subsequence, the larger the length of the sliding window corresponding to the filtering process is set.

[0021] Preferably, the filtering processing module adaptively adjusts the target sliding window according to the data characteristics in the target sliding window to obtain a filtered subsequence, including:

[0022] According to the target sliding window, an initial sliding window of the first data is obtained in the target subsequence, wherein the length of the initial sliding window is equal to the length of the target sliding window;

[0023] According to the data features in the initial sliding window of the first data, an adaptive sliding window of the first data is obtained. The filtered denoised data corresponding to the first data is obtained using the adaptive sliding window. According to the length of the adaptive sliding window, an initial sliding window of the second data in the target subsequence is obtained. The second data is used as the first data. According to the method for obtaining the filtered denoised data corresponding to the first data, the filtered denoised data corresponding to the second data is obtained. Until the last data in the target subsequence is traversed, the filtered denoised data corresponding to each data in the target subsequence is obtained, and all the filtered denoised data are combined into a filtered subsequence corresponding to the target subsequence.

[0024] Preferably, the filtering processing module obtains the adaptive sliding window of the first data according to the data features in the initial sliding window of the first data, including:

[0025] Obtaining a data chaos degree corresponding to the initial sliding window of the first data according to a data chaos feature in the initial sliding window of the first data;

[0026] Constructing a target data change curve of the target subsequence in a two-dimensional plane, wherein the horizontal axis of the two-dimensional plane represents acupuncture time and the vertical axis represents skin electrical signal data value, respectively obtaining the curvature of the curve between each two adjacent data in the target data change curve, and obtaining the degree of data non-smoothness corresponding to the initial sliding window of the first data based on the curvature of the curve between each two adjacent data in the initial sliding window of the first data;

[0027] Performing a weighted summation on the data chaos degree and the data non-smoothness degree to obtain a window adjustment degree corresponding to the initial sliding window of the first data;

[0028] Obtain comparison results between the window adjustment degree and a preset first adjustment threshold and a preset second adjustment threshold, and obtain an adaptive sliding window of the first data based on the comparison results, wherein the first adjustment threshold is less than the second adjustment threshold.

[0029] Preferably, the filtering processing module obtains the data disorder degree corresponding to the initial sliding window of the first data according to the data disorder feature in the initial sliding window of the first data, including:

[0030] According to the data in the initial sliding window of the first data, respectively obtaining data information entropy and data standard deviation, performing normalization processing on the data information entropy to obtain a first normalization result, and performing normalization processing on the data standard deviation to obtain a second normalization result;

[0031] A weighted sum is performed on the first normalization result and the second normalization result to obtain a data disorder degree corresponding to the initial sliding window of the first data.

[0032] Preferably, the filtering processing module obtains the degree of data non-smoothness corresponding to the initial sliding window of the first data according to the curvature of the curve between every two adjacent data in the initial sliding window of the first data, including:

[0033] The curvatures of the curves between every two adjacent data in the initial sliding window of the first data are combined into a curvature data sequence, and the absolute value of the curvature difference between every two adjacent curve curvatures in the curvature data sequence is obtained to obtain the average absolute value of the curvature difference;

[0034] Normalizing the absolute value of the mean curvature difference to obtain a third normalized result, and using the third normalized result as the data non-smoothness corresponding to the initial sliding window of the first data.

[0035] Preferably, the filtering processing module obtains the adaptive sliding window of the first data according to the comparison result, including:

[0036] If the comparison result shows that the window adjustment degree is less than the first adjustment threshold, reducing the length of the initial sliding window of the first data to obtain a reduced length, and using the reduced length as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data;

[0037] If the comparison result shows that the window adjustment degree is greater than or equal to the first adjustment threshold, and the window adjustment degree is less than or equal to the second adjustment threshold, using the length of the initial sliding window of the first data as the length of the adaptive sliding window of the first data, to obtain the adaptive sliding window of the first data;

[0038] If the comparison result is that the window adjustment degree is greater than the second adjustment threshold, the length of the initial sliding window of the first data is increased to obtain the increased length, and the increased length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

[0039] Preferably, the filtering processing module obtains an initial sliding window of the second data in the target subsequence according to the length of the adaptive sliding window, including:

[0040] The length of the adaptive sliding window of the first data is used as the length of the initial sliding window of the second data in the target subsequence to obtain the initial sliding window of the second data.

[0041] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0042] The present invention provides an artificial intelligence-based acupuncture sensation data acquisition and control system, comprising a data acquisition module for acquiring a skin electrical signal data sequence within a preset time period during acupuncture treatment on a patient using an intelligent acupuncture device, and dividing the skin electrical signal data sequence into a preset number of subsequences based on data differences in the skin electrical signal data sequence; a data division module for setting a sliding window for filtering processing for each subsequence based on data fluctuations in each subsequence, acquiring a subsequence and sliding window corresponding to a maximum length as a target subsequence and a target sliding window based on the length of the sliding window corresponding to each subsequence, and filtering each subsequence other than the target subsequence using a non-target sliding window to obtain a corresponding filtered subsequence; a window adjustment module for adaptively adjusting the target sliding window based on data features in the target sliding window during filtering processing of the target subsequence using the target sliding window to obtain a filtered subsequence; and a parameter control module for combining all filtered subsequences into a target smoothed data sequence, and performing real-time control of preset acupuncture parameters in the intelligent acupuncture device based on the target smoothed data sequence. Among them, the collected skin electrical signal data sequence is divided into a preset number of subsequences, and a corresponding sliding window is set for each subsequence according to the fluctuation degree corresponding to each subsequence. The greater the fluctuation degree, the longer the corresponding sliding window. Then, when each subsequence is subjected to SG filtering processing, the sliding window of the subsequence with a small fluctuation degree is kept unchanged, and the sliding window of the subsequence with a large fluctuation degree is adaptively adjusted, and the corresponding sequence after filtering the skin electrical signal data sequence is obtained, that is, the target smoothed data sequence, thereby eliminating the loss of skin electrical signal data details in the sliding window or the decrease in denoising effect due to the use of a fixed-length sliding window, ensuring the denoising effect when filtering the skin electrical signal data sequence, and improving the accuracy of subsequent acupuncture parameter adjustment based on the target smoothed data sequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 This is a structural block diagram of an artificial intelligence-based acupuncture sensation data acquisition and control system provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0045] The embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.

[0046] It should be noted that the terms "first," "second," and the like in the specification of the present disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0047] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for description.

[0048] See also Figure 1 , is a structural block diagram of an artificial intelligence-based acupuncture sensation data acquisition and control system provided in the first embodiment of the present invention, such as Figure 1 As shown, the system may include:

[0049] The data processing module 11 is used to obtain a skin electrical signal data sequence within a preset time period during the process of performing acupuncture on a patient using the intelligent acupuncture device, and divide the skin electrical signal data sequence into a preset number of subsequences based on the data differences in the skin electrical signal data sequence.

[0050] Electrodermal activity (EDA), also known as galvanic skin reflex, is a characteristic of the human body and a physiological or psychological state caused by external stimuli. It uses an EDA sensor to measure the conductivity, current flow, or current level change between two selected points on the skin (usually between two fingers, but also between two palms or between two feet). When a person is stimulated and in a state of intense passion (such as anger), the instantaneous and large fluctuations generated are the galvanic skin response, which is an accurate indicator of emotion. Therefore, in an embodiment of the present invention, taking skin electrode signal data as an example, first, an intelligent acupuncture device (such as an intelligent acupuncture robot) is used to automatically identify the patient's acupoints, and an EDA sensor is placed on the patient's finger. The doctor controls the robotic arm of the intelligent acupuncture robot through a smart tablet system to perform acupuncture on the patient, thereby collecting the skin electrode signal data sequence generated by the patient during a complete acupuncture treatment course. A complete acupuncture treatment course includes the acupuncture preparation period before needle insertion, the acupuncture treatment period (including the three steps of needle insertion, needle movement, and needle removal), and the acupuncture end period after needle removal.

[0051] It should be noted that the time period set in the embodiment of the present invention is the time required for a complete acupuncture treatment course, which is not limited and can be set according to the specific real-time scenario.

[0052] After obtaining a skin electrical signal data sequence within a complete acupuncture treatment course, the existing technology uses Savitzky-Golay filtering (SG filtering) processing technology to directly filter the collected skin electrical signal data sequence. However, since the degree of fluctuation of the skin electrical signal data in different acupuncture periods in the skin electrical signal data sequence is different, when using a fixed-size sliding window to perform SG filtering on the skin electrical signal data sequence, if the sliding window is set too large, the skin electrical signal data in the sliding window will be over-smoothed, resulting in the loss of some skin electrical signal data details in the skin electrical signal data sequence. If the sliding window is set too small, the denoising effect of the skin electrical signal data in the sliding window will be reduced, thereby affecting the accuracy of the subsequent acupuncture parameter adjustment in the intelligent acupuncture device.

[0053] To solve the above problems, in an embodiment of the present invention, first, the collected skin electrical signal data sequence is divided into corresponding subsequences according to different acupuncture periods, wherein one acupuncture period corresponds to one subsequence. Then, according to the data fluctuation in each subsequence, a sliding window of different lengths for filtering processing is set for each subsequence, thereby eliminating the loss of some data details or the decrease in denoising effect due to the use of a fixed sliding window, improving the denoising effect of each subsequence, and thereby improving the overall denoising effect of the skin electrical signal data sequence.

[0054] Considering that the data characteristics during the acupuncture treatment period are significantly different from those of the other two periods, and the skin electrical signal data during the acupuncture preparation period will remain relatively stable, when the acupuncture treatment period occurs (i.e., when the needle is inserted), the skin electrical signal data will fluctuate greatly, that is, the skin electrical signal data will show an instantaneous growth trend. When the acupuncture end period occurs (i.e., after the needle is removed), the skin electrical signal data will gradually tend to be stable and remain relatively stable for a period of time thereafter. Therefore, based on the change characteristics of the skin electrical signal data in each acupuncture period, the subsequence corresponding to each acupuncture period is identified and divided. Therefore, in the embodiment of the present invention, the specific method of dividing the skin electrical signal data sequence into multiple subsequences is as follows:

[0055] (1) For any data except the last data in the skin electrical signal data sequence, obtain the absolute value of the difference between the data and the next data after the data, and use the absolute value of the difference as the first change degree of the data;

[0056] obtaining a preset target number of data after the data, forming a target data sequence with the data, and obtaining a standard deviation of the target data sequence as a second degree of change of the data;

[0057] In one embodiment, considering that the end period of acupuncture will not contain only one data, that is, the end data of acupuncture will not be the last data in the skin electrical signal data sequence, the first change degree and the second change degree of the last data in the skin electrical signal data sequence are not calculated.

[0058] Then, for the i-th data except the last data in the skin electrical signal data sequence, the absolute value of the difference between the i-th data and the (i+1)-th data is obtained, and the absolute value of the difference is used as the first change degree of the i-th data.

[0059] Assuming that the number of data in the skin electrical signal data sequence is N, if i is N-3, the i-th data and the two data after the i-th data are composed of the target data sequence, and the standard deviation of the target data sequence is obtained as the second degree of change of the i-th data, that is, the preset target number is 2. If i is N-2, the i-th data and the one data after the i-th data are composed of the target data sequence, and the standard deviation of the target data sequence is obtained as the second degree of change of the i-th data, that is, the preset target number is 1.

[0060] If i is not N-2 or N-3, then the three data after the i-th data are obtained in the skin electrical signal data sequence, and the target data sequence is formed together with the i-th data. The standard deviation of the target data sequence is obtained as the second degree of change of the i-th data, that is, the preset target number set in the embodiment of the present invention is 3, which is not limited and can be set according to the specific scenario.

[0061] (2) According to the first change degree and the second change degree of each data, respectively obtain the acupuncture start data and the acupuncture end data in the skin electrical signal data sequence.

[0062] Specifically, according to the first change degree of each data, at least one maximum first change degree is obtained to form a maximum data sequence, and the data corresponding to the first maximum first change degree in the maximum data sequence is used as the acupuncture start data in the skin electrical signal data sequence;

[0063] Perform weighted summation on the first degree of change and the second degree of change of the data to obtain the overall degree of change of the data. According to the overall degree of change of each data, obtain at least one minimum overall degree of change to form a minimum data sequence. Obtain data corresponding to the first minimum overall degree of change in the minimum data sequence as the acupuncture end data in the skin electrical signal data sequence.

[0064] In one embodiment, taking the i-th data as an example, based on the first change degree and the second change degree of the i-th data, the expression for obtaining the overall change degree of the i-th data is:

[0065]

[0066] in, Indicates the overall change degree of the i-th data, Indicates the first degree of change of the i-th data, Indicates the second degree of change of the i-th data, represents the weight of the first degree of change, The weight representing the second degree of change.

[0067] It should be noted that The smaller the value of , the smaller the difference between the i-th data and the i+1-th data in the skin electrical signal data sequence, that is, the less likely the i-th data is to undergo a mutation, indicating a greater possibility of the end of acupuncture, and the smaller the overall change of the corresponding i-th data; The smaller the value of , the smaller the data difference between the i-th data and the subsequent data in the skin electrical signal data sequence, that is, the smaller the degree of fluctuation, indicating that the possibility that the i-th data is the end of acupuncture is greater, and the corresponding overall change degree of the i-th data is smaller.

[0068] Preferably, the weight of the first degree of change and the weight of the second degree of change set in the embodiment of the present invention are empirical values, that is, There is no restriction on this, and it can be set according to the specific real-time scenario.

[0069] According to the above method for obtaining the first change degree and the overall change degree of the i-th data, the first change degree and the overall change degree of each data except the last data in the skin electrical signal data sequence are obtained.

[0070] Since the skin electrical signal data during the acupuncture preparation period will remain in a relatively stable state, the corresponding degree of change is small. When the acupuncture treatment period occurs, the skin electrical signal data will show an instantaneous increase in change at the moment of needle insertion. After the needle is inserted, due to various needle manipulations, the skin electrical signal data will also show an instantaneous increase in change at a certain moment. That is, the degree of change of the skin electrical signal data during the acupuncture treatment period is relatively high. Therefore, according to the first change degree of each data except the last data in the skin electrical signal data sequence, the first data of the acupuncture treatment period is obtained. Specifically, according to the first change degree of each data except the last data, the maximum first change degree is obtained, and the data corresponding to the maximum first change degree is used as the acupuncture start data in the skin electrical signal data sequence. If there are multiple maximum first change degrees, the data corresponding to the maximum first change degree are arranged in time series to obtain the maximum data sequence, and the data corresponding to the minimum moment is selected as the acupuncture start data in the skin electrical signal data sequence, that is, the data corresponding to the first maximum first change degree in the maximum data sequence is selected as the acupuncture start data in the skin electrical signal data sequence.

[0071] Since the skin electrical signal data gradually tends to be stable after the needle is removed and remains in a relatively stable state for a period of time thereafter, and the overall change degree of each data in the end period of acupuncture is small, the first data of the end period of acupuncture is obtained according to the overall change degree of each data except the last data in the skin electrical signal data sequence. Specifically, the minimum overall change degree is obtained according to the overall change degree of each data except the last data, and the data corresponding to the minimum overall change degree is used as the acupuncture end data in the skin electrical signal data sequence. If there are multiple minimum overall change degrees, the data corresponding to the minimum overall change degree are arranged in time series to obtain the minimum data sequence, and the data corresponding to the minimum moment is selected as the acupuncture end data in the skin electrical signal data sequence, that is, the data corresponding to the first minimum overall change degree in the minimum data sequence is selected as the acupuncture end data in the skin electrical signal data sequence.

[0072] (2) Using the acupuncture start data and the acupuncture end data as truncation points, the skin electrical signal data sequence is divided into a preset number of subsequences.

[0073] The acquired acupuncture start data and acupuncture end data are used as truncation points in the skin electrical signal data sequence. The skin electrical signal data sequence is divided into three subsequences using the two truncation points, wherein one acupuncture period corresponds to one subsequence.

[0074] The data analysis module 12 is configured to set a sliding window for filtering each subsequence based on the data fluctuations in each subsequence, and obtain the subsequence and sliding window corresponding to the maximum length as the target subsequence and target sliding window based on the length of the sliding window corresponding to each subsequence.

[0075] After obtaining the subsequence corresponding to each acupuncture period, considering that the fluctuation degree of the subsequences corresponding to different acupuncture periods is different, for the subsequence with a large fluctuation degree, when filtering the skin electrical signal data within its sliding window, it is necessary to combine more skin electrical signal data before and after, and the corresponding sliding window length should be larger. For the subsequence with a small fluctuation degree, when filtering the skin electrical signal data within its sliding window, it is not necessary to have too much skin electrical signal data, and the corresponding sliding window length should be smaller. Therefore, the same size sliding window cannot be directly used to filter the three subsequences, which is likely to result in poor filtering effect. Therefore, in the embodiment of the present invention, a sliding window for filtering is set for each subsequence according to the data fluctuation of each subsequence, including:

[0076] The variance of each subsequence is obtained respectively, and each variance is used as the fluctuation degree of the corresponding subsequence. According to the fluctuation degree of each subsequence, the length of the sliding window for filtering processing is set for each subsequence respectively, wherein the greater the fluctuation degree of the subsequence, the larger the length of the sliding window for filtering processing is set.

[0077] In one embodiment, assuming that the subsequences corresponding to the acupuncture preparation period, the acupuncture treatment period, and the acupuncture end period are A1, A2, and A3, respectively, the variance of each subsequence is calculated as the fluctuation degree of the corresponding subsequence, and the fluctuation degree of each subsequence is recorded as S1, S2, and S3, respectively. The lengths of the corresponding sliding windows set for filtering processing are L1, L2, and L3, respectively. If the fluctuation degrees of the three subsequences are in the order of S2>S1>S3, then the lengths of the corresponding sliding windows set for filtering processing are in the order of L2>L1>L3.

[0078] It should be noted that the length of the sliding window used for filtering processing set for each subsequence in the embodiment of the present invention is an empirical value, that is, L1=5, L2=7, and L3=3. There is no limitation on this and it can be set according to specific scenarios.

[0079] After obtaining the sliding window corresponding to each subsequence, considering that the fluctuation degree of subsequence S1 during the acupuncture preparation period and subsequence S3 during the acupuncture end period is relatively small, that is, the skin electrode signal data contained in S1 and S3 have less details and lower analysis value, if L1 and L3 are further adjusted, the denoising effect of the corresponding filtered subsequence will not be very beneficial and the calculation amount will increase. Therefore, L1 and L3 are kept unchanged.

[0080] Furthermore, considering that the fluctuation of subsequence S2 during the acupuncture treatment period is too large, that is, the skin electrode signal data contained in S2 contains more details, if L2 is directly used for filtering, the denoising effect of the subsequence corresponding to S2 after filtering cannot be guaranteed. In order to eliminate the loss of some data details or the decline in denoising effect caused by the use of a fixed sliding window, thereby improving the denoising effect of L2, the sliding window L2 of subsequence S2 corresponding to the acupuncture treatment period is adaptively adjusted to improve the filtering effect of subsequence S2. The specific adjustment method is described below. It is worth noting that, for the convenience of clear description below, the subsequence S2 corresponding to the acupuncture treatment period is used as the target subsequence, the sliding window L2 corresponding to subsequence S2 is used as the target sliding window, and the sliding windows corresponding to subsequence S1 during the acupuncture preparation period and subsequence S3 during the acupuncture end period are used as non-target sliding windows.

[0081] The filtering processing module 13 is used to use the non-target sliding window to filter each subsequence except the target subsequence to obtain the corresponding filtered subsequence. In the process of filtering the target subsequence using the target sliding window, the target sliding window is adaptively adjusted according to the data characteristics in the target sliding window to obtain the filtered subsequence.

[0082] When performing SG filtering on each subsequence, it is considered that the fluctuation degree of the subsequence S1 in the acupuncture preparation period and the subsequence S3 in the acupuncture end period is relatively small, that is, the skin electrical signal data contained in S1 and S3 have fewer details and lower analysis value. Therefore, for S1 and S3, the corresponding non-target sliding windows (L1 and L3) are used to perform SG filtering on S1 and S3 respectively to obtain the corresponding filtered subsequences, thereby ensuring the denoising effect of S1 and S3 while reducing the amount of calculation. It is worth noting that the SG filtering processing technology is an existing technology and will not be repeated here.

[0083] Furthermore, considering that the fluctuation of S2 is too large, that is, the skin electrode signal data contained in S2 has more details, in order to further improve the denoising effect of S2, therefore, in the embodiment of the present invention, when using the target sliding window to filter the target subsequence, the target sliding window is adaptively adjusted according to the data characteristics in the target sliding window to obtain the filtered subsequence. The specific process of obtaining the filtered subsequence corresponding to subsequence S2 is as follows:

[0084] According to the target sliding window, an initial sliding window of the first data is obtained in the target subsequence. It is worth noting that the length of the target sliding window is equal to the length of the initial sliding window of the first data.

[0085] According to the data characteristics in the initial sliding window of the first data, an adaptive sliding window of the first data is obtained. Using the adaptive sliding window of the first data, SG filtering is performed on all data in the adaptive sliding window of the first data to obtain filtered denoised data corresponding to the first data. The length of the adaptive sliding window of the first data is used as the length of the initial sliding window of the second data to obtain the initial sliding window of the second data. Similarly, according to the method for obtaining the filtered denoised data corresponding to the first data, the adaptive sliding window of the second data and the filtered denoised data corresponding to the second data are obtained. The length of the adaptive sliding window of the second data is used as the length of the initial sliding window of the third data to obtain the initial sliding window of the third data. Then, according to the method for obtaining the filtered denoised data corresponding to the first data, the adaptive sliding window of the third data and the filtered denoised data corresponding to the third data are obtained. And so on, until the filtered denoised data corresponding to the last data in the target subsequence is obtained, and all the filtered denoised data are combined into a filtered subsequence corresponding to the target subsequence.

[0086] It should be noted that the SG filtering processing technology is an existing technology and will not be described in detail here.

[0087] In the process of obtaining the adaptive sliding window of the first data, considering that noise data has a great influence on the adjustment of the sliding window, when performing SG filtering on the target subsequence, if there is more noise data in the initial sliding window, some data details may be over-smoothed, thereby affecting the accuracy of the filtering result. In this case, the length of the initial sliding window needs to be increased to weaken the influence of noise points on the data in the initial sliding window. If there is less noise data in the initial sliding window, the length of the initial sliding window can be appropriately reduced to amplify the data details in the initial sliding window. Therefore, the initial sliding window is adaptively adjusted according to the degree of chaos and non-smoothness of the data in the initial sliding window. Therefore, according to the data characteristics in the initial sliding window of the first data, the adaptive sliding window of the first data is obtained, including:

[0088] (1) According to the data chaos feature in the initial sliding window of the first data, the data chaos degree corresponding to the initial sliding window of the first data is obtained.

[0089] Specifically, according to the data in the initial sliding window of the first data, data information entropy and data standard deviation are respectively obtained, the data information entropy is normalized to obtain a first normalized result, and the data standard deviation is normalized to obtain a second normalized result;

[0090] A weighted sum is performed on the first normalization result and the second normalization result to obtain a data disorder degree corresponding to the initial sliding window of the first data.

[0091] In one embodiment, the expression for obtaining the data chaos degree corresponding to the initial sliding window of the first data is:

[0092]

[0093] in, It represents the degree of data confusion corresponding to the initial sliding window of the first data, and H represents the data information entropy corresponding to the initial sliding window of the first data. Indicates the data standard deviation corresponding to the initial sliding window of the first data, represents the weight of data information entropy, It represents the weight of the data standard deviation, and norm() represents the normalization function.

[0094] It should be noted that the larger the value of H is, the more chaotic the data in the initial sliding window of the first data is, indicating that the possibility of more noise data in the initial sliding window is greater, and the degree of data chaos corresponding to the initial sliding window of the first data is greater; The larger the value of , the more discrete the data in the initial sliding window of the first data is, that is, the greater the degree of discreteness of the data, the greater the degree of chaos of the corresponding data.

[0095] Preferably, the weight of the data information entropy and the weight of the data standard deviation set in the embodiment of the present invention are empirical values, that is, There is no restriction on this, and it can be set according to the specific real-time scenario.

[0096] (2) Constructing a target data change curve of the target subsequence in a two-dimensional plane, wherein the horizontal axis of the two-dimensional plane is the acupuncture time and the vertical axis is the skin electrical signal data value, respectively obtaining the curve curvature between each two adjacent data in the target data change curve, and obtaining the data non-smoothness corresponding to the initial sliding window of the first data based on the curve curvature between each two adjacent data in the initial sliding window of the first data.

[0097] Specifically, the curvatures of the curves between every two adjacent data in the initial sliding window of the first data are combined into a curvature data sequence, and the absolute value of the curvature difference between every two adjacent curve curvatures in the curvature data sequence is obtained to obtain the average absolute value of the curvature difference;

[0098] Normalizing the absolute value of the mean curvature difference to obtain a third normalized result, and using the third normalized result as the data non-smoothness corresponding to the initial sliding window of the first data.

[0099] In one embodiment, after obtaining the data chaos degree corresponding to the initial sliding window of the first data, considering that more chaotic the data in the initial sliding window does not necessarily mean that there is more noise data content therein, it may also be caused by the characteristics of the data itself. Therefore, the non-smoothness of the data in the first initial sliding window is used to perform the next step of evaluating the noise data content of the data in the initial sliding window. The less smooth the data is, the higher the noise level of the data in the initial sliding window is.

[0100] A target data change curve of the target subsequence is constructed in a two-dimensional plane, wherein the horizontal axis of the two-dimensional plane is the acupuncture time and the vertical axis is the skin electrode signal data value. The curvature of the curve between each two adjacent data in the target data change curve is obtained respectively. The curvature of the curve of each two adjacent data in the initial sliding window of the first data is combined into a curvature data sequence. According to the absolute value of the difference between the curvatures of each two adjacent curves in the curvature data sequence, the data non-smoothness corresponding to the initial sliding window of the first data is obtained. The expression of the data non-smoothness corresponding to the initial sliding window of the first data is:

[0101]

[0102] in, Indicates the degree of non-smoothness of the data corresponding to the initial sliding window of the first data, represents the curvature of the jth curve in the curvature data sequence, represents the curvature of the j+1th curve in the curvature data sequence, I represents the total number of curve curvatures in the curvature data sequence, || represents the absolute value, and norm() represents the normalization function.

[0103] It should be noted that The larger the value of , the more significant the change in the curvature of the jth curve and the j+1th curve in the curvature data sequence. That is, the greater the curvature of the data change curve at the data point corresponding to the curvature of the jth curve, indicating that the data in the initial sliding window of the first data has changed significantly. The larger the value of , the higher the data noise level in the initial sliding window, and the greater the degree of non-smoothness of the data corresponding to the initial sliding window of the first data.

[0104] (3) Performing a weighted summation of the data chaos degree and the data non-smoothness degree to obtain a window adjustment degree corresponding to the initial sliding window of the first data.

[0105] In one embodiment, after obtaining the data chaos degree and data non-smoothness degree corresponding to the initial sliding window of the first data, considering that the higher the data chaos degree and the greater the data non-smoothness degree in the initial sliding window, the higher the data noise content in the initial sliding window, the more necessary it is to enlarge the initial sliding window to weaken the influence of noise points on the data in the initial sliding window. Conversely, the lower the data chaos degree and the smaller the data non-smoothness degree in the initial sliding window, the less the data noise content in the initial sliding window, and the initial sliding window can be appropriately reduced to amplify the details of the data in the initial sliding window. Therefore, according to the data chaos degree and the data non-smoothness degree, the window adjustment degree corresponding to the initial sliding window of the first data is obtained. The expression for obtaining the window adjustment degree corresponding to the initial sliding window of the first data is:

[0106]

[0107] in, Indicates the window adjustment degree corresponding to the initial sliding window of the first data, Indicates the degree of data chaos corresponding to the initial sliding window of the first data, Indicates the degree of non-smoothness of the data corresponding to the initial sliding window of the first data, The weight representing the degree of data chaos, A weight indicating the degree of non-smoothness of the data.

[0108] It should be noted that The larger the value is, the more chaotic the data in the initial sliding window of the first data is, that is, the greater the degree of fluctuation is, indicating that the possibility of more noise data in the initial sliding window is greater, and the length of the initial sliding window needs to be increased. The larger the window adjustment degree corresponding to the initial sliding window of the first data is; The larger the value is, the more significant the changes are in the data in the initial sliding window of the first data. This indicates that the data noise level in the initial sliding window is higher, and the longer the initial sliding window needs to be, the greater the window adjustment degree corresponding to the initial sliding window of the first data.

[0109] Preferably, the weights of the data chaos degree and the data non-smoothness degree set in the embodiment of the present invention are empirical values, that is, There is no restriction on this, and it can be set according to the specific real-time scenario.

[0110] (4) Obtaining comparison results between the window adjustment degree and a preset first adjustment threshold and a preset second adjustment threshold, and obtaining an adaptive sliding window of the first data based on the comparison results, wherein the first adjustment threshold is less than the second adjustment threshold.

[0111] Specifically, if the comparison result shows that the window adjustment degree is less than the first adjustment threshold, the length of the initial sliding window of the first data is reduced to obtain a reduced length, and the reduced length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data;

[0112] If the comparison result shows that the window adjustment degree is greater than or equal to the first adjustment threshold, and the window adjustment degree is less than or equal to the second adjustment threshold, using the length of the initial sliding window of the first data as the length of the adaptive sliding window of the first data, to obtain the adaptive sliding window of the first data;

[0113] If the comparison result is that the window adjustment degree is greater than the second adjustment threshold, the length of the initial sliding window of the first data is increased to obtain the increased length, and the increased length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

[0114] In one embodiment, the first adjustment threshold is set to 0.3 and the second adjustment threshold is set to 0.6. There is no limitation here. If the window adjustment degree corresponding to the initial sliding window of the first data is less than the preset first adjustment threshold, that is, , it indicates that the data noise content in the current initial sliding window is relatively low. The initial sliding window can be appropriately reduced to amplify the details of the data in the initial sliding window. Therefore, the length of the initial sliding window of the first data is reduced. For example, the length of the initial sliding window is adjusted from the initial setting of 7 to 5, and the reduced length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

[0115] If the window adjustment degree corresponding to the initial sliding window of the first data is greater than or equal to the preset first adjustment threshold, and the window adjustment degree corresponding to the initial sliding window of the first data is less than or equal to the preset second adjustment threshold, that is, When , it indicates that the length of the initial sliding window of the first data is relatively appropriate, and there is no need to adjust the length of the initial sliding window. That is, the length of the initial sliding window of the first data is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

[0116] If the window adjustment degree corresponding to the initial sliding window of the first data is greater than the preset second adjustment threshold, that is, , it indicates that the data noise content in the current initial sliding window is higher, which may cause some data details to be over-smoothed. In order to weaken the influence of noise data on normal data in the initial sliding window, the length of the initial sliding window of the first data is increased. For example, the length of the initial sliding window is adjusted from the initial setting of 7 to 9, and the increased length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

[0117] At this point, the filtered subsequence corresponding to each subsequence is obtained.

[0118] The parameter control module 14 is used to combine all filtered subsequences into a target smoothed data sequence, and to perform real-time control on the preset acupuncture parameters in the intelligent acupuncture device according to the target smoothed data sequence.

[0119] After obtaining all filtered subsequences, all filtered subsequences are combined into a target smoothed data sequence. At this time, the SG filtering processing of the skin electrodermal signal data sequence is completed. Then, based on the target smoothed data sequence, the doctor further analyzes how acupuncture affects the patient's emotions and physiological state by regulating the autonomic nervous system during a complete acupuncture course, further explores the physiological mechanism of acupuncture, and updates and adjusts the acupuncture parameters such as acupuncture angle, acupuncture depth, and acupuncture speed in the intelligent acupuncture robot in real time based on a series of analysis results, so that the intelligent acupuncture robot can provide more accurate auxiliary acupuncture for patients and automatically record the acupuncture data within each complete acupuncture course. When the patient undergoes the next complete acupuncture course, the intelligent acupuncture robot can formulate a personalized acupuncture treatment plan based on the patient's historical treatment response, and the doctor controls the robotic arm of the intelligent acupuncture robot through the tablet system, so that it can accurately perform processes such as needle insertion, needle movement, and needle removal, thereby providing more accurate acupuncture treatment for patients.

[0120] It should be noted that the focus of the present invention is on how to optimize the SG filtering processing of the skin electrical signal data sequence, and the real-time adjustment of the preset acupuncture parameters in the intelligent acupuncture device based on the filtered skin electrical signal data sequence (i.e., the target smoothed data sequence) belongs to the existing technology and will not be described in detail here.

[0121] In summary, the present invention provides an artificial intelligence-based acupuncture sensation data acquisition and control system, comprising a data acquisition module for acquiring a skin electrical signal data sequence within a preset time period during acupuncture treatment on a patient using an intelligent acupuncture device, and dividing the skin electrical signal data sequence into a preset number of subsequences based on data differences in the skin electrical signal data sequence; a data division module for setting a sliding window for filtering processing for each subsequence based on data fluctuations in each subsequence, acquiring a subsequence and sliding window corresponding to the maximum length as a target subsequence and a target sliding window based on the length of the sliding window corresponding to each subsequence, and filtering each subsequence other than the target subsequence using a non-target sliding window to obtain a corresponding filtered subsequence; a window adjustment module for adaptively adjusting the target sliding window based on data features in the target sliding window during filtering processing of the target subsequence using the target sliding window to obtain a filtered subsequence; and a parameter control module for combining all filtered subsequences into a target smoothed data sequence, and performing real-time control of preset acupuncture parameters in the intelligent acupuncture device based on the target smoothed data sequence. Among them, the collected skin electrical signal data sequence is divided into a preset number of subsequences, and a corresponding sliding window is set for each subsequence according to the fluctuation degree corresponding to each subsequence. The greater the fluctuation degree, the longer the corresponding sliding window. Then, when each subsequence is subjected to SG filtering processing, the sliding window of the subsequence with a small fluctuation degree is kept unchanged, and the sliding window of the subsequence with a large fluctuation degree is adaptively adjusted, and the corresponding sequence after filtering the skin electrical signal data sequence is obtained, that is, the target smoothed data sequence, thereby eliminating the loss of skin electrical signal data details in the sliding window or the decrease in denoising effect due to the use of a fixed-length sliding window, ensuring the denoising effect when filtering the skin electrical signal data sequence, and improving the accuracy of subsequent acupuncture parameter adjustment based on the target smoothed data sequence.

[0122] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. An artificial intelligence-based acupuncture sensation data collection and control system, characterized in that: The system comprises: A data processing module, used to obtain a skin electrical signal data sequence within a preset period of time during the process of performing acupuncture on a patient using the intelligent acupuncture device, and divide the skin electrical signal data sequence into a preset number of subsequences according to data differences in the skin electrical signal data sequence; A data analysis module, configured to set a sliding window for filtering for each of the subsequences according to data fluctuations in each of the subsequences, and to obtain a subsequence and a sliding window corresponding to a maximum length as a target subsequence and a target sliding window according to the length of the sliding window corresponding to each of the subsequences; A filtering processing module, configured to use a non-target sliding window to perform filtering processing on each subsequence except the target subsequence, to obtain a corresponding filtered subsequence; in the process of using the target sliding window to perform filtering processing on the target subsequence, according to the data features in the target sliding window, the target sliding window is adaptively adjusted to obtain a filtered subsequence; A parameter control module, used to form a target smooth data sequence from all filtered subsequences, and to perform real-time control on the preset acupuncture parameters in the intelligent acupuncture device according to the target smooth data sequence; The data processing module divides the skin electrical signal data sequence into a preset number of subsequences according to the data difference in the skin electrical signal data sequence, including: For any data except the last data in the skin electrical signal data sequence, obtain the absolute value of the difference between the data and the next data after the data, and use the absolute value of the difference as the first change degree of the data; Acquire a preset target number of data after the data, form a target data sequence with the data, and acquire a standard deviation of the target data sequence as a second variation degree of the data; According to the first change degree and the second change degree of each data except the last data in the skin electrical signal data sequence, respectively obtaining the acupuncture start data and the acupuncture end data in the skin electrical signal data sequence; Using the acupuncture start data and the acupuncture end data as cutoff points, dividing the skin electrical signal data sequence into a preset number of subsequences; The filtering processing module adaptively adjusts the target sliding window according to the data features in the target sliding window to obtain a filtered subsequence, including: According to the target sliding window, an initial sliding window of the first data is obtained in the target subsequence, wherein the length of the initial sliding window is equal to the length of the target sliding window; According to the data features in the initial sliding window of the first data, an adaptive sliding window of the first data is obtained, and the denoised data corresponding to the first data after filtering is obtained by using the adaptive sliding window. According to the length of the adaptive sliding window, an initial sliding window of the second data in the target subsequence is obtained, and the second data is used as the first data. According to the method for obtaining the denoised data corresponding to the first data after filtering, the denoised data corresponding to the second data after filtering is obtained, until the last data in the target subsequence is traversed, and the denoised data corresponding to each data in the target subsequence is obtained, and all the denoised data after filtering are combined into a filtered subsequence corresponding to the target subsequence; The filtering processing module obtains the adaptive sliding window of the first data according to the data features in the initial sliding window of the first data, including: According to the data chaos feature in the initial sliding window of the first data, obtaining the data chaos degree corresponding to the initial sliding window of the first data; Constructing a target data change curve of the target subsequence in a two-dimensional plane, wherein the horizontal axis of the two-dimensional plane is the acupuncture time, and the vertical axis is the skin electrical signal data value, respectively obtaining the curve curvature between every two adjacent data in the target data change curve, and obtaining the data non-smoothness corresponding to the initial sliding window of the first data according to the curve curvature between every two adjacent data in the initial sliding window of the first data; Performing a weighted summation on the data disorder degree and the data non-smoothness degree to obtain a window adjustment degree corresponding to the initial sliding window of the first data; Obtaining comparison results between the window adjustment degree and a preset first adjustment threshold and a preset second adjustment threshold, respectively, and obtaining an adaptive sliding window of the first data according to the comparison results, wherein the first adjustment threshold is less than the second adjustment threshold; The filtering processing module obtains the adaptive sliding window of the first data according to the comparison result, including: If the comparison result is that the window adjustment degree is less than the first adjustment threshold, the length of the initial sliding window of the first data is reduced to obtain a reduced length, and the reduced length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data; If the comparison result is that the window adjustment degree is greater than or equal to the first adjustment threshold, and the window adjustment degree is less than or equal to the second adjustment threshold, taking the length of the initial sliding window of the first data as the length of the adaptive sliding window of the first data, to obtain the adaptive sliding window of the first data; If the comparison result is that the window adjustment degree is greater than the second adjustment threshold, the length of the initial sliding window of the first data is increased to obtain the increased length, and the increased length is used as the length of the adaptive sliding window of the first data to obtain the adaptive sliding window of the first data.

2. The artificial intelligence-based acupuncture sensation data acquisition and control system according to claim 1, characterized in that: The data processing module obtains acupuncture start data and acupuncture end data in the skin electrical signal data sequence according to the first change degree and the second change degree of each data except the last data in the skin electrical signal data sequence, respectively, including: According to the first change degree of each of the data, at least one maximum first change degree is obtained to form a maximum data sequence, and the data corresponding to the first maximum first change degree in the maximum data sequence is used as the acupuncture start data in the skin electrical signal data sequence; A weighted sum is performed on the first degree of change and the second degree of change of the data to obtain the overall degree of change of the data. According to the overall degree of change of each data, at least one minimum overall degree of change is obtained to form a minimum data sequence. The data corresponding to the first minimum overall degree of change in the minimum data sequence is used as the acupuncture end data in the skin electrical signal data sequence.

3. The artificial intelligence-based acupuncture sensation data acquisition and control system according to claim 1, characterized in that: In the data analysis module, according to the data fluctuation in each of the subsequences, a sliding window for filtering is set for each of the subsequences, including: The variance of each subsequence is obtained respectively, and each variance is used as the fluctuation degree of the corresponding subsequence. According to the fluctuation degree of each subsequence, the length of the sliding window corresponding to the filtering process is set for each subsequence respectively, wherein the greater the fluctuation degree of the subsequence, the greater the length of the sliding window corresponding to the filtering process.

4. The artificial intelligence-based acupuncture sensation data acquisition and control system according to claim 1, characterized in that: The filtering processing module obtains the data disorder degree corresponding to the initial sliding window of the first data according to the data disorder feature in the initial sliding window of the first data, including: According to the data in the initial sliding window of the first data, respectively obtain data information entropy and data standard deviation, perform normalization processing on the data information entropy to obtain a first normalization result, and perform normalization processing on the data standard deviation to obtain a second normalization result; A weighted sum is performed on the first normalization result and the second normalization result to obtain a data disorder degree corresponding to the initial sliding window of the first data.

5. The artificial intelligence-based acupuncture sensation data acquisition and control system according to claim 1, characterized in that: The filtering processing module obtains the data non-smoothness corresponding to the initial sliding window of the first data according to the curve curvature between every two adjacent data in the initial sliding window of the first data, including: The curvatures of the curves between every two adjacent data in the initial sliding window of the first data are combined into a curvature data sequence, and the absolute values ​​of the curvature differences between every two adjacent curve curvatures in the curvature data sequence are respectively obtained to obtain the average absolute value of the curvature difference; The absolute value of the mean curvature difference is normalized to obtain a third normalized result, and the third normalized result is used as the data non-smoothness corresponding to the initial sliding window of the first data.

6. The artificial intelligence-based acupuncture sensation data acquisition and control system according to claim 1, characterized in that: The filtering processing module obtains an initial sliding window of the second data in the target subsequence according to the length of the adaptive sliding window, including: The length of the adaptive sliding window of the first data is used as the length of the initial sliding window of the second data in the target subsequence to obtain the initial sliding window of the second data.

Images