Pelvic floor muscle pressure noise detection method and system for pelvic floor muscle function screening

Abstract

The present invention relates to the technical field of data denoising, and particularly to a pelvic floor muscle pressure noise detection method and system for pelvic floor muscle function screening. The method comprises the following steps: acquiring pelvic floor muscle pressure time series data, segmenting the pelvic floor muscle pressure time series data to obtain a plurality of pressure sequence segments; screening out suspected noise sequence segments from each pressure sequence segment based on the length of each pressure sequence segment; obtaining a recovery trend stability of the suspected noise sequence segment based on a pressure change trend during the process of the suspected noise sequence segment returning to normal, combined with a speed at which the suspected noise sequence segment returns to normal; and obtaining an abnormal factor of the suspected noise sequence segment based on a normal state maintenance duration after the suspected noise sequence segment; and screening out a noise sequence segment from each suspected noise sequence segment based on the abnormal factor, thereby improving the accuracy of noise recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of data denoising, and in particular to a pelvic floor muscle pressure noise detection method and system for pelvic floor muscle function screening. Background Art

[0002] Pelvic floor muscle function screening is an examination used to evaluate the function of the pelvic floor muscles. The pelvic floor muscle pressure data can most directly and quantitatively reflect the physiological state and function of the pelvic floor muscles. It has the advantages of being non-invasive, convenient, and fast, and can obtain a large amount of effective data in a short time. Therefore, it is widely used in pelvic floor muscle function screening.

[0003] Pelvic floor muscle pressure detection requires high-precision sensors to accurately capture subtle pressure changes. During long-term use, the sensors may experience stability issues such as noise interference. There are many types of noise, and some types of noise are often confused with normal physiological phenomena of the body. For example, short-term noise such as electromagnetic interference and normal physiological phenomena such as abdominal pressure have sudden and short-term effects on pelvic floor muscle pressure data. When performing noise detection on pelvic floor muscle pressure data, it is easy to cause low noise recognition accuracy. For example, short-term noise may be identified as normal data. Summary of the Invention

[0004] In order to solve the technical problem of low noise recognition accuracy when performing noise detection on pelvic floor muscle pressure data, the present invention aims to provide a pelvic floor muscle pressure noise detection method and system for pelvic floor muscle function screening. The technical solutions adopted are as follows:

[0005] In a first aspect of the present invention, a pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening is provided, comprising:

[0006] Acquiring pelvic floor muscle pressure time series data, and segmenting the pelvic floor muscle pressure time series data to obtain multiple pressure sequence segments;

[0007] Based on the length of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment;

[0008] According to the pressure change trend of the suspected noise sequence segment during the process of returning to normal, combined with the speed at which the suspected noise sequence segment returns to normal, the stability of the recovery trend of the suspected noise sequence segment is obtained;

[0009] Obtaining an abnormality factor of the suspected noise sequence segment based on the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment;

[0010] According to the abnormal factors, noise sequence segments are obtained by screening out the suspected noise sequence segments.

[0011] In an exemplary embodiment, segmenting the pelvic floor muscle pressure time series data to obtain multiple pressure sequence segments includes:

[0012] The pelvic floor muscle pressure time series data is segmented using an adaptive piecewise constant approximation method to obtain a plurality of pressure sequence segments.

[0013] In an exemplary embodiment, the screening of the suspected noise sequence segments from the pressure sequence segments based on the length of each pressure sequence segment includes:

[0014] Obtaining the mode of length from the lengths of each pressure sequence segment;

[0015] Obtaining the difference between the length of each pressure sequence segment and the length of the mode;

[0016] According to the length difference and the number of occurrences of the mode, the period anomaly degree of each pressure sequence segment is obtained, wherein the period anomaly degree is proportional to the length difference and the number of occurrences;

[0017] According to the degree of periodic anomaly of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment.

[0018] In an exemplary embodiment, the process of obtaining the speed at which the suspected noise sequence segment returns to normal includes:

[0019] Obtaining a target sequence segment of a first suspected noise sequence segment, where the first suspected noise sequence segment is any suspected noise sequence segment, and the target sequence segment is a pressure sequence segment corresponding to the mode that appears for the first time after the first suspected noise sequence segment;

[0020] Obtaining a duration of a recovery phase sequence segment of the first suspected noise sequence segment, where the recovery phase sequence segment is a period between the first suspected noise sequence segment and its corresponding target sequence segment;

[0021] Obtaining a characteristic value difference between a characteristic value of the first suspected noise sequence segment and a characteristic value of the corresponding target sequence segment;

[0022] A speed of the first suspected noise sequence segment returning to normal is obtained according to the duration of the recovery phase sequence segment and the eigenvalue difference; the speed is inversely proportional to the duration of the recovery phase sequence segment and inversely proportional to the eigenvalue difference.

[0023] In an exemplary embodiment, the process of obtaining the pressure change trend during the recovery of the suspected noise sequence segment to normal includes:

[0024] Obtaining each peak in the recovery phase sequence segment;

[0025] The number ratio of the first peak is obtained, where the first peak is a peak that meets the following condition: the peak value is greater than the peak value of the next adjacent peak.

[0026] In an exemplary embodiment, obtaining the recovery trend stability of the suspected noise sequence segment based on the pressure change trend during the process of the suspected noise sequence segment returning to normal and the speed at which the suspected noise sequence segment returns to normal includes:

[0027] The recovery trend stability of the first suspected noise sequence segment is obtained based on the speed at which the first suspected noise sequence segment returns to normal, the proportion of the number of first peaks corresponding to the first suspected noise sequence segment, and the degree of peak change fluctuation of the first peak of the recovery phase sequence segment of the first suspected noise sequence segment; the recovery trend stability is proportional to the speed, proportional to the proportion of the number, and inversely proportional to the degree of peak change fluctuation.

[0028] In an exemplary embodiment, the process of obtaining the duration of the normal state after the suspected noise sequence segment includes:

[0029] The duration of the period between the target sequence segment of the first suspected noise sequence segment and the second suspected noise sequence segment is obtained as the normal state maintenance duration; the second suspected noise sequence segment is the suspected noise sequence segment that appears for the first time after the target sequence segment.

[0030] In an exemplary embodiment, obtaining the abnormality factor of the suspected noise sequence segment based on the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment includes:

[0031] An abnormality factor of the first suspected noise sequence segment is obtained based on the recovery trend stability of the first suspected noise sequence segment and the normal state maintenance time of the first suspected noise sequence segment; the abnormality factor is inversely proportional to the recovery trend stability and inversely proportional to the normal state maintenance time.

[0032] In an exemplary embodiment, screening the suspected noise sequence segments to obtain the noise sequence segments according to the abnormal factor includes:

[0033] The suspected noise sequence segments corresponding to the abnormal factors greater than or equal to the preset abnormal threshold are regarded as noise sequence segments.

[0034] In a second aspect of the present invention, a pelvic floor muscle pressure noise detection system for pelvic floor muscle function screening is provided, comprising: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-mentioned pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening when the program instructions are executed.

[0035] The present invention has the following beneficial effects: by segmenting pelvic floor muscle pressure time series data to obtain multiple pressure sequence segments, and subsequently analyzing each pressure sequence segment, the reliability of the pelvic floor muscle pressure time series data can be improved, thereby improving the accuracy of subsequent data denoising; first, based on the length of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment, and then the suspected noise sequence segments are used as analysis objects, and according to the data change characteristics when an anomaly occurs and the characteristics of noise interference, the recovery trend stability of each suspected noise sequence segment is analyzed, thereby obtaining the abnormal factor of each suspected noise sequence segment, and the noise sequence segments are screened according to the abnormal factor, which can improve the accuracy of noise recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a flow chart of a pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening provided by one embodiment of the present invention;

[0037] Figure 2 is a waveform diagram of pelvic floor muscle pressure time series data provided by an embodiment of the present invention;

[0038] Figure 3 is a flowchart of screening suspected noise sequence segments provided by one embodiment of the present invention;

[0039] Figure 4 This is a flow chart of obtaining a speed at which a suspected noise sequence segment is restored to normal, provided by one embodiment of the present invention;

[0040] Figure 5 This is a flow chart for obtaining a pressure change trend provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] To further illustrate the technical means and effects employed by the present invention to achieve its intended objectives, the following detailed description of the specific embodiments, structures, features, and effects of the present invention is provided in conjunction with the accompanying drawings and preferred embodiments. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this invention belongs. The data and information collected in this application have been obtained with full consent and authorization, and the collection, use, and processing of relevant information must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0043] The application scenario of the pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening provided by the present embodiment is: detecting the user's pelvic floor muscle pressure by a pelvic floor muscle pressure measuring instrument, specifically, sensing the user's pelvic floor muscle pressure by the pressure sensor in the pelvic floor muscle pressure measuring instrument. The pressure data of the pelvic floor muscle directly reflects the contraction and relaxation function of the user's pelvic floor muscle. However, when collecting pelvic floor muscle pressure data, it is inevitable that it will be interfered with by the user itself and environmental factors, and the pressure sensor itself may have certain measurement errors, resulting in abnormal noise data fluctuations in the collected pressure data. The purpose of the pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening provided by the present embodiment is to improve the noise recognition accuracy of pelvic floor muscle pressure data.

[0044] like Figure 1 As shown, the present embodiment provides a pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening, comprising the following steps:

[0045] Step 1: Acquire pelvic floor muscle pressure time series data, segment the pelvic floor muscle pressure time series data, and obtain multiple pressure sequence segments.

[0046] Step 2: Based on the length of each pressure sequence segment, the suspected noise sequence segment is screened from each pressure sequence segment.

[0047] Step 3: Based on the pressure change trend and pressure fluctuation during the recovery process of the suspected noise sequence segment to normal, combined with the speed at which the suspected noise sequence segment recovers to normal, the recovery trend stability of the suspected noise sequence segment is obtained.

[0048] Step 4: Based on the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment, the abnormal factor of the suspected noise sequence segment is obtained.

[0049] Step 5: Based on the abnormal factors, noise sequence segments are screened from each suspected noise sequence segment.

[0050] The specific implementation process of each step of a pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening provided by this embodiment is described below in conjunction with the accompanying drawings.

[0051] Step 1: Acquire pelvic floor muscle pressure time series data, segment the pelvic floor muscle pressure time series data, and obtain multiple pressure sequence segments.

[0052] The user's pelvic floor muscle pressure is collected according to a preset collection frequency. The collection frequency is set according to the actual situation, for example, the collection frequency is once per second.

[0053] A monitoring period is preset, which includes multiple collection moments. Multiple pelvic floor muscle pressure data within the monitoring period are obtained, thereby forming pelvic floor muscle pressure time series data in a time sequence. The duration of the monitoring period can be set according to actual needs, for example, the entire pelvic floor muscle pressure detection process can be used as the monitoring period.

[0054] The contraction force of the pelvic floor muscles is reflected by the changes in pelvic floor muscle pressure. The higher the pelvic floor muscle pressure, the greater the corresponding muscle contraction force. Normal pelvic floor muscle pressure data usually show periodic changes. This is because the pelvic floor muscles are affected by autonomous control and periodic contraction, such as Figure 2 As shown, the horizontal axis is time, and the vertical axis is the amplitude of the pelvic floor muscle pressure data.

[0055] If pelvic floor muscle pressure recovers too slowly, it indicates the muscles aren't contracting properly or are being significantly affected by noise. Sustained high pressure may be related to abdominal pressure. The pelvic floor muscles' spontaneous contraction and relaxation exhibit a rhythmic pattern, with the intervals between contraction and relaxation being relatively stable.

[0056] In order to facilitate subsequent processing, the pelvic floor muscle pressure time series data is segmented to obtain multiple pressure sequence segments. The specific segmentation method is set according to actual needs. In an exemplary embodiment, in order to improve the reliability of data segmentation, and the pelvic floor muscle pressure data in the same pressure sequence segment are relatively relevant, and the pelvic floor muscle pressure data in different pressure sequence segments are relatively different, this embodiment uses the Adaptive Piecewise Constant Approximation (APCA) method to segment the pelvic floor muscle pressure time series data to obtain multiple pressure sequence segments. The adaptive piecewise constant approximation method is a commonly used time series data segmentation method and will not be described in detail. After segmentation using the adaptive piecewise constant approximation method, each pressure sequence segment obtained has an APCA constant value (i.e., APCA approximation), such as the pressure mean of each pressure sequence segment.

[0057] Step 2: Based on the length of each pressure sequence segment, the suspected noise sequence segment is screened from each pressure sequence segment.

[0058] After obtaining each pressure sequence segment, since each pressure sequence segment has regularity and periodicity under normal circumstances, the suspected noise sequence segment can be screened from each pressure sequence segment according to the length of each pressure sequence segment. Figure 3 As shown in Figure 2, a specific screening process for suspected noise sequence segments is given:

[0059] Step 2-1: Obtain the mode of the lengths of each pressure sequence segment.

[0060] Obtain the length of each pressure sequence segment, that is, the number of acquisition moments contained in each pressure sequence segment. Construct a length sequence based on the lengths of each pressure sequence segment. Obtain the mode of the lengths in this length sequence. The mode is the data value that occurs most frequently in a set of data; in other words, obtain the length that occurs most frequently. If there are multiple modes, select any one for analysis. Since the mode is the length that occurs most frequently, it can be understood that the pressure sequence segment corresponding to the mode is the normal pressure sequence segment.

[0061] Step 2-2: Obtain the difference between the length of each pressure sequence segment and the length of the mode.

[0062] The difference between the length of each pressure sequence segment and the length of the mode is obtained. Specifically, the absolute value of the difference between the length of each pressure sequence segment and the mode is calculated as the length difference between the length of each pressure sequence segment and the mode.

[0063] Step 2-3: Based on the length difference and the number of times the mode occurs, the degree of periodic anomaly of each pressure sequence segment is obtained.

[0064] The greater the difference between the length of a pressure sequence segment and the length of the mode, the greater the difference between the length of the pressure sequence segment and the length of the most common pressure sequence segment, and the more abnormal the pressure sequence segment's period is, that is, the greater the degree of period anomaly. The number of times the mode appears is used as the credibility of the mode; the more times the mode appears, the higher the corresponding mode's credibility. Therefore, based on the length difference and the number of times the mode appears, the degree of period anomaly for each pressure sequence segment is determined. The degree of period anomaly is proportional to the length difference and the number of times the mode appears.

[0065] In an exemplary embodiment, a specific calculation formula for the degree of periodic anomaly of a pressure sequence segment is given below:

[0066] ;

[0067] in, Indicates the degree of cycle anomaly of the xth pressure sequence segment, Indicates the length of the xth pressure sequence segment, Indicates the majority, It represents the absolute value of the difference between the length of the x-th pressure sequence segment and the mode, Indicates the number of times the mode appears, norm indicates the normalization function. The normalization method here can be: obtain the maximum and minimum values ​​of the product of the absolute value of the difference between the length of all pressure sequence segments and the mode and the number of times the mode appears, and then use the maximum and minimum value normalization method to normalize the pressure sequence segments. Normalize it and put it into the numerical range of 0-1 to facilitate subsequent comparison.

[0068] Step 2-4: Based on the degree of periodic anomaly of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment.

[0069] The periodicity of the pressure sequence segments reflects the contraction and relaxation patterns of the pelvic floor muscles, typically exhibiting regular fluctuations. However, interference from abdominal pressure and external environmental factors can disrupt the periodicity of pelvic floor muscle pressure data. Abdominal pressure is closely related to breathing and abdominal muscle activity. During breathing, the pressure generated by the abdominal muscles can disrupt the periodicity of pelvic floor muscle pressure.

[0070] Therefore, the degree of periodic anomaly of each pressure sequence segment reflects the degree of periodic anomaly of each pressure sequence segment. The greater the degree of periodic anomaly, the more abnormal the periodic state. Therefore, according to the degree of periodic anomaly of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment. In an exemplary embodiment, a periodic anomaly threshold is preset. The numerical range of the preset periodic anomaly threshold is 0-1, and its specific numerical value is set according to actual judgment needs. If the noise denoising judgment needs to be more accurate, the preset periodic anomaly threshold can be set smaller, so that more pressure sequence segments are judged as suspected noise sequence segments. Accordingly, more suspected noise sequence segments need to be processed subsequently. If the noise denoising judgment needs to be looser, the preset periodic anomaly threshold can be set larger, so that fewer pressure sequence segments are judged as suspected noise sequence segments. Accordingly, fewer suspected noise sequence segments need to be processed subsequently. In an exemplary embodiment, the preset periodic anomaly threshold is 0.7 as an example.

[0071] The periodic abnormality degree of each pressure sequence segment is compared with a preset periodic abnormality degree threshold, and the pressure sequence segment corresponding to the periodic abnormality degree greater than or equal to the preset periodic abnormality degree threshold is determined as a suspected noise sequence segment.

[0072] Step 3: Based on the pressure change trend during the recovery process of the suspected noise sequence segment to normal, combined with the speed at which the suspected noise sequence segment recovers to normal, the recovery trend stability of the suspected noise sequence segment is obtained.

[0073] Under normal circumstances, during the process of contraction and relaxation, the pressure of the pelvic floor muscles usually recovers quickly, and the pressure curve shows a rapid drop back to the resting state, showing the elasticity and good function of the muscles.

[0074] The suspected noise sequence segment is taken as the key analysis object. The suspected noise sequence segment will gradually return to normal. During the process of returning to normal, the pressure change trend of the pelvic floor muscle will affect the stability of the recovery trend of the suspected noise sequence segment. Moreover, the speed at which the suspected noise sequence segment returns to normal will also determine the stability of the recovery trend of the suspected noise sequence segment. Therefore, according to the pressure change trend during the process of the suspected noise sequence segment returning to normal and the speed at which the suspected noise sequence segment returns to normal, the stability of the recovery trend of the suspected noise sequence segment is obtained.

[0075] In an exemplary embodiment, Figure 4 As shown in Figure 2, a specific process for obtaining the speed at which a suspected noise sequence segment returns to normal is given:

[0076] Step 3-1: Obtain a target sequence segment of a first suspected noise sequence segment.

[0077] The pressure sequence segment corresponding to the mode is the pressure sequence segment with the largest number of length occurrences, that is, the most frequently occurring pressure sequence segment. Therefore, the pressure sequence segment corresponding to the mode is the pressure sequence segment under normal circumstances, that is, the pressure sequence segment corresponding to the mode cannot be a suspected noise sequence segment.

[0078] For ease of explanation, assume that the first suspected noise sequence segment is any suspected noise sequence segment. The search continues from the first suspected noise sequence segment, ending at the first occurrence of a pressure sequence segment corresponding to the mode. The pressure sequence segment corresponding to the first occurrence of the mode after the first suspected noise sequence segment is obtained, and this pressure sequence segment is defined as the target sequence segment for the first suspected noise sequence segment.

[0079] Step 3-2: Obtain the duration of the recovery phase sequence segment of the first suspected noise sequence segment.

[0080] The first suspected noise sequence segment is a suspected abnormal pressure sequence segment, and the target sequence segment of the first suspected noise sequence segment is the target sequence segment under normal conditions. Therefore, the period between the first suspected noise sequence segment and the target sequence segment of the first suspected noise sequence segment can be understood as the recovery phase of the first suspected noise sequence segment returning to normal. Then, the recovery phase sequence segment is set to be the period between the first suspected noise sequence segment and the target sequence segment of the first suspected noise sequence segment. And the duration of the recovery phase sequence segment of the first suspected noise sequence segment is obtained. The duration of the recovery phase sequence segment of the first suspected noise sequence segment is the time interval between the first suspected noise sequence segment and its corresponding target sequence segment, and the time interval is the time interval between the end moment of the first suspected noise sequence segment and the start moment of its corresponding target sequence segment.

[0081] Step 3-3: Obtain a characteristic value difference between the characteristic value of the first suspected noise sequence segment and the characteristic value of its corresponding target sequence segment.

[0082] Obtain the characteristic value of the first suspected noise sequence segment and the characteristic value of the target sequence segment of the first suspected noise sequence segment. The characteristic value reflects the characteristics of the pressure data in the corresponding pressure sequence segment. In an exemplary embodiment, the characteristic value is the average value of the pelvic floor muscle pressure in the corresponding pressure sequence segment, which can also be understood as the APCA approximation of the corresponding pressure sequence segment. Then, the characteristic value of the first suspected noise sequence segment is the average value of the pelvic floor muscle pressure of the first suspected noise sequence segment, that is, the APCA approximation of the first suspected noise sequence segment; the characteristic value of the target sequence segment of the first suspected noise sequence segment is the average value of the pelvic floor muscle pressure of the target sequence segment of the first suspected noise sequence segment, that is, the APCA approximation of the target sequence segment of the first suspected noise sequence segment.

[0083] Obtain a characteristic value difference between the characteristic value of the first suspected noise sequence segment and the characteristic value of the corresponding target sequence segment, specifically, the absolute value of the characteristic value difference. The characteristic value difference represents the difference in pelvic floor muscle pressure between abnormal and normal corresponding to the first suspected noise sequence segment.

[0084] Step 3-4: Obtain the speed at which the first suspected noise sequence segment returns to normal according to the duration and characteristic value difference of the sequence segment in the recovery phase.

[0085] The longer the duration of the recovery phase sequence segment, the longer it takes for the first suspected noise sequence segment to recover from abnormality to normal, and the slower the speed at which the first suspected noise sequence segment recovers to normal; the greater the difference in eigenvalues, the greater the difference in pelvic floor muscle pressure between the abnormal and normal corresponding to the first suspected noise sequence segment, and the slower the speed at which the first suspected noise sequence segment recovers to normal.

[0086] Therefore, the speed of the first suspected noise sequence segment returning to normal is obtained according to the duration and eigenvalue difference of the recovery phase sequence segment. The speed is inversely proportional to the duration of the recovery phase sequence segment and inversely proportional to the eigenvalue difference.

[0087] In an exemplary embodiment, a specific method for calculating the speed is given as follows:

[0088] ;

[0089] in, Indicates the speed at which the yth suspected noise sequence segment returns to normal, It represents the duration of the recovery phase sequence segment of the yth suspected noise sequence segment, Indicates the eigenvalue difference between the eigenvalue of the y-th suspected noise sequence segment and the eigenvalue of the target sequence segment of the y-th suspected noise sequence segment.

[0090] Express The negative correlation normalization method here can be: obtain the maximum and minimum values ​​of the product of the duration of the recovery phase sequence segment corresponding to each suspected noise sequence segment and the characteristic value difference, and then use the maximum and minimum value normalization method to normalize the Normalize and finally calculate the value 1 and the normalized The difference between Negative correlation normalization.

[0091] The slower the yth suspected noise sequence segment returns to normal, the stronger the noise interference or the longer the noise impact. The faster the yth suspected noise sequence segment returns to normal, the faster it can return to a resting state (i.e., a normal state) in a short period of time. This indicates that the yth suspected noise sequence segment is only a temporary fluctuation or interference, such as a short-term external electromagnetic interference, which is sudden and transient, or is affected by abdominal pressure.

[0092] According to the above content, in addition to the normal physiological phenomenon of abdominal pressure that will cause the pelvic floor muscle pressure data to have a low periodicity, there are two types of noise that can cause low periodicity. One type of abnormality requires a longer recovery time, that is, the corresponding suspected noise sequence segment recovers to the normal state more slowly. The other type of abnormality requires a shorter recovery time, and the abdominal pressure will also recover to normal on its own in a short time. The interference caused by this short-term noise and abdominal pressure on the pelvic floor muscle pressure data is similar in recovery speed. Therefore, further analysis is needed to better distinguish between short-term noise and abdominal pressure in the pelvic floor muscle pressure data.

[0093] To further distinguish the impact of short-term noise and abdominal pressure data in pressure sequences with faster recovery speeds, we compared the recovery trends and durations of multiple suspected noise sequences and analyzed the changing patterns during the recovery process. If a pressure sequence shows very unstable recovery or takes a long time to recover, it is more likely to be noise. The recovery process generally refers to the process by which the pelvic floor muscle pressure signal gradually returns to normal after being disturbed by noise.

[0094] Then, the pressure change trend during the process of the suspected noise sequence segment returning to normal is obtained. In an exemplary embodiment, as shown in FIG. Figure 5 As shown, a specific process for obtaining the pressure change trend during the recovery of the suspected noise sequence segment to normal is given below:

[0095] Step 3-5: Obtain each peak in the recovery phase sequence segment.

[0096] Acquire each peak in the recovery phase sequence segment of the yth suspected noise sequence segment, that is, each maximum pressure value.

[0097] Step 3-6: Obtain the number ratio of the first peak, where the first peak is a peak that meets the following conditions: the peak value is greater than the peak value of the next adjacent peak.

[0098] Observe the peaks in the recovery phase of the yth suspected noise sequence segment, that is, the amplitude of the pelvic floor muscle pressure fluctuations during the recovery process. If the amplitude of each pressure fluctuation gradually decreases and approaches normal during the recovery process, it indicates a relatively stable recovery trend. If the pressure value fluctuates widely and irregularly, such as a sharp rise and then a sharp drop, the recovery trend is unstable. Large fluctuations indicate the presence of significant noise.

[0099] For each peak in the recovery phase of the yth suspected noise sequence segment, calculate the The peak value of the first peak and the The difference between the peak values ​​of the ,like Greater than 0 (i.e. The peak value of the first peak is greater than the peak value of the next peak adjacent to it), then the The first peak is recorded as the first peak, and the first peak is a normal peak. Then, the first peak is a peak that meets the following conditions: the peak value is greater than the peak value of the next peak adjacent to it.

[0100] Then, the ratio of the number of first peaks in the recovery phase of the yth suspected noise sequence segment is obtained, that is, the ratio of the number of first peaks in the recovery phase of the yth suspected noise sequence segment to the total number of peaks in the recovery phase of the yth suspected noise sequence segment. A larger ratio of the number of first peaks indicates more normal peaks and a more stable recovery trend of the yth suspected noise sequence segment.

[0101] Obtain the peak value fluctuation of the first peak of the recovery phase sequence segment of the first suspected noise sequence segment. Taking the yth suspected noise sequence segment as an example, obtain the peak values ​​of each first peak of the recovery phase sequence segment of the yth suspected noise sequence segment and arrange them in time sequence to obtain a first peak peak sequence. Then, calculate the difference sequence of the first peak peaks. Finally, calculate the variance of this difference sequence as the peak value fluctuation of the first peak of the recovery phase sequence segment of the yth suspected noise sequence segment. A larger variance indicates more uneven differences between adjacent normal peaks during the recovery process and a more unstable recovery trend.

[0102] Finally, according to the pressure change trend during the process of the suspected noise sequence segment returning to normal, combined with the speed at which the suspected noise sequence segment returns to normal, the recovery trend stability of the suspected noise sequence segment is obtained. In an exemplary embodiment, the recovery trend stability of the first suspected noise sequence segment is obtained based on the speed at which the first suspected noise sequence segment returns to normal, the number percentage of the first peak corresponding to the first suspected noise sequence segment, and the peak change fluctuation degree of the first peak of the recovery phase sequence segment of the first suspected noise sequence segment. The faster the speed at which the first suspected noise sequence segment returns to normal, that is, the greater the speed, the more stable the recovery trend of the first suspected noise sequence segment, that is, the greater the recovery trend stability of the first suspected noise sequence segment, and the recovery trend stability is proportional to the speed. Through the above analysis, the recovery trend stability is proportional to the number percentage and inversely proportional to the peak change fluctuation degree.

[0103] In an exemplary embodiment, a specific calculation formula for the stability of the recovery trend is given as follows:

[0104] ;

[0105] in, represents the stability of the recovery trend of the yth suspected noise sequence segment, Indicates the number of the first peaks of the recovery phase sequence segment of the yth suspected noise sequence segment, Indicates the total number of peaks in the recovery phase of the y-th suspected noise sequence segment, It represents the proportion of the first peaks in the recovery phase of the y-th suspected noise sequence segment. The greater the proportion of the first peaks, that is, the greater the proportion of the normal peaks, the more stable the recovery trend of the y-th suspected noise sequence segment.

[0106] It represents the variance of the difference sequence of the first peak value sequence of the recovery phase of the y-th suspected noise sequence segment. The larger the variance, the more uneven the differences between adjacent normal peaks in the recovery process of the y-th suspected noise sequence segment, and the more unstable the recovery trend. Express The negative correlation normalization here can be: obtain the maximum and minimum values ​​of the variance corresponding to each suspected noise sequence segment, and then use the maximum and minimum value normalization method to normalize the Normalize and finally calculate the value 1 and the normalized The difference between Negative correlation normalization.

[0107] The speed at which the yth suspected noise sequence segment returns to normal The faster it is, the more stable the recovery trend will be. Express Normalization, the normalization method here can be: obtain the maximum and minimum values ​​of the speed corresponding to each suspected noise sequence segment, and then use the maximum and minimum value normalization method to normalize the speed. Perform normalization.

[0108] Step 4: Based on the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment, the abnormal factor of the suspected noise sequence segment is obtained.

[0109] This step requires further analysis of the duration after the suspected noise sequence segment returns to normal, i.e., how long the normal state lasts. After returning to normal, normal pelvic floor muscle pressure data should remain relatively constant for a certain period of time. However, if the normal state is quickly broken again, with new fluctuations or anomalies, i.e., if the normal state lasts for a short period of time, the suspected noise sequence segment is more likely to be noise.

[0110] Therefore, it is necessary to obtain the normal state maintenance duration after the suspected noise sequence segment. In an exemplary embodiment, for the first suspected noise sequence segment, the suspected noise sequence that appears for the first time after the target sequence segment of the first suspected noise sequence segment is obtained is defined as the second suspected noise sequence segment. Then, the second suspected noise sequence segment indicates that an abnormal situation occurs again after the first suspected noise sequence segment returns to a normal state. Then, the starting moment of the second suspected noise sequence segment is the abnormal state start moment from the normal state to the abnormal state. The duration of the period between the target sequence segment of the first suspected noise sequence segment and the second suspected noise sequence segment is obtained as the normal state maintenance duration of the first suspected noise sequence segment. The normal state maintenance duration is the time interval between the end moment of the target sequence segment of the first suspected noise sequence segment and the start moment of the second suspected noise sequence segment.

[0111] Then, based on the stability of the recovery trend of the first suspected noise sequence segment and the duration of the normal state of the first suspected noise sequence segment, the abnormality factor of the first suspected noise sequence segment is obtained. The higher the stability of the recovery trend of the first suspected noise sequence segment, the less abnormal the first suspected noise sequence segment is, that is, the smaller the abnormality factor is. Therefore, the abnormality factor is inversely proportional to the stability of the recovery trend. The shorter the duration of the normal state of the first suspected noise sequence segment, the sooner a new abnormality appears after returning to the normal state, that is, the sooner it enters the abnormal state, that is, the larger the abnormality factor is. Therefore, the abnormality factor is inversely proportional to the duration of the normal state.

[0112] In an exemplary embodiment, a specific calculation formula of the abnormality factor is given as follows:

[0113] ;

[0114] in, It represents the abnormal factor of the y-th suspected noise sequence segment. The abnormal factor is specifically the abnormal factor of the recovery trend. It indicates the duration of the normal state corresponding to the y-th suspected noise sequence segment. The shorter the normal state duration, the greater the abnormality.

[0115] Express The normalization method here can be: obtain the maximum and minimum values ​​of the product of the recovery trend stability and the normal state maintenance time corresponding to each suspected noise sequence segment, and then use the maximum and minimum value normalization method to normalize Perform normalization.

[0116] Step 5: Based on the abnormal factors, noise sequence segments are screened from each suspected noise sequence segment.

[0117] Step 4 is used to obtain the abnormality factor of each suspected noise sequence segment. The larger the abnormality factor, the more abnormal the corresponding suspected noise sequence segment is, that is, the more likely it is that noise exists. Therefore, according to the abnormality factor, the noise sequence segment is screened from each suspected noise sequence segment.

[0118] In an exemplary embodiment, an abnormality threshold is preset, and the numerical range of the preset abnormality threshold is 0-1. The specific value of the preset abnormality threshold is further set according to actual needs. If the requirements for noise detection are stricter, the preset abnormality threshold can be set smaller, thereby enabling more suspected noise sequence segments to be identified as noise sequence segments. Correspondingly, if the requirements for noise detection are looser, the preset abnormality threshold can be set larger, thereby enabling fewer suspected noise sequence segments to be identified as noise sequence segments. In an exemplary embodiment, the preset abnormality threshold is 0.7, for example.

[0119] The abnormality factor of each suspected noise sequence segment is compared with a preset abnormality threshold. The suspected noise sequence segments corresponding to abnormality factors greater than or equal to the preset abnormality threshold are obtained. The suspected noise sequence segments corresponding to abnormality factors greater than or equal to the preset abnormality threshold are then used as noise sequence segments. This allows for accurate noise identification of pelvic floor muscle pressure time series data.

[0120] In subsequent practical applications, the abnormal factor of the z-th noise sequence segment can also be As the weight of all pelvic floor muscle pressure data in the zth noise sequence segment. The value 1 is used as the weight of all pelvic floor muscle pressure data in all pressure sequence segments except the noise sequence segment. In this way, the weight of each pelvic floor muscle pressure data in the pelvic floor muscle pressure time series data is obtained, and the weighted filtering algorithm is used to filter and denoise the pelvic floor muscle pressure time series data.

[0121] This embodiment also provides a pelvic floor muscle pressure noise detection system for pelvic floor muscle function screening, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-mentioned pelvic floor muscle pressure noise detection method embodiment for pelvic floor muscle function screening when the program instructions are executed.

[0122] In an exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the above-mentioned embodiment of the pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening.

[0123] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0124] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening, characterized in that: include: Acquiring pelvic floor muscle pressure time series data, and segmenting the pelvic floor muscle pressure time series data to obtain multiple pressure sequence segments; Based on the length of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment; According to the pressure change trend of the suspected noise sequence segment during the process of returning to normal, combined with the speed at which the suspected noise sequence segment returns to normal, the stability of the recovery trend of the suspected noise sequence segment is obtained; Obtaining an abnormality factor of the suspected noise sequence segment based on the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment; According to the abnormal factor, a noise sequence segment is obtained by screening each suspected noise sequence segment; The process of obtaining the speed at which the suspected noise sequence segment returns to normal includes: Obtaining a target sequence segment for the first suspected noise sequence segment, where the first suspected noise sequence segment is any suspected noise sequence segment, and the target sequence segment is a pressure sequence segment corresponding to the mode of the lengths of the pressure sequence segments that appear for the first time after the first suspected noise sequence segment; Obtaining a duration of a recovery phase sequence segment of the first suspected noise sequence segment, where the recovery phase sequence segment is a period between the first suspected noise sequence segment and its corresponding target sequence segment; Obtaining a characteristic value difference between a characteristic value of the first suspected noise sequence segment and a characteristic value of the corresponding target sequence segment; Obtaining a speed of the first suspected noise sequence segment returning to normal based on the duration of the recovery phase sequence segment and the characteristic value difference; wherein the speed is inversely proportional to the duration of the recovery phase sequence segment and inversely proportional to the characteristic value difference; The process of obtaining the pressure change trend during the process of the suspected noise sequence segment returning to normal includes: Obtaining each peak in the recovery phase sequence segment; Obtaining a ratio of the number of first peaks, where the first peak is a peak that meets the following conditions: the peak value is greater than the peak value of the next adjacent peak; The stability of the recovery trend of the suspected noise sequence segment is obtained based on the pressure change trend during the process of the suspected noise sequence segment returning to normal and the speed at which the suspected noise sequence segment returns to normal, including: The recovery trend stability of the first suspected noise sequence segment is obtained based on the speed at which the first suspected noise sequence segment returns to normal, the proportion of the number of first peaks corresponding to the first suspected noise sequence segment, and the degree of peak change fluctuation of the first peak of the recovery phase sequence segment of the first suspected noise sequence segment; the recovery trend stability is proportional to the speed, proportional to the proportion of the number, and inversely proportional to the degree of peak change fluctuation.

2. The pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to claim 1, characterized in that: The pelvic floor muscle pressure time series data is segmented to obtain a plurality of pressure sequence segments, including: The pelvic floor muscle pressure time series data is segmented using an adaptive piecewise constant approximation method to obtain a plurality of pressure sequence segments.

3. The pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to claim 1, characterized in that: The method of screening the suspected noise sequence segments from the pressure sequence segments based on the length of each pressure sequence segment includes: Obtaining the mode of length from the lengths of each pressure sequence segment; Obtaining the difference between the length of each pressure sequence segment and the length of the mode; According to the length difference and the number of occurrences of the mode, the period anomaly degree of each pressure sequence segment is obtained, wherein the period anomaly degree is proportional to the length difference and the number of occurrences; According to the degree of periodic anomaly of each pressure sequence segment, suspected noise sequence segments are screened from each pressure sequence segment.

4. The pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to claim 1, characterized in that: The process of obtaining the duration of the normal state after the suspected noise sequence segment includes: The duration of the period between the target sequence segment of the first suspected noise sequence segment and the second suspected noise sequence segment is obtained as the normal state maintenance duration; the second suspected noise sequence segment is the suspected noise sequence segment that appears for the first time after the target sequence segment.

5. The pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to claim 4, characterized in that: The abnormal factor of the suspected noise sequence segment is obtained according to the stability of the recovery trend and the duration of the normal state after the suspected noise sequence segment, including: An abnormality factor of the first suspected noise sequence segment is obtained based on the recovery trend stability of the first suspected noise sequence segment and the normal state maintenance time of the first suspected noise sequence segment; the abnormality factor is inversely proportional to the recovery trend stability and inversely proportional to the normal state maintenance time.

6. The pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to claim 1, characterized in that: The step of screening the suspected noise sequence segments to obtain the noise sequence segments according to the abnormal factors includes: The suspected noise sequence segments corresponding to the abnormal factors greater than or equal to the preset abnormal threshold are regarded as noise sequence segments.

7. A pelvic floor muscle pressure noise detection system for pelvic floor muscle function screening, characterized in that: include: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is configured to implement the pelvic floor muscle pressure noise detection method for pelvic floor muscle function screening according to any one of claims 1 to 6 when the program instructions are executed.

Images