A method for recognizing a pulse wave based on a P-P interval sequence pulse rate irregularity
By acquiring radial artery pulse signals, constructing a PP interval time series and performing threshold processing, the problem of inaccurate identification of intermittent and knotted pulses using traditional Chinese medicine experience was solved, achieving accurate identification and digitization of pulse signals.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHONGBEI UNIV
- Filing Date
- 2023-06-27
- Publication Date
- 2026-05-19
AI Technical Summary
Traditional Chinese medicine relies on experience to accurately identify intermittent and knotted pulses, which makes the digitization of TCM pulse diagnosis difficult.
By acquiring pulse signals at the radial artery, a PP interval time series was constructed, threshold processing was performed, pulse signal categories were identified, and pulse rate variability was determined using differential calculation and threshold processing. A pulse rate variability signal map was then plotted to distinguish between intermittent pulse, knotted pulse, and normal pulse rate.
It improves the accuracy of pulse signal recognition and can effectively distinguish between normal pulse rate, intermittent pulse and knotted pulse, laying the foundation for the digitalization of TCM pulse diagnosis.
Smart Images
Figure CN116530950B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of digital research on pulse diagnosis in traditional Chinese medicine, and more specifically, to a method for identifying intermittent pulse abnormalities based on PP interval sequences. Background Technology
[0002] The human circulatory system relies on the heart and blood vessels to transport nutrients to cells throughout the body to maintain daily metabolism. The heartbeat is influenced by both its own rhythmic electrical activity and the autonomic nervous system. Stimulation of the sympathetic nervous system increases myocardial contraction, leading to a faster heartbeat; while excitation of the vagus nerve reduces the conduction velocity of the atrioventricular node, slowing the heartbeat. Arrhythmia is a manifestation of abnormal heartbeats, as obstructions in the origin or conduction of the heartbeat can cause abnormalities in the frequency and rhythm of the pulse. Under normal physiological conditions, the heart and its conduction functions operate normally, and the autonomic nervous system maintains a dynamic balance. However, when the heart or other parts of the body are attacked or damaged by disease, the balance of the autonomic nervous system is disrupted, leading to cardiovascular disorders and significant changes in the frequency and rhythm of the pulse.
[0003] The pulse wave is a physical reflection of the cardiovascular system and is closely related to various physiological changes. The pulse is produced by the combined action of the heart, blood, and arterial walls, and its frequency and rhythm are crucial characteristic parameters. During the onset of cardiovascular disease, the radial artery pulsation shows significant changes. Therefore, extracting changes in frequency and rhythm from pulse signals is of great significance for the study of cardiovascular diseases.
[0004] Intermittent and knotted pulses are important pulse patterns reflecting changes in frequency and rhythm. Generally, the identification of intermittent and knotted pulses relies primarily on the personal experience and intuition of traditional Chinese medicine (TCM) practitioners, and different TCM practitioners may yield different results. Furthermore, the identification depends on the practitioner's working state; even the same practitioner may yield different results at different times. This presents a challenge for the digitization of TCM pulse diagnosis. In other words, relying solely on TCM practitioners' experience to identify intermittent and knotted pulses is insufficiently accurate, thus hindering the digitization of TCM pulse diagnosis. Summary of the Invention
[0005] The purpose of this invention is to address the shortcomings of the prior art by providing a method for identifying intermittent pulse based on PP interval sequence pulse rate abnormalities, thereby solving the problem that relying on traditional Chinese medicine experience to identify intermittent and knotted pulses is not accurate enough.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0007] This application provides a method for identifying intermittent pulse based on PP interval sequence pulse rate abnormalities. The method includes the following steps: S1, acquiring the sample to be tested; S2, extracting the time information of the same feature points from continuous pulse signals to construct the corresponding PP interval time series; S3, performing threshold processing on the PP interval time series, and obtaining the pulse signal category based on the threshold processing result.
[0008] Furthermore, acquiring the sample to be tested refers to acquiring the pulse signal at the radial artery of the wrist of the person being tested, which is directly acquired through a pulse acquisition device; the acquired parameters include at least a one-dimensional time series, signal sampling rate, and measurement time.
[0009] Furthermore, the pulse signal measurement time is greater than 10 minutes, the signal sampling rate is 1kHz, and the one-dimensional time series refers to the numerical values representing the amplitude of the pulse signal vibration arranged in chronological order.
[0010] Furthermore, extracting the time information of the same feature points from continuous pulse signals to construct the corresponding PP interval time series includes: extracting the time information corresponding to the local maxima in each cardiac cycle of the pulse signal to obtain a time series array. ,in, Indicates the first Time information corresponding to the same feature points in each cardiac cycle.
[0011] Furthermore, extracting the time information of the same feature points from continuous pulse signals and constructing the corresponding PP interval time series also includes: performing difference calculations on the time series array sequentially to obtain the PP interval time series. ,in, This represents the result of the difference calculation. The expression for the difference calculation is: Where N is the number of identical feature points in the time series array. Indicates the first The PP interval corresponding to each cardiac cycle and They represent the first The and the first Time information corresponding to the same feature points in each cardiac cycle.
[0012] Furthermore, thresholding is performed on the PP interval time series, and the pulse signal categories obtained based on the thresholding results are as follows:
[0013] PP interval time series Thresholding is performed to obtain a threshold, and the pulse signal category is determined based on the rhythm of the threshold; the expression for thresholding is:
[0014]
[0015] in, For the threshold, and It is the first The and the first The time value of each PP interval.
[0016] Furthermore, thresholding is performed on the PP interval time series. Based on the thresholding results, the pulse signal categories include: intermittent pulse with periodic peaks when the threshold is greater than 0; knotted pulse with irregular peaks when the threshold is greater than 0; and normal pulse rate with a straight line without peaks when the threshold is greater than 0.
[0017] Furthermore, thresholding is performed on the PP interval time series, and the pulse signal category is obtained based on the thresholding result. This also includes plotting a pulse rate variability signal graph, with the threshold obtained after thresholding as the vertical axis and the number of beats as the horizontal axis.
[0018] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention extracts local maxima in the pulse signal as common feature points, performs differential calculations on the time information corresponding to these common feature points to obtain the corresponding PP interval time series; then, it performs threshold processing on the PP interval time series to obtain pulse rate variation information, and determines intermittent pulse, knotted pulse, and normal pulse rate based on rhythms with a threshold greater than 0. This invention performs threshold processing on the PP interval sequence, converting normal pulse rate bands to 0, resulting in a higher signal-to-noise ratio for abnormal pulse rate bands, making them easier to display, thereby effectively improving the recognition accuracy. This method can effectively identify normal pulse rate, intermittent pulse, and knotted pulse, laying the foundation for the digitalization of traditional Chinese medicine pulse diagnosis. Attached Figure Description
[0019] Figure 1 A schematic diagram illustrating a method for identifying arrhythmic pulses based on PP interval sequence pulse rate abnormalities provided by the present invention;
[0020] Figure 2 This is a schematic diagram of the pulse signal obtained in step S1 of the method for identifying arrhythmic pulse based on PP interval sequence provided by the present invention.
[0021] Figure 3 The continuous pulse signals of normal pulse and intermittent pulse obtained in step S1 of the method for identifying intermittent pulse arrhythmia based on PP interval sequence provided by the present invention are as follows: Figure 3 (a) The pulse signal of a healthy person; Figure 3 (b) and Figure 3 (c) These are the intermittent pulse signal and the knot pulse signal, respectively;
[0022] Figure 4This is the PP interval diagram after threshold processing in step S3 of the method for identifying arrhythmias based on PP interval sequence provided by the present invention. Figure 4 (a) The pulse signal of a healthy person; Figure 4 (b) and Figure 4 (c) are the intermittent pulse signal and the knot pulse signal, respectively. Detailed Implementation
[0023] To make the implementation process of this invention clearer, a detailed description will be provided below in conjunction with the accompanying drawings.
[0024] This invention provides a method for identifying intermittent pulses based on PP interval sequence pulse rate aberrations, such as... Figure 1 As shown, the method includes the following steps:
[0025] S1, Obtain the sample to be tested;
[0026] The test sample in this invention is a pulse signal, which is an amplitude curve that varies over time. The pulse signal can be from an existing database or obtained directly by a pulse acquisition device. Any pulse signal with an amplitude curve that varies over time is suitable for the identification method of this invention. In this embodiment, the pulse signal sample is self-tested data, obtained from the radial artery at the wrist using a self-built pulse acquisition device. The acquired pulse signal is as follows: Figure 2 As shown, the fluctuations are displayed as a function of time. The acquired signal, with time as the independent variable, becomes a one-dimensional time series. The characteristics of a pulse signal include at least a one-dimensional time series, a signal sampling rate, and a measurement time. In this embodiment, the signal sampling rate is 1kHz, the continuous measurement time is greater than 10 minutes, and the measurement time for a single sample is greater than 5 minutes. The samples in this embodiment include healthy individuals and patients with cardiovascular diseases, such as... Figure 3 (a) The pulse signal of a healthy person; Figure 3 (b) and Figure 3 (c) are the intermittent pulse signal and the knot pulse signal, respectively.
[0027] Before proceeding to the next steps, denoising processing is required, which uses ensemble empirical mode decomposition. For details of the denoising process, please refer to the method disclosed in the invention patent application entitled "A Pulse Signal Acquisition and Processing Method Based on EEMD-CSI Algorithm" with publication number "CN115227210A".
[0028] S2, extract the time information of the same feature points from continuous pulse signals to construct the corresponding PP interval time series;
[0029] The time interval between two adjacent pulse waves generated by a heartbeat is called the PP interval (Pulse-to-Pulse, PP). Identical feature points are extracted from the one-dimensional time series of the pulse signal, and then the differences between adjacent feature points are calculated to construct the corresponding PP interval time series. Specifically, local maxima points are extracted from each cycle of the pulse signal, i.e., the aforementioned identical feature points, and the time information of these identical feature points is extracted from continuous pulse signals. Time series array This represents multiple feature points corresponding to the pulse signal, where... Indicates the first Feature points corresponding to the pulse signal during the cardiac cycle. A sampling rate sufficient for the pulse signal is set, and the time series array is sequentially differentially calculated to obtain the corresponding PP interval time series; the expression for the differential calculation is: Where N is the number of identical feature points in the time series array. Indicates the first The PP interval of a pulse signal sequence is calculated by taking the difference between all the intervals in the time series array. The PP interval time series is the sum of all the differences in the time series array. The PP interval of all individual pulse segments in a pulse signal time series array is represented as... This corresponds to one cardiac cycle of the left ventricle.
[0030] S3, threshold processing is performed on the PP interval time series, and the pulse signal category is obtained based on the threshold processing result.
[0031] The PP interval time series obtained in step S2 Threshold processing is then performed. The pulse rate threshold classification is defined as follows: ,in and It is the first The and the first The time values of each PP interval are numerically related to the PP interval time series. The corresponding values in are equal, that is , N represents the number of identical feature points in the time series array. Thresholding is derived from the following expression, which uses the thresholding method to generate fuzzy regularized discrete time series with PP intervals. :
[0032]
[0033] The obtained threshold is the value of the above expression. This means that when the threshold is less than 0.2s, the output is zero; when the threshold is greater than 0.2s, the output is the original value. A time difference greater than 0.2s between adjacent pulses is considered a conduction block. The pulse pattern is judged based on the rhythm of the threshold greater than 0, that is, the regularity of the points where the threshold is greater than 0. Periodic peaks indicate intermittent pulse, irregular peaks indicate knotted pulse, and no peaks indicate a normal pulse rate. This is because a normal pulse rate is smooth and relatively regular, with little difference in the PP interval; therefore, for thresholds less than or equal to 0.2s, the output value is always 0. Intermittent pulse is a slow and regular intermittent pulse, such as... Figure 3 As shown in (b), a regular pulse rate delay occurs; the intermittent pulse is characterized by an indefinite number of beats, i.e., after a complete pulse beat, there is a pause or a small beat appears prematurely, followed by a complete or incomplete compensatory pause, such as... Figure 3 As shown in (c), irregular pulse rate delays and premature beats are observed. Specifically, this can be observed by plotting a pulse rate variability signal.
[0034] Because the pulse signal is not a strictly periodic sequence, the PP interval and They are not necessarily equal. In fact, the pulsation cycle of the human pulse signal exhibits minute variations, and the difference in the PP interval is known as pulse rate variability (PRV). This invention uses the differences in PRV to identify and distinguish normal pulse signals (pulse signals corresponding to healthy individuals), intermittent pulse signals (abnormal signals), and knotted pulse signals (abnormal signals).
[0035] Using a two-dimensional array to record the PP interval time series PP interval time series after thresholding For the PRV signal, establish... dimensional discrete time series The first row of this discrete time series is used to record the PP interval time series. The second row records the PP interval time series after thresholding. A pulse rate variability signal was plotted. Specifically, the thresholded PP interval was plotted on the ordinate (i.e., the threshold value was plotted on the ordinate), and the number of beats was plotted on the x-axis. As shown in the thresholding process above, intervals with a difference of less than 0.2 seconds between adjacent beats were defined as 0; otherwise, no change was observed. The results are as follows... Figure 4 As shown. Figure 4 for Figure 3 Pulse rate variability signal graph corresponding to pulse signal, Figure 4 (a) The pulse signal of a healthy person is shown as a straight line with all vertical axes being 0; Figure 4 (b) and Figure 4 (c) are the intermittent pulse signal and the knot pulse signal, respectively. Figure 4(b) The points with a threshold greater than 0 exhibit periodic peaks, which is regular. Figure 4 (c) The points with a threshold greater than 0 show irregular changes and are irregular.
[0036] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for identifying intermittent pulse abnormalities based on PP interval sequences, characterized in that, The method includes the following steps: S1, Obtain the sample to be tested; S2, extract the time information of the same feature points from continuous pulse signals to construct the corresponding PP interval time series; S3, perform thresholding on the PP interval time series, and obtain the pulse signal category based on the thresholding result; For the PP interval time series The threshold processing is performed to obtain a threshold, and the pulse signal category is determined based on the rhythm of the threshold; the expression for the threshold processing is: in, For the threshold value, 1 < k ≤ N-1, and It is the first The and the first The time values of each PP interval are numerically related to the PP interval time series. The corresponding values in the data are equal; If the threshold is greater than 0 and shows periodic peaks, it is an intermittent pulse; if the threshold is greater than 0 and shows irregular peaks, it is a knotted pulse; if the threshold is greater than 0 and shows a straight line with no peaks, it is a normal pulse rate. The step of extracting time information of the same feature points from continuous pulse signals and constructing corresponding PP interval time series includes: extracting the time information corresponding to the local maxima in each cardiac cycle of the pulse signal to obtain a time series array. ,in, Indicates the first Time information corresponding to the same feature points in each cardiac cycle; The step of extracting time information of the same feature points from continuous pulse signals and constructing the corresponding PP interval time series further includes: performing difference calculations on the time series array sequentially to obtain the PP interval time series. ,in, This represents the result of the difference calculation, and the expression for the difference calculation is: ,in, N The number of identical feature points in the time series array. Indicates the first The PP interval corresponding to each cardiac cycle and They represent the first The and the first Time information corresponding to the same feature points in each cardiac cycle.
2. The method for identifying arrhythmic pulses based on PP interval sequence pulse rate abnormalities according to claim 1, characterized in that, The acquisition of the test sample refers to the acquisition of the pulse signal at the radial artery of the wrist of the test subject, which is directly acquired through a pulse acquisition device; the acquired parameters include at least a one-dimensional time series, signal sampling rate, and measurement time.
3. The method for identifying arrhythmic pulses based on PP interval sequence pulse rate abnormalities according to claim 2, characterized in that, The measurement time of the pulse signal is greater than 10 minutes, the signal sampling rate is 1 kHz, and the one-dimensional time series refers to the numerical values representing the amplitude of the pulse signal vibration arranged in chronological order.
4. The method for identifying arrhythmic pulses based on PP interval sequence pulse rate abnormalities according to claim 3, characterized in that, The thresholding process for the PP interval time series and the determination of the pulse signal category based on the thresholding result further includes: plotting a pulse rate variability signal graph, using the threshold obtained after thresholding as the vertical axis and the number of beats as the horizontal axis.