An oscillography anti-interference algorithm based on adaptive threshold and waveform feature matching, a computer and a storage medium

Abstract

This invention discloses an oscilloscope-based anti-interference algorithm, a computer, and a storage medium based on adaptive threshold and waveform feature matching. The oscilloscope-based anti-interference algorithm includes: if a signal segment simultaneously satisfies the following conditions: its pulse period is higher than a period target threshold, its pulse amplitude is higher than an amplitude target threshold, and its waveform similarity to a standard pulse waveform template is lower than a similarity target threshold, then the signal segment is identified as an interference signal segment; and the interference signal segment is removed and filled with the average value of the normal pulse signal adjacent to the interference signal segment; wherein the period target threshold, the amplitude target threshold, and the similarity target threshold are all dynamic. The beneficial effect of this invention is that it avoids measurement deviations caused by the user speaking or slight upper arm movements, and has a better anti-interference effect.

Description

Technical Field

[0001] This invention relates to the field of dynamic blood pressure monitoring devices, and in particular, to an oscillometric anti-interference algorithm, a computer, and a storage medium based on adaptive threshold and waveform feature matching. Background Technology

[0002] The oscillometric method is the core measurement principle used in ambulatory blood pressure monitoring (ABPM). During deflation, the ABPM pressure sensor samples the raw oscillometric pulse signal, analyzes the pulse wave envelope characteristics, and then calculates systolic and diastolic blood pressure. The user's talking or slight arm movements are the most common interference factors for the oscillometric method and the main cause of measurement deviations. When speaking, the vibration of muscles in the throat and shoulder is transmitted to the cuff, generating interference waveforms with frequencies close to the pulse signal. When the upper arm moves slightly, muscle contraction and joint movement cause abnormal fluctuations in cuff pressure, generating false oscillation signals. These interference signals can mask the normal pulse signal, leading to algorithmic misinterpretation and affecting measurement results, thus causing difficulties in clinical diagnosis. Summary of the Invention

[0003] The technical problem to be solved by the present invention is that the user's speech or upper arm micro-movements interfere with the accuracy of blood pressure measurement. The present invention provides an oscilloscope-based anti-interference algorithm, a computer and a storage medium based on adaptive threshold and waveform feature matching.

[0004] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: an oscilloscope-based anti-interference algorithm based on adaptive threshold and waveform feature matching, comprising:

[0005] If a signal segment simultaneously satisfies the following conditions: its pulse period is higher than the period target threshold, its pulse amplitude is higher than the amplitude target threshold, and its waveform similarity to the standard pulse waveform template is lower than the similarity target threshold, then the signal segment is determined to be an interference signal segment; and,

[0006] The interference signal segments are removed and filled with the average value of the normal pulse signals adjacent to the interference signal segments;

[0007] Among them, the period target threshold, the amplitude target threshold, and the similarity target threshold are all dynamic.

[0008] The beneficial effects are at least as follows: it avoids measurement deviations caused by the user's speech or slight movements of the upper arm, and has a better anti-interference effect.

[0009] As a preferred embodiment of the oscilloscope-based anti-interference algorithm, the target thresholds for period, amplitude, and similarity are initially set based on the normal range of human pulse parameters. Specifically, the target threshold for period is initially set to 0.6-1.2 seconds, the target threshold for amplitude is initially set to 30%-70% of the maximum amplitude of the original oscilloscope pulse signal, and the target threshold for similarity is initially set to 0.75. If the average value of three consecutive pulse cycles deviates from the target threshold for period by ±10%, the target threshold for period is adjusted to ±5% of that average value. If the average value of three consecutive pulse amplitudes deviates from the target threshold for amplitude by ±15%, the target threshold for amplitude is adjusted to ±8% of that average value. If the SNR ≥ 20dB, the target threshold for similarity is adjusted to 0.75; if 15dB ≤ SNR < 20dB, the target threshold for similarity is adjusted to 0.65; and if SNR < 15dB, the target threshold for similarity is adjusted to 0.55. The beneficial effects are at least as follows: the threshold is automatically adjusted based on real-time pulse characteristics and interference intensity, adapting to people of different ages and blood pressure levels. Whether it is a slow pulse in the elderly or a rapid pulse in young people, it can be accurately adapted, as well as to scenarios with different interference intensities, without the need for manual parameter adjustment, making it highly practical.

[0010] As a preferred solution for the oscilloscope-based anti-interference algorithm, the standard pulse waveform template is established based on human pulse signal samples of different ages, genders, and blood pressure levels, representing three types of pulse waveforms: normal pulse, bradycardia, and tachycardia. The standard pulse waveform template is established using the K-means clustering algorithm.

[0011] As a preferred embodiment of the oscilloscope-based anti-interference algorithm, the padding is performed using time-segmented padding. Further, the first 1 / 3 of the interference signal segment is filled with the mean of the preceding signal, the middle 1 / 3 of the interference signal segment is filled with a weighted average of the means of the preceding and following signals, and the last 1 / 3 of the interference signal segment is filled with the mean of the following signal.

[0012] As a preferred solution for the oscilloscope-based anti-interference algorithm, an adaptive Kalman filter is used to denoise the original oscilloscope pulse signal. The filtering state equation is X(k) = A × X(k-1) + W(k), and the observation equation is Z(k) = H × X(k) + V(k), where A and H are both set to 1. The noise variance is updated in real time, and the update formula is adjusted to: variance(k) = 0.7 × variance(k-1) + 0.3 × |Z(k) - X(k)|. If the variance ≥ 0.01, the gain is set to 0.8-0.9; if the variance < 0.01, the gain is set to 0.5-0.6.

[0013] As a preferred solution for the oscilloscope-based anti-interference algorithm, the signal envelope curve is fitted using a double-exponential fitting formula. Further, the fitting formula is E(t) = A1 × e^(-αt) + A2 × e^(-βt), where E(t) is the envelope amplitude, t is time, and α and β are the fitting parameters. The parameters are solved using the least squares method.

[0014] Another technical solution of the present invention is as follows: A computer includes a processor and a memory, wherein the memory stores at least one instruction, which is loaded and executed by the processor to implement the operation performed by the oscilloscope anti-interference algorithm. The beneficial effects are at least as follows: it avoids measurement deviations caused by the user speaking or slight movements of the upper arm, and has a better anti-interference effect.

[0015] Another technical solution of the present invention is as follows: a storage medium storing at least one instruction, which is loaded and executed by a processor to implement the operation performed by the oscilloscope anti-interference algorithm described above. The beneficial effects include at least: avoiding measurement deviations caused by the user's speech or slight upper arm movements, and exhibiting better anti-interference performance.

[0016] In addition to the technical problems solved by the present invention, the technical features constituting the technical solutions, and the beneficial effects brought about by the technical features of these technical solutions as described above, other technical problems that the present invention can solve, other technical features contained in the technical solutions, and the beneficial effects brought about by these technical features will be further described in detail with reference to the accompanying drawings. Attached Figure Description

[0017] Figure 1 This is a flowchart of the oscilloscope-based anti-interference algorithm of the present invention.

[0018] Figure 2 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. It should be noted that these descriptions of embodiments are intended to aid in understanding the invention and do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0020] See Figure 1 The figure illustrates an oscillometric anti-interference algorithm based on adaptive threshold and waveform feature matching. This algorithm, without altering the existing hardware configuration of the ambulatory blood pressure monitor, achieves accurate identification and signal rejection of interference from user speech or upper arm micro-movements solely through software algorithms, thereby improving blood pressure measurement accuracy.

[0021] The oscilloscope-based anti-interference algorithm specifically includes:

[0022] Step S1, Oscilloscope Pulse Signal Acquisition: Receive the raw oscilloscope pulse signal acquired by the pressure sensor of the ambulatory blood pressure monitor. In practice, the sampling frequency is set to 100-200Hz, the sampling period is set to 30-60 seconds, and it is necessary to ensure that the acquired raw oscilloscope pulse signal is free from distortion caused by loose connections.

[0023] Step S2, Adaptive Noise Suppression Filtering: Adaptive Kalman filtering is used to denoise the raw oscilloscope pulse signal. Specifically, the filtering state equation is X(k) = A × X(k-1) + W(k), and the observation equation is Z(k) = H × X(k) + V(k), where A and H are both set to 1, W(k) is the process noise, and V(k) is the observation noise. The noise variance is updated in real-time using the following formula: variance(k) = 0.7 × variance(k-1) + 0.3 × |Z(k) - X(k)|. Testing has shown that the update formula effectively suppresses the influence of external electromagnetic interference on the filtering effect after multiple rounds of noise scenario testing, ensuring a stable and reliable filtering process without abnormal fluctuations. The filtering gain is adjusted according to the real-time signal variance. If the variance ≥ 0.01 (indicating strong interference), the gain is set to 0.8-0.9 to enhance the denoising effect. If the variance is less than 0.01 (indicating weak interference), the gain should be set to 0.5-0.6 to avoid over-filtering and loss of pulse characteristics, ensuring that the denoised signal retains the core pulse characteristics and does not affect the accuracy of subsequent measurements.

[0024] Step S3, Pulse Signal Feature Extraction: From the denoised oscilloscope pulse original signal, extract the pulse period (time interval between adjacent peaks), pulse amplitude (difference between peak and trough values), and waveform similarity (cosine similarity between the current signal and the standard pulse template). In practice, the feature extraction error is controlled within a fixed range (e.g., within 2%).

[0025] Step S4, Adaptive Threshold Dynamic Adjustment: includes sub-steps S41 and S42.

[0026] Sub-step S41, initial setting of target thresholds: Initially set the period target threshold, amplitude target threshold, and similarity target threshold based on the normal range of human pulse parameters. Specifically, the period target threshold is initially set to 0.6-1.2 seconds, the amplitude target threshold is initially set to 30%-70% of the maximum amplitude of the original oscilloscope pulse signal, and the similarity target threshold is initially set to 0.75. Testing has shown that these initial settings are suitable for the vast majority of users, avoiding misjudgments caused by unreasonable initial threshold settings.

[0027] Sub-step S42, dynamic adjustment of the target threshold: If the average value of three consecutive pulse cycles deviates from the target threshold of the cycle by ±10%, the target threshold of the cycle is adjusted to the average value ±5%. If the average value of three consecutive pulse amplitudes deviates from the target threshold of the amplitude by ±15%, the target threshold of the amplitude is adjusted to the average value ±8%. The similarity target threshold is dynamically adjusted according to the signal SNR: 0.75 when SNR ≥ 20dB (i.e., indicating weak interference), 0.65 when 15dB ≤ SNR < 20dB (i.e., indicating moderate interference), and 0.55 when SNR < 15dB (i.e., indicating strong interference), ensuring that the threshold adjustment does not cause algorithm abnormalities and can adapt to different groups of people and interference scenarios.

[0028] Step S5, waveform feature matching and interference identification: includes sub-steps 51, 52 and 53.

[0029] Sub-step 51: Obtaining the standard pulse waveform template. In practice, pulse signal samples from individuals of different ages, genders, and blood pressure levels are collected (e.g., 1000 groups of users). A K-means clustering algorithm is used to establish three classes, corresponding to normal pulse, bradycardia (pulse frequency < 60 beats / minute), and tachycardia (pulse frequency > 100 beats / minute), respectively. The K-means iteration count is 50, and the initial cluster center is selected from the sample mean. Testing has confirmed the template's stability and universality, with the template similarity error controlled within 2%.

[0030] Sub-step S52, Feature matching: Match the denoised oscilloscope pulse original signal with the standard pulse waveform template one by one, and calculate the waveform similarity.

[0031] Sub-step S53, interference determination: If a signal segment simultaneously meets the three conditions of pulse period being higher than the period target threshold, pulse amplitude being higher than the amplitude target threshold, and similarity with the standard pulse waveform template being lower than the similarity target threshold, then it is determined that there is speech or upper arm micro-movement interference and the signal segment is marked as an interference signal segment.

[0032] Step S6, Interference Signal Removal and Signal Filling: For interference signal segments, the signal is filled using the average of adjacent normal pulse signals. Specifically, the average of normal signals from an equal number (e.g., 30) sampling points before and after the interference signal segment is taken, and the signal is filled in time segments (e.g., the average of the preceding signal is used for the first 1 / 3 of the interference segment, the weighted average of the preceding and following signals is used for the middle 1 / 3, and the average of the following signal is used for the last 1 / 3). Testing shows that time-segmented filling can control the error within 4%, ensuring that the filled signal is continuous and consistent with the normal pulse signal, meeting the requirements for subsequent blood pressure calculation.

[0033] Step S7, Blood Pressure Calculation: First, fit the signal envelope curve using a double exponential fitting formula. Specifically, the fitting formula is determined as E(t) = A1 × e^(-αt) + A2 × e^(-βt) (where E(t) is the envelope amplitude, t is time, and α and β are fitting parameters). The parameters are solved using the least squares method. Then, based on the correspondence between the signal envelope curve and cuff pressure, calculate the systolic blood pressure, diastolic blood pressure, and mean arterial pressure. Finally, output the measurement results.

[0034] The table below compares the oscilloscope-based anti-interference algorithm with traditional algorithms.

[0035]

[0036] The oscilloscope-based anti-interference algorithm controls the measurement deviation within ±3.0 mmHg under interference scenarios, far below the ±5.0 mmHg limit specified in the new standard; the interference recognition accuracy is 97.8%, higher than the 95% requirement of the new standard, accurately identifying interference from speech and upper arm micro-movements, avoiding misjudgments; the electromagnetic interference suppression capability is ≥40dB, meeting the requirements for use in complex electromagnetic environments such as hospitals; the feature extraction error is 1.7%, lower than the limit of the new standard, providing reliable support for subsequent calculations; the software runs continuously for 72 hours without abnormalities, far exceeding the 48 hours required by the new standard, ensuring operational stability. The oscilloscope-based anti-interference algorithm complies with GB 9706.1-2020 and YY 9706.230-2023 standards.

[0037] See Figure 2 This application provides an electronic device. The electronic device includes a processor 1 and a memory 2. The processor 1 is connected to the memory 2. The processor 1 is used to execute a computer program stored in the memory 2, so that the electronic device executes the oscilloscope-based anti-interference algorithm.

[0038] In specific implementation, the processor 1 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0039] In specific implementation, the memory 2 includes various media capable of storing program code, such as ROM, RAM, magnetic disk, USB flash drive, memory card or optical disk.

[0040] This application provides a storage medium. The storage medium stores a computer program. When the computer program is executed by a processor, it implements the oscilloscope-based anti-interference algorithm.

[0041] The above description merely illustrates embodiments of the present invention and is quite specific and detailed; however, it should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. An oscilloscope-based anti-interference algorithm based on adaptive threshold and waveform feature matching, characterized in that, include: If a signal segment simultaneously satisfies the following conditions: its pulse period is higher than the period target threshold, its pulse amplitude is higher than the amplitude target threshold, and its waveform similarity to the standard pulse waveform template is lower than the similarity target threshold, then the signal segment is determined to be an interference signal segment; and the interference signal segment is removed and filled with the average value of the normal pulse signal adjacent to the interference signal segment; wherein, the period target threshold, the amplitude target threshold, and the similarity target threshold are all dynamic.

2. The oscilloscope-based anti-interference algorithm according to claim 1, characterized in that, The target threshold for the cycle is configured as follows: the target threshold for the cycle is initially set according to the normal range of human pulse parameters; then, if the average value of three consecutive pulse cycles deviates from the target threshold for the cycle by ±10%, the target threshold for the cycle is adjusted to the average value ±5%; further, the target threshold for the cycle is initially set to 0.6-1.2 seconds.

3. The oscilloscope-based anti-interference algorithm according to claim 1, characterized in that, The amplitude target threshold is configured as follows: the amplitude target threshold is initially set according to the normal human pulse parameter range; then, if the average value of three consecutive pulse amplitudes deviates from the amplitude target threshold by ±15%, the amplitude target threshold is adjusted to the average value ±8%; furthermore, the amplitude target threshold is initially set to 30%-70% of the maximum amplitude of the original oscilloscope pulse signal.

4. The oscilloscope-based anti-interference algorithm according to claim 1, characterized in that, The similarity target threshold is configured as follows: the similarity target threshold is initially set according to the normal human pulse parameter range; then, if SNR≥20dB, the similarity target threshold is adjusted to 0.75, if 15dB≤SNR<20dB, the similarity target threshold is adjusted to 0.65, and if SNR<15dB, the similarity target threshold is adjusted to 0.55; furthermore, the similarity target threshold is initially set to 0.

75.

5. The oscilloscope-based anti-interference algorithm according to any one of claims 1 to 4, characterized in that, The standard pulse waveform template is established based on human pulse signal samples of different ages, genders, and blood pressure levels, and includes three types of pulse waveform templates: normal pulse, bradycardia, and tachycardia. Furthermore, the standard pulse waveform template is established using the K-means clustering algorithm.

6. The oscilloscope-based anti-interference algorithm according to any one of claims 1 to 4, characterized in that, The padding is done in time-segmented manner; furthermore, the first 1 / 3 of the interference signal segment is filled with the mean of the preceding signal, the middle 1 / 3 of the interference signal segment is filled with a weighted average of the mean of the preceding and following signals, and the last 1 / 3 of the interference signal segment is filled with the mean of the following signal.

7. The oscilloscope-based anti-interference algorithm according to any one of claims 1 to 4, characterized in that, An adaptive Kalman filter is used to denoise the raw oscilloscope pulse signal. Further, the filtering state equation is X(k) = A × X(k-1) + W(k), and the observation equation is Z(k) = H × X(k) + V(k), where A and H are both set to 1. The noise variance is updated in real time, with the update formula adjusted as follows: variance(k) = 0.7 × variance(k-1) + 0.3 × |Z(k) - X(k)|. If the variance ≥ 0.01, the gain is set to 0.8-0.9; if the variance < 0.01, the gain is set to 0.5-0.

6.

8. The oscilloscope-based anti-interference algorithm according to any one of claims 1 to 4, characterized in that, The signal envelope curve is fitted using a double exponential fitting formula.

9. A computer, characterized in that, The computer includes a processor and a memory, the memory storing at least one instruction, which is loaded and executed by the processor to perform the operation of the oscilloscope anti-interference algorithm according to any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium stores at least one instruction, which is loaded and executed by a processor to perform the operation of the oscilloscope anti-interference algorithm according to any one of claims 1 to 8.

Images