Wearable heartbeat monitoring device and method thereof

Abstract

The invention relates to the technical field of computers, and provides a wearable heartbeat monitoring device and method, and the device comprises a central processing unit, an electrode plate unit, an R wave detection unit, a heartbeat frequency monitoring unit, and a heartbeat abnormity monitoring unit. The electrode slice unit is used for collecting electrocardiosignals of a user in real time; the R-wave detection unit is used for splitting the electrocardiosignal into a plurality of electrocardiosignal segments and carrying out R-wave feature point detection on each electrocardiosignal segment to obtain an R-wave feature point of each electrocardiosignal segment; the heartbeat frequency monitoring unit is used for determining a heartbeat cycle based on the time interval of the R-wave feature points of the two adjacent electrocardiosignal segments and determining the current heartbeat frequency of the user based on a plurality of continuous heartbeat cycles; and the abnormal heartbeat monitoring unit is used for comparing the current heartbeat frequency with a preset normal heartbeat frequency range and judging whether the user is in an abnormal heartbeat state or not. According to the embodiment of the invention, the heartbeat monitoring accuracy and reliability are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a wearable heartbeat monitoring device and a method thereof. Background Art

[0002] Currently, a heart rate monitoring method based on photoplethysmography (PPG) is widely used in wearable devices. This method uses an optical sensor on the wearable device to emit light to the skin, then receives the reflected or transmitted light signal and extracts heart rate information based on changes in the light signal. However, this method has a significant drawback: it is susceptible to motion artifacts and interference from ambient light. When a user exercises, the movement of the body causes the contact between the optical sensor and the skin to change, resulting in large fluctuations in the received light signal, seriously affecting the accuracy of heart rate monitoring. Furthermore, changes in the intensity of ambient light can interfere with the acquisition of the light signal, causing errors in the monitoring results. This makes it impossible to meet the user's demand for accurate heart rate monitoring in complex scenarios. Summary of the Invention

[0003] The present invention provides a wearable heartbeat monitoring device and method thereof, aiming to improve the accuracy and reliability of heartbeat monitoring.

[0004] In a first aspect, the present invention provides a wearable heartbeat monitoring device, comprising a central processing unit, an electrode unit, an R-wave detection unit, a heartbeat frequency monitoring unit, and a heartbeat abnormality monitoring unit; the central processing unit is connected to the electrode unit, the R-wave detection unit, the heartbeat frequency monitoring unit, and the heartbeat abnormality monitoring unit, respectively, to manage each unit; Electrode unit, used to collect the user's electrocardiogram signals in real time; An R wave detection unit, configured to split the ECG signal into a plurality of ECG signal segments, and perform R wave feature point detection on each ECG signal segment to obtain an R wave feature point of each ECG signal segment; A heart rate monitoring unit, configured to determine a heart rate cycle based on a time interval between R-wave feature points of two adjacent ECG signal segments, and to determine a user's current heart rate based on a plurality of consecutive heart rate cycles; The abnormal heartbeat monitoring unit is used to compare the current heartbeat frequency with a preset normal heartbeat frequency range to determine whether the user is in an abnormal heartbeat state.

[0005] In a second aspect, the present invention further provides a blood pressure monitoring method, which is implemented based on the wearable heart rate monitoring device described in the first aspect, and the blood pressure monitoring method includes: The user's ECG signals are collected in real time based on the electrode unit on the wearable device; Splitting the ECG signal into multiple ECG signal segments, and performing R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment; Determine a heartbeat cycle based on the time interval between R-wave feature points of two adjacent ECG signal segments, and determine the user's current heartbeat frequency based on multiple consecutive heartbeat cycles; The current heart rate is compared with a preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

[0006] According to the blood pressure monitoring method provided by the present invention, R wave feature point detection is performed on each ECG signal segment to obtain the R wave feature point of each ECG signal segment, including: For each ECG signal segment, based on a first ECG energy intensity feature in the range of 0.1-0.6 Hz, a second ECG energy intensity feature in the range of 0.6-1.5 Hz, a third ECG energy intensity feature in the range of 1.5-2.2 Hz, and a fourth ECG energy intensity feature in the range of 2.2-3 Hz of each ECG signal feature point, respectively determine a first R-wave peak feature point, a second R-wave peak feature point, a third R-wave peak feature point, and a fourth R-wave peak feature point; determining a target signal-to-noise ratio based on a first signal-to-noise ratio of the first segment of the ECG signal, a second signal-to-noise ratio of the middle ECG signal, and a third signal-to-noise ratio of the last segment of the ECG signal; An R wave feature point is determined based on the first R wave peak feature point, the second R wave peak feature point, the third R wave peak feature point, the fourth R wave peak feature point, and the target signal-to-noise ratio.

[0007] According to the blood pressure monitoring method provided by the present invention, based on the first electrocardiogram energy intensity characteristic of each electrocardiogram signal characteristic point in the range of 0.1-0.6 Hz, the first R wave peak characteristic point of each electrocardiogram signal segment is determined, including: For each ECG signal segment, the first ECG energy intensity feature of each ECG signal feature point in the range of 0.1-0.6 Hz is sampled in the time domain to obtain an energy intensity sequence. With each sampling point as the center, a window of preset length is selected before and after to construct an energy-time fluctuation matrix. Each matrix element in the energy-time fluctuation matrix represents the energy difference between the current point and the adjacent point. Determine the local energy gradient curvature based on the transverse gradient and longitudinal gradient of each position in the energy time fluctuation matrix; the transverse gradient represents the rate of change of energy on the time axis, and the longitudinal gradient represents the rate of change of the energy difference between adjacent time points; Traversing the energy intensity sequence, determining a target energy extreme point corresponding to an energy intensity greater than the energy intensities of a first preset number of points before and after the energy extreme point; Determine the first R wave peak feature point of each electrocardiogram signal segment based on the target energy extreme value point; The local energy gradient curvature The calculation formula is: in, Indicates the The lateral gradient at each position, Indicates the The longitudinal gradient at each position, represents the time derivative of the transverse gradient, It represents the time derivative of the longitudinal gradient, which characterizes the curvature of energy change.

[0008] According to the blood pressure monitoring method provided by the present invention, determining the first R wave peak feature point of each electrocardiogram signal segment based on the target energy extreme point includes: Calculating a curvature response value based on the local energy gradient curvature of the target energy extreme point; constructing an adjacent extreme point association network based on the time interval and curvature response value difference between each first energy extreme point in the target energy extreme point and a second preset number of second energy extreme points adjacent to it; Determining an importance index based on an edge weight of each first energy extreme value point in the adjacent extreme value point association network and a curvature response value of its corresponding second energy extreme value point in the adjacent extreme value point association network; The first R-wave peak feature point of each electrocardiogram signal segment is determined based on the importance index of each first energy extreme value point.

[0009] In one embodiment, the calculation formula of the curvature response value of the target energy extreme point is as follows: in, Indicates the The curvature response value of the target energy extreme point, Indicates the The energy intensity of the target energy extreme point, represents the average energy intensity of the energy intensity sequence, Represents the energy intensity standard deviation of the energy intensity series.

[0010] In one embodiment, the edge weight calculation formula of each first energy extreme value point in the adjacent extreme value point association network is as follows: in, Indicates the The edge weight of the first energy extreme point in the network associated with the adjacent extreme points, Indicates the The first energy extreme point and its adjacent The time interval between the second energy extreme points, Indicates the The first energy extreme point and its adjacent The difference in curvature response values ​​between the second energy extreme points, Indicates the difference threshold of curvature response values.

[0011] According to the blood pressure monitoring method provided by the present invention, determining the first R wave peak feature point of each electrocardiogram signal segment based on the importance index of each first energy extreme value point includes: For each ECG signal segment, the first energy extreme point whose importance index is greater than or equal to the preset index threshold is determined as the preliminary R-wave peak feature point: For each R-peak feature point in the preliminary R-peak feature points, calculating the second-order derivative of the local energy gradient curvature corresponding to the R-peak feature point in the energy-time fluctuation matrix; Traverse each second-order derivative and determine the target R peak feature point where the second-order derivative changes from positive to negative; If the target R peak feature point belongs to the preliminary R peak feature point and the number of the target R peak feature point is 1, the target R peak feature point is determined as the first R peak feature point; If the target R peak feature point belongs to the preliminary R peak feature point, and the number of the target R peak feature points is greater than 1, the target R peak feature point corresponding to the importance index with the largest value is determined as the first R peak feature point.

[0012] In a third aspect, the present invention further provides an electronic device comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby implementing any of the above-mentioned blood pressure monitoring methods.

[0013] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, wherein the storage medium stores a computer software program, and when the computer software program is executed by a processor, it implements any of the blood pressure monitoring methods described above.

[0014] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which implements any of the blood pressure monitoring methods described above when executed by a processor.

[0015] The wearable heart rate monitoring device provided by the embodiments of the present invention, on the one hand, detects R-wave feature points based on the electrophysiological characteristics of the electrocardiogram (ECG) signal itself. These electrophysiological features have clear physiological significance and stability and are not affected by ambient light. Therefore, regardless of how the ambient light changes, as long as a valid ECG signal can be collected, the R-wave feature points can be accurately detected, and the heart rate cycle can be calculated. Furthermore, determining the heart rate using multiple consecutive heart rate cycles can effectively reduce the potential errors in a single interval, improving the accuracy and reliability of the heart rate calculation, and thus accurately and reliably determining whether the user is experiencing an abnormal heart rate, thereby improving the accuracy and reliability of the heart rate monitoring device. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 1 is a schematic structural diagram of a wearable heart rate monitoring device provided by the present invention; Figure 2 1 is a flow chart of the blood pressure monitoring method provided by the present invention; Figure 3 An embodiment diagram of an electronic device provided by an embodiment of the present invention; Figure 4 An embodiment diagram of a computer-readable storage medium provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0018] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0019] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0020] Optional, see Figure 1 As shown, Figure 1 This is a structural diagram of the wearable heartbeat monitoring device provided by the present invention. The wearable heartbeat monitoring device includes a central processing unit, an electrode unit, an R-wave detection unit, a heart rate monitoring unit, and a heartbeat abnormality monitoring unit.

[0021] Optionally, the central processing unit in the embodiment of the present invention is respectively connected to the electrode unit, the R-wave detection unit, the heart rate monitoring unit and the heartbeat abnormality monitoring unit to manage each unit.

[0022] Optionally, the electrode unit collects the user's electrocardiogram signal in real time.

[0023] Optionally, the R wave detection unit splits the ECG signal into multiple ECG signal segments, and performs R wave feature point detection on each ECG signal segment to obtain the R wave feature point of each ECG signal segment.

[0024] Optionally, the heart rate monitoring unit determines the heart rate cycle based on the time interval between R-wave feature points of two adjacent electrocardiogram signal segments, and determines the user's current heart rate based on multiple consecutive heart rate cycles.

[0025] Optionally, the abnormal heartbeat monitoring unit compares the current heartbeat frequency with a preset normal heartbeat frequency range to determine whether the user is in an abnormal heartbeat state.

[0026] The embodiment of the present invention detects R-wave feature points based on the electrophysiological characteristics of the electrocardiogram signal itself. These electrophysiological features have clear physiological significance and stability and are not affected by ambient light. Therefore, no matter how the ambient light changes, as long as a valid electrocardiogram signal can be collected, the R-wave feature points can be accurately detected and the heartbeat cycle can be calculated. Furthermore, determining the heart rate through multiple consecutive heartbeat cycles can effectively reduce the errors that may exist in a single interval, improve the accuracy and reliability of the heart rate calculation, and thus accurately and reliably determine whether the user is in an abnormal heartbeat state, thereby improving the accuracy and reliability of the heartbeat monitoring device.

[0027] Reference Figure 2 , Figure 2 FIG2 is a flow chart of the blood pressure monitoring method provided by the present invention. In the embodiment of the present invention, the execution subject of the blood pressure monitoring method is a wearable device. Therefore, the blood pressure monitoring method includes: Step 10: Collect the user's electrocardiogram signal in real time based on the electrode unit on the wearable device.

[0028] Optionally, the electrode unit carried on a wearable device (such as a smartwatch, smart bracelet, etc.) is usually made of conductive materials and can be in close contact with the user's skin to form a circuit loop. Therefore, when the heart performs electrical activity, it generates weak bioelectric signals, which are transmitted to the skin surface through human tissue. The electrode unit senses the potential changes on the skin surface and converts them into electrical signals. After processing by the signal conditioning circuit (such as amplification, filtering, etc.), it forms an electrocardiogram signal that can be recognized by the device. The electrocardiogram signal is output in real time in the form of a time series, reflecting the entire process of the heart's electrical activity.

[0029] In one embodiment, a smartwatch is used as an example. Two stainless steel metal electrodes are installed on its back. When the user wears the watch, the electrodes adhere tightly to the skin on the wrist. The watch integrates a signal acquisition circuit, including an instrumentation amplifier, a low-pass filter, and an analog-to-digital converter (ADC). Bioelectric signals generated by cardiac electrical activity are transmitted to the surface of the wrist skin, generating a potential difference between the two electrodes. The instrumentation amplifier amplifies this potential difference (by approximately 1000 times), a low-pass filter filters out high-frequency noise (with a cutoff frequency set at 100 Hz), and the ADC samples the filtered signal at a sampling rate of 1000 Hz, converting the analog signal into a digital ECG signal, which is stored in the watch's memory in real time.

[0030] Step 20: Split the ECG signal into multiple ECG signal segments, and perform R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment.

[0031] Furthermore, because ECG signals are continuous time series, they need to be split into multiple segments of fixed duration to facilitate processing and improve detection efficiency. Typically, the duration of each segment (e.g., 5 seconds, 10 seconds, etc.) is set based on the characteristics of the ECG signal and subsequent processing requirements. Adjacent segments can overlap to a certain extent (e.g., 1 second) to avoid feature loss due to segmentation.

[0032] Furthermore, the R wave is the peak with the largest amplitude and the most obvious characteristics in the ECG signal. The accuracy of its detection directly affects the calculation of subsequent heartbeat cycle and frequency. Among them, the detection method is usually based on the electrophysiological characteristics of the ECG signal, such as amplitude, slope, second-order derivative, etc.

[0033] Therefore, R wave feature point detection is performed on each ECG signal segment to obtain the R wave feature point of each ECG signal segment, as specifically described in the process from step 201 to step 203 .

[0034] Step 30 : determining a heartbeat cycle based on a time interval between R-wave feature points of two adjacent ECG signal segments, and determining the user's current heartbeat frequency based on a plurality of consecutive heartbeat cycles.

[0035] Furthermore, the time interval between two adjacent R-wave feature points is a heartbeat cycle (RR interval). Because the ECG signal is split into multiple segments, when processing adjacent segments, it is necessary to associate the last R-wave feature point of the previous segment with the first R-wave feature point of the next segment and calculate the time interval between them as a heartbeat cycle.

[0036] Furthermore, to reduce potential errors in individual heartbeat cycles (such as individual RR interval anomalies caused by motion, noise, and other factors), multiple consecutive heartbeat cycles (such as 5 or 10) are typically selected, their average calculated, and then converted to beats per minute (i.e., heart rate). The specific conversion formula is: heart rate (beats / minute) = 60 / average heartbeat cycle (seconds).

[0037] In one embodiment, assuming that the time of the last R-wave feature point of the previous ECG signal segment is 10.8s, and the time of the first R-wave feature point of the next segment is 12.0s, the time interval of the R-wave feature points is 1.2s, that is, the heartbeat cycle is 1.2 seconds. Continuing to process, the next few heartbeat cycles are 1.1s, 1.2s, 1.0s, 1.1s, and 1.2s respectively. Select these 5 consecutive heartbeat cycles and calculate the average value: (1.2+1.1+1.2+1.0+1.1) / 5=1.12 seconds. The current heart rate is: 60 / 1.12≈53.57 times / minute.

[0038] Step 40: Compare the current heart rate with a preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

[0039] Furthermore, the preset normal heart rate range is usually set according to different groups of people (such as age, gender, physical condition, etc.). Generally speaking, the normal resting heart rate range for adults is 60-100 beats / minute, but the normal heart rate for athletes may be lower (such as around 50 beats / minute). The wearable device will store the normal heart rate range for different users, or allow the user to set it according to their own situation. The current heart rate calculated in step 30 is compared with the normal range: if the current heart rate is less than the lower limit of the normal range or greater than the upper limit of the normal range, it is judged that the user is in an abnormal heartbeat state (such as bradycardia or tachycardia); otherwise, it is judged to be in a normal state.

[0040] Continuing with the above embodiment, the wearable device presets a normal heart rate range of 60-100 beats / minute for an average adult user. The current heart rate calculated in step 30 is 53.57 beats / minute, which is less than the lower limit of the normal range of 60 beats / minute. Therefore, the user is judged to be in an abnormal heartbeat state (bradycardia). At this point, the wearable device will alert the user through vibration, screen prompts, etc., and record the abnormality in historical data for the user to review later or seek medical reference. If the current heart rate is 75 beats / minute, which is within the normal range, the user's heartbeat is judged to be normal, and no alarm is issued.

[0041] The embodiment of the present invention detects R-wave feature points based on the electrophysiological characteristics of the electrocardiogram signal itself. These electrophysiological features have clear physiological significance and stability and are not affected by ambient light. Therefore, no matter how the ambient light changes, as long as a valid electrocardiogram signal can be collected, the R-wave feature points can be accurately detected and the heartbeat cycle can be calculated. Furthermore, determining the heart rate through multiple consecutive heartbeat cycles can effectively reduce the errors that may exist in a single interval, improve the accuracy and reliability of the heart rate calculation, and thus accurately and reliably determine whether the user is in an abnormal heartbeat state, thereby improving the accuracy and reliability of the heartbeat monitoring device.

[0042] In one embodiment, steps 201 to 203 are described as follows: Step 201, for each ECG signal segment, based on the first ECG energy intensity feature in the range of 0.1-0.6 Hz, the second ECG energy intensity feature in the range of 0.6-1.5 Hz, the third ECG energy intensity feature in the range of 1.5-2.2 Hz, and the fourth ECG energy intensity feature in the range of 2.2-3 Hz of each ECG signal feature point, respectively determine the first R-wave peak feature point, the second R-wave peak feature point, the third R-wave peak feature point, and the fourth R-wave peak feature point.

[0043] Optionally, in the field of bioelectric signals, different frequency components of the ECG signal correspond to different characteristics of cardiac electrophysiological activity. The 0.1-0.6 Hz frequency band mainly reflects the baseline drift and slowly changing components of the ECG signal, the 0.6-1.5 Hz frequency band is related to the low-frequency characteristics of the QRS complex, the 1.5-2.2 Hz frequency band corresponds to the main energy distribution of the R wave, and the 2.2-3 Hz frequency band contains the high-frequency details of the rising edge of the R wave. For each ECG signal segment, determine the first ECG energy intensity feature in the range of 0.1-0.6 Hz, the second ECG energy intensity feature in the range of 0.6-1.5 Hz, the third ECG energy intensity feature in the range of 1.5-2.2 Hz, and the fourth ECG energy intensity feature in the range of 2.2-3 Hz for each ECG signal feature point.

[0044] Furthermore, according to the first ECG energy intensity feature, the second ECG energy intensity feature, the third ECG energy intensity feature and the fourth ECG energy intensity feature, the first R-wave peak feature point, the second R-wave peak feature point, the third R-wave peak feature point and the fourth R-wave peak feature point are respectively determined. The process principle of determining the corresponding R-wave peak feature point for each ECG energy intensity feature is the same. Therefore, the embodiment of the present invention is explained in detail by taking the determination of the first R-wave peak feature point according to the first ECG energy intensity feature as an example, specifically the process from step 2011 to step 2014.

[0045] Step 202 : determining a target signal-to-noise ratio based on a first signal-to-noise ratio of the first segment of the ECG signal, a second signal-to-noise ratio of the middle ECG signal, and a third signal-to-noise ratio of the last segment of the ECG signal.

[0046] Furthermore, the signal-to-noise ratio (SNR) reflects the ratio of effective components to noise in an ECG signal, which directly affects the reliability of feature point detection. In an embodiment of the present invention, the starting part (first signal segment), the middle part (middle signal), and the ending part (last signal segment) of an ECG signal segment are selected to calculate the first signal-to-noise ratio of the first ECG signal segment, the second signal-to-noise ratio of the middle ECG signal segment, and the third signal-to-noise ratio of the last ECG signal segment, respectively. Taking into account the possible edge effects of the signal during acquisition (such as noise in the starting segment caused by loose wearing), the embodiment of the present invention uses the weighting of the three SNR values ​​to highlight the signal in the middle stable segment, while suppressing the influence of possible abnormal noise at the beginning and end, and finally obtaining the target signal-to-noise ratio.

[0047] In one embodiment, in a 10-second ECG signal segment, the first 3 seconds (0-3s) are the first ECG signal segment, the middle 4 seconds (3-7s) are the second ECG signal segment, and the last 3 seconds (7-10s) are the third ECG signal segment. For the first ECG signal segment, assuming the effective signal power is 8mV² and the noise power is 2mV², ; The effective signal power in the middle section is 12mV², and the noise power is 1mV². ; The effective signal power of the last segment is 9mV², and the noise power is 3mV². The target signal-to-noise ratio calculation of the embodiment of the present invention adopts a nonlinear function Mapping the three SNRs highlights the differences in high SNR areas. The calculation formula is:

[0048] Substitute the numerical calculation:

[0049] Step 203 : Determine an R wave feature point based on the first R wave peak feature point, the second R wave peak feature point, the third R wave peak feature point, the fourth R wave peak feature point, and the target signal-to-noise ratio.

[0050] Furthermore, the R-wave peak feature points in different frequency bands reflect the characteristics of the R wave in different frequency dimensions, while the target signal-to-noise ratio characterizes the overall signal quality. By constructing a multidimensional feature fusion model, the time coordinates of the R-wave peak feature points in the four frequency bands are jointly optimized with the target signal-to-noise ratio. Specifically, the four feature points are regarded as candidate points, the time distance difference between each candidate point and other candidate points is calculated, and an energy function is constructed in combination with the target signal-to-noise ratio. The final R-wave feature point is determined by solving the minimum value of the energy function, so that this point has the best consistency among the characteristics of each frequency band and adapts to the current signal noise level.

[0051] Continuing with the above embodiment, the time of the four R peak characteristic points is: 、 , target signal-to-noise ratio , construct the energy function: Among them, the weight According to the correlation between each frequency band and R wave: 、 .

[0052] for .

[0053] It is clear that when , The minimum, so the final R wave characteristic point is determined to be 3.4s (the 3400th sampling point).

[0054] The embodiment of the present invention utilizes the complementarity of the electrophysiological characteristics of ECG signals in different frequency bands. The 0.1-0.6Hz frequency band suppresses the influence of baseline drift, the 1.5-2.2Hz frequency band captures the main energy of the R wave, the 2.2-3Hz frequency band enhances the details of the rising edge, and the 0.6-1.5Hz frequency band assists in locating the range of the QRS complex. The joint optimization of the feature points of the four frequency bands can effectively resist the interference of noise in a single frequency band. On the other hand, based on the nonlinear fusion model of the signal segment SNR, the detection algorithm's sensitivity to noise is dynamically adjusted, improving detection accuracy in high SNR scenarios and suppressing false peaks in low SNR scenarios. Ultimately, the detection error of the R wave feature points is reduced to less than 50ms, providing a more reliable time reference for subsequent heart cycle calculations.

[0055] In one embodiment, steps 2011 to 2014 are described as follows: In step 2011, for each ECG signal segment, the first ECG energy intensity feature within the 0.1-0.6 Hz range of each ECG signal feature point is sampled in the time domain to obtain an energy intensity sequence. An energy-time fluctuation matrix is ​​constructed by selecting a window of a preset length before and after each sampling point. Each matrix element in the energy-time fluctuation matrix represents the energy difference between the current point and an adjacent point.

[0056] Optionally, when processing ECG signals collected by the wearable device, time-domain sampling of the first ECG energy intensity feature in the 0.1-0.6 Hz frequency band is performed, discretizing the continuous energy feature into an energy intensity sequence. A window of preset length (e.g., five sampling points before and after each sampling point) is selected around each sampling point to construct an energy-time fluctuation matrix. Each element in the matrix represents the energy difference between the current point and the adjacent points. This captures subtle energy variations over time, forming a two-dimensional matrix structure.

[0057] In one embodiment, taking a 10-second ECG signal segment (sampling rate 1000 Hz) collected by a smartwatch as an example, the energy intensity sequence is calculated in the 0.1-0.6 Hz frequency band. Taking the 3200th sampling point as the center, select a window of 5 points before and after, and construct an 11×11 energy-time fluctuation matrix M: , , for example, the matrix elements express , reflecting the energy difference between the two sampling points.

[0058] Step 2012: Determine the local energy gradient curvature based on the transverse gradient and longitudinal gradient at each position in the energy time fluctuation matrix. The transverse gradient represents the rate of change of energy on the time axis, and the longitudinal gradient represents the rate of change of the energy difference between adjacent time points.

[0059] Furthermore, the local energy gradient curvature is determined based on the transverse gradient and longitudinal gradient of each position in the energy time fluctuation matrix, where the transverse gradient represents the rate of change of energy on the time axis, and the longitudinal gradient represents the rate of change of the energy difference between adjacent time points. The calculation formula is: in, Indicates the The lateral gradient at each position, Indicates the The longitudinal gradient at each position, represents the time derivative of the transverse gradient, It represents the time derivative of the longitudinal gradient, which characterizes the curvature of energy change.

[0060] Step 2013 , traverse the energy intensity sequence to determine a target energy extreme point whose energy intensity is greater than the energy intensities of a first preset number of points before and after it.

[0061] Furthermore, the energy intensity sequence is traversed to find an energy extreme point that meets specific conditions. In this embodiment of the present invention, the target energy extreme point must have an energy intensity greater than the energy intensities of a first preset number of points before and after it (e.g., three points before and after it). This multi-neighborhood comparison method effectively avoids false peaks caused by noise and improves the accuracy of extreme point detection.

[0062] In one embodiment, in the energy intensity sequence In the point , check whether it satisfies: , assuming that in the sequence, and satisfy , but Determine the target energy extreme point.

[0063] Step 2014: determine the first R-wave peak feature point of the ECG signal segment based on the target energy extreme value point.

[0064] Furthermore, the first R-wave peak feature point of each ECG signal segment is determined according to the target energy extreme value point, as specifically described in the process from step 20141 to step 20144 .

[0065] The embodiment of the present invention captures the subtle changes in energy in the time domain through the energy-time fluctuation matrix. The horizontal and vertical gradient analysis describes the energy change trend from different angles, and the local energy gradient curvature quantifies the degree of curvature of the change. Therefore, it can effectively highlight the characteristics of the R-wave peak, improve the detection accuracy of the R-wave peak feature points, and provide a more accurate time reference for subsequent heartbeat cycle calculations.

[0066] In one embodiment, steps 20141 to 20144 are described as follows: Step 20141, calculating the curvature response value based on the local energy gradient curvature of the target energy extreme point.

[0067] Furthermore, for each ECG signal segment, the corresponding curvature response value is calculated according to the local energy gradient curvature of each target energy extreme point. The calculation formula of the curvature response value of the target energy extreme point is as follows: in, Indicates the The curvature response value of the target energy extreme point, Indicates the The energy intensity of the target energy extreme point, represents the average energy intensity of the energy intensity sequence, Represents the energy intensity standard deviation of the energy intensity series.

[0068] Step 20142: construct an adjacent extreme point association network based on the time interval and curvature response value difference between each first energy extreme point in the target energy extreme point and a second preset number of second energy extreme points adjacent to it.

[0069] Furthermore, based on the time interval and curvature response value difference between each first energy extreme point in the target energy extreme point and a second preset number of second energy extreme points adjacent to it, an adjacent extreme point association network is constructed, wherein the edge weight calculation formula of each first energy extreme point in the adjacent extreme point association network is as follows: ;in, Indicates the The first energy extreme point and its adjacent The time interval between the second energy extreme points, Indicates the The first energy extreme point and its adjacent The difference in curvature response values ​​between the second energy extreme points, Indicates the difference threshold of curvature response values.

[0070] Step 20143: Determine an importance index based on the edge weight of each first energy extreme value point in the adjacent extreme value point association network and the curvature response value of its corresponding second energy extreme value point in the adjacent extreme value point association network.

[0071] Furthermore, the importance index is determined based on the edge weight of each first energy extreme point in the adjacent extreme point association network and the curvature response value of its corresponding second energy extreme point in the adjacent extreme point association network. The specific formula for the importance index of each first energy extreme point is: .

[0072] in, Indicates the The importance index of the first energy extreme point, For the The set of all second energy extreme value points connected to the first energy extreme value points.

[0073] Step 20144: determine the first R-wave peak feature point of each ECG signal segment based on the importance index of each first energy extreme value point.

[0074] Furthermore, the first R-wave peak feature point of each electrocardiogram signal segment is determined according to the importance index of each first energy extreme value point, as specifically described in steps 201441 to 201445.

[0075] The nonlinear transformation of the curvature response value in the embodiment of the present invention enhances the distinguishability of features, the association network captures the spatiotemporal correlation between R-peaks, and the importance index evaluation comprehensively considers local and global features, thereby improving the detection robustness of R-peak feature points and enhancing the reliability of ECG monitoring of wearable devices.

[0076] In one embodiment, steps 201441 to 201445 are described as follows: Step 201441: For each ECG signal segment, the first energy extreme value point whose importance index is greater than or equal to a preset index threshold is determined as a preliminary R-wave peak feature point.

[0077] Optionally, in wearable ECG signal processing, the importance index reflects the likelihood of each extreme point being an R-peak. By setting a preset index threshold (e.g., 1.2), extreme points with an importance index greater than or equal to the threshold are selected as preliminary R-peak feature points. This effectively eliminates low-importance noise points and retains the most likely R-peak candidate points.

[0078] In one embodiment, the importance index of multiple extreme points is: The preset index threshold is 1.2, and the extreme point that meets the conditions is , so these three points are determined as preliminary R-peak feature points.

[0079] Step 201442: For each R-peak feature point in the preliminary R-peak feature points, calculate the second-order derivative of the local energy gradient curvature corresponding to it in the energy-time fluctuation matrix.

[0080] Furthermore, for each preliminary R-peak feature point, the second-order derivative of the local energy gradient curvature is calculated at the corresponding energy-time fluctuation matrix location. The second-order derivative reflects the acceleration of the curvature change and can identify the inflection point of the curvature change. At the R-peak location, the local energy gradient curvature typically increases first and then decreases, and its second-order derivative changes from positive to negative. This characteristic can be used to accurately locate the apex of the R-peak.

[0081] The embodiment of the present invention uses the central difference method to calculate the second-order derivative, which will not be described in detail here.

[0082] Step 201443, traverse each second-order derivative and determine the target R peak feature point where the second-order derivative changes from positive to negative.

[0083] Next, we traverse the second-order derivative sequence of each preliminary R-peak feature point, searching for the point where the second-order derivative changes from positive to negative. This point corresponds to the inflection point of the curvature change, typically the apex of the R-peak. By examining the sign change of the second-order derivative, we can accurately identify the location of the R-peak and eliminate false peaks.

[0084] Continuing with the above embodiment, for the preliminary R peak feature point , calculate its second-order derivative sequence: . . .

[0085] therefore, It is determined as the target R peak feature point.

[0086] Step 201444: If the target R-peak feature point belongs to the preliminary R-peak feature point and the number of the target R-peak feature point is 1, the target R-peak feature point is determined as the first R-peak feature point.

[0087] Furthermore, if the target R-peak feature point belongs to the preliminary R-peak feature point and the number is 1, then this point is directly determined as the first R-peak feature point. This situation indicates that in the ECG signal segment, only one point satisfies both the importance index threshold and the curvature change condition and is the most likely R-peak location.

[0088] In one embodiment, it is assumed that in a certain ECG signal segment, the initial R wave peak feature point is , and its second-order derivative changes from positive to negative, that is, the target R peak feature point is only A point. At this time, directly A point. At this time, directly Determined as the first R peak feature point.

[0089] Step 201445: If the target R peak feature point belongs to the preliminary R peak feature point and the number of the target R peak feature points is greater than 1, the target R peak feature point corresponding to the largest importance index is determined as the first R peak feature point.

[0090] Furthermore, if the target R-peak feature point is one of the preliminary R-peak feature points and the number is greater than 1, the importance indexes of these points are compared and the point with the largest value is selected as the first R-peak feature point. This method can prioritize the most representative and reliable points among multiple candidate points, improving the accuracy of R-peak detection.

[0091] Continuing in step 201443, the target R peak feature point is determined to be and . Compare their importance index, Therefore, Determined as the first R peak feature point.

[0092] In the embodiment of the present invention, threshold screening eliminates low-quality candidate points, second-order derivative analysis accurately locates the inflection point of curvature change, and importance index comparison selects the optimal solution among multiple candidate points. Therefore, the multi-level decision-making mechanism achieves high-precision positioning of the R-wave peak feature point, thereby improving the reliability of ECG monitoring of wearable devices.

[0093] See also Figure 3 , Figure 3 This is a diagram of an embodiment of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented: The user's ECG signals are collected in real time based on the electrode unit on the wearable device; Splitting the ECG signal into multiple ECG signal segments, and performing R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment; Determine a heartbeat cycle based on the time interval between R-wave feature points of two adjacent ECG signal segments, and determine the user's current heartbeat frequency based on multiple consecutive heartbeat cycles; Compare the current heart rate with the preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

[0094] See also Figure 4 , Figure 4 Detailed description of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. Figure 4As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, the following steps are implemented: The user's ECG signals are collected in real time based on the electrode unit on the wearable device; Splitting the ECG signal into multiple ECG signal segments, and performing R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment; Determine a heartbeat cycle based on the time interval between R-wave feature points of two adjacent ECG signal segments, and determine the user's current heartbeat frequency based on multiple consecutive heartbeat cycles; Compare the current heart rate with the preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

[0095] In another aspect, the present invention further provides a computer program product, which includes a computer program. The computer program may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the blood pressure monitoring method provided by the above methods. The blood pressure monitoring method includes: The user's ECG signals are collected in real time based on the electrode unit on the wearable device; Splitting the ECG signal into multiple ECG signal segments, and performing R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment; Determine a heartbeat cycle based on the time interval between R-wave feature points of two adjacent ECG signal segments, and determine the user's current heartbeat frequency based on multiple consecutive heartbeat cycles; Compare the current heart rate with the preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

[0096] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0097] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A wearable heart rate monitoring device, characterized in that: It includes a central processing unit, an electrode unit, an R-wave detection unit, a heart rate monitoring unit, and an abnormal heartbeat monitoring unit; the central processing unit is connected to the electrode unit, the R-wave detection unit, the heart rate monitoring unit, and the abnormal heartbeat monitoring unit, respectively, to manage each unit; Electrode unit, used to collect the user's electrocardiogram signals in real time; An R wave detection unit, configured to split the ECG signal into a plurality of ECG signal segments, and perform R wave feature point detection on each ECG signal segment to obtain an R wave feature point of each ECG signal segment; A heart rate monitoring unit, configured to determine a heart rate cycle based on a time interval between R-wave feature points of two adjacent ECG signal segments, and to determine a user's current heart rate based on a plurality of consecutive heart rate cycles; The abnormal heartbeat monitoring unit is used to compare the current heartbeat frequency with a preset normal heartbeat frequency range to determine whether the user is in an abnormal heartbeat state.

2. A blood pressure monitoring method, implemented based on the wearable heart rate monitoring device according to claim 1, characterized in that: The blood pressure monitoring method comprises: The user's ECG signals are collected in real time based on the electrode unit on the wearable device; Splitting the ECG signal into multiple ECG signal segments, and performing R-wave feature point detection on each ECG signal segment to obtain the R-wave feature point of each ECG signal segment; Determine a heartbeat cycle based on the time interval between R-wave feature points of two adjacent ECG signal segments, and determine the user's current heartbeat frequency based on multiple consecutive heartbeat cycles; The current heart rate is compared with a preset normal heart rate range to determine whether the user is in an abnormal heart rate state.

3. The blood pressure monitoring method according to claim 2, wherein: The detecting of R wave feature points of each ECG signal segment to obtain the R wave feature points of each ECG signal segment includes: For each ECG signal segment, based on a first ECG energy intensity feature in the range of 0.1-0.6 Hz, a second ECG energy intensity feature in the range of 0.6-1.5 Hz, a third ECG energy intensity feature in the range of 1.5-2.2 Hz, and a fourth ECG energy intensity feature in the range of 2.2-3 Hz of each ECG signal feature point, respectively determine a first R-wave peak feature point, a second R-wave peak feature point, a third R-wave peak feature point, and a fourth R-wave peak feature point; determining a target signal-to-noise ratio based on a first signal-to-noise ratio of the first segment of the ECG signal, a second signal-to-noise ratio of the middle ECG signal, and a third signal-to-noise ratio of the last segment of the ECG signal; An R wave feature point is determined based on the first R wave peak feature point, the second R wave peak feature point, the third R wave peak feature point, the fourth R wave peak feature point, and the target signal-to-noise ratio.

4. The blood pressure monitoring method according to claim 3, wherein: Determining a first R-wave peak feature point of each ECG signal segment based on a first ECG energy intensity feature of each ECG signal feature point in the range of 0.1-0.6 Hz includes: For each ECG signal segment, the first ECG energy intensity feature of each ECG signal feature point in the range of 0.1-0.6 Hz is sampled in the time domain to obtain an energy intensity sequence. With each sampling point as the center, a window of preset length is selected before and after to construct an energy-time fluctuation matrix. Each matrix element in the energy-time fluctuation matrix represents the energy difference between the current point and the adjacent point. Determine the local energy gradient curvature based on the transverse gradient and longitudinal gradient of each position in the energy time fluctuation matrix; the transverse gradient represents the rate of change of energy on the time axis, and the longitudinal gradient represents the rate of change of the energy difference between adjacent time points; Traversing the energy intensity sequence, determining a target energy extreme point corresponding to an energy intensity greater than the energy intensities of a first preset number of points before and after the energy extreme point; Determine the first R wave peak feature point of each electrocardiogram signal segment based on the target energy extreme value point; The local energy gradient curvature The calculation formula is: in, Indicates the The lateral gradient at each position, Indicates the The longitudinal gradient at each position, represents the time derivative of the transverse gradient, It represents the time derivative of the longitudinal gradient, which characterizes the curvature of energy change.

5. The blood pressure monitoring method according to claim 4, wherein: The determining of the first R wave peak feature point of each electrocardiogram signal segment based on the target energy extreme point includes: Calculating a curvature response value based on the local energy gradient curvature of the target energy extreme point; constructing an adjacent extreme point association network based on the time interval and curvature response value difference between each first energy extreme point in the target energy extreme point and a second preset number of second energy extreme points adjacent to it; Determining an importance index based on an edge weight of each first energy extreme value point in the adjacent extreme value point association network and a curvature response value of its corresponding second energy extreme value point in the adjacent extreme value point association network; The first R-wave peak feature point of each electrocardiogram signal segment is determined based on the importance index of each first energy extreme value point.

6. The blood pressure monitoring method according to claim 4, wherein: The calculation formula of the curvature response value of the target energy extreme point is as follows: in, Indicates the The curvature response value of the target energy extreme point, Indicates the The energy intensity of the target energy extreme point, represents the average energy intensity of the energy intensity sequence, Represents the energy intensity standard deviation of the energy intensity series.

7. The blood pressure monitoring method according to claim 4, wherein: The edge weight calculation formula of each first energy extreme point in the associated network of adjacent extreme points is as follows: in, Indicates the The edge weight of the first energy extreme point in the network associated with the adjacent extreme points, Indicates the The first energy extreme point and its adjacent The time interval between the second energy extreme points, Indicates the The first energy extreme point and its adjacent The difference in curvature response values ​​between the second energy extreme points, Indicates the difference threshold of curvature response values.

8. The blood pressure monitoring method according to claim 4, wherein: The determining of the first R wave peak feature point of each electrocardiogram signal segment based on the importance index of each first energy extreme value point includes: For each ECG signal segment, the first energy extreme value point whose importance index is greater than or equal to a preset index threshold is determined as a preliminary R-wave peak feature point; For each R-peak feature point in the preliminary R-peak feature points, calculating the second-order derivative of the local energy gradient curvature corresponding to the R-peak feature point in the energy-time fluctuation matrix; Traverse each second-order derivative and determine the target R peak feature point where the second-order derivative changes from positive to negative; If the target R peak feature point belongs to the preliminary R peak feature point and the number of the target R peak feature point is 1, the target R peak feature point is determined as the first R peak feature point; If the target R peak feature point belongs to the preliminary R peak feature point, and the number of the target R peak feature points is greater than 1, the target R peak feature point corresponding to the importance index with the largest value is determined as the first R peak feature point.

9. An electronic device comprising: The memory and the processor are characterized in that a computer software program is stored in the memory, and when the processor reads and executes the computer software program, the blood pressure monitoring method according to any one of claims 2 to 8 is implemented.

10. A non-transitory computer-readable storage medium, characterized in that The storage medium stores a computer software program, which, when executed by a processor, implements the blood pressure monitoring method according to any one of claims 2 to 8.

Images