ECG Signal Processing Method, Portable ECG Acquisition Device and Storage Medium

Abstract

Embodiments of the present disclosure provide a portable electrocardiogram (ECG) acquisition device, an ECG signal processing method therefor, and a computer-readable information storage medium. The ECG signal processing method includes: acquiring continuously acquired ECG signals; sampling the ECG signals in a target time period to obtain ECG sampling data; dividing the ECG sampling data into multiple sub-period ECG sampling data sets; determining the average value of data changes within each sub-period; obtaining a first data set of inter-sub-period data change values by calculating the inter-sub-period data change values between adjacent sub-periods; calculating the root mean square of differences (RMSSD) for the items in the first data set; determining whether an event conforming to specific arrhythmia characteristics has occurred based on the first data set and the RMSSD; if it is determined that an event conforming to specific arrhythmia characteristics has occurred, recording information about the event and sending the recorded event information.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to electrocardiogram (ECG) data processing, and particularly to a portable ECG acquisition device, an ECG signal processing method therefor, and a computer-readable information storage medium. Background Art

[0002] The myocardial cell membrane of the human body is a semipermeable membrane. When in a resting state, a certain number of positively charged cations are arranged outside the membrane, and the same number of negatively charged anions are arranged inside the membrane. The potential outside the membrane is higher than that inside the membrane, which is called the polarized state. In the resting state, since myocardial cells in all parts of the heart are in the polarized state and there is no potential difference, the potential curve recorded by the current recorder is flat, which is the isoelectric line of the surface electrocardiogram. When myocardial cells are stimulated by a certain intensity, the permeability of the cell membrane changes, and a large number of cations rush into the membrane within a short time, making the potential inside the membrane change from negative to positive. This process is called depolarization.

[0003] For the whole heart, the potential change during the sequential depolarization process of myocardial cells from the endocardium to the epicardium is recorded by the current recorder as the depolarization wave, that is, the P wave of the atrium and the QRS wave of the ventricle on the surface electrocardiogram. After the cell depolarization is completed, the cell membrane discharges a large number of cations again, making the potential inside the membrane change from positive to negative and return to the original polarized state. This process proceeds from the epicardium to the endocardium and is called repolarization. Similarly, the potential change during the repolarization process of myocardial cells is recorded by the current recorder as the repolarization wave. Since the repolarization process is relatively slow, the repolarization wave is lower than the depolarization wave. The repolarization wave of the atrium is low and buried in the depolarization wave of the ventricle, making it difficult to identify on the surface electrocardiogram. The repolarization wave of the ventricle is manifested as the T wave on the surface electrocardiogram. After all myocardial cells are completely repolarized and return to the polarized state again, there is no potential difference between myocardial cells in each part, and the isoelectric line is recorded on the surface electrocardiogram.

[0004] Figure 1 is a schematic diagram showing the changes in cardiac electrical activity in different time periods of the ECG waveform. Referring to Figure 1 , the ECG waveform of the human body usually includes the following several wave bands.

[0005] 1. P wave: Represents atrial electrical activity. The normal time limit is 0.08 - 0.11 seconds, and it is a small upward arch on the electrocardiogram. Usually, when the heart starts to contract, the sinoatrial node first emits an electrical signal that conducts to the atrium, causing depolarization of the atrium. At this time, the atrium will be excited, and then atrial contraction occurs.

[0006] When the atrium is enlarged and there is an abnormality in the conduction between the two atria, the P wave can be manifested as a tall and pointed or bimodal P wave.

[0007] 2. PR Segment: The impulse is conducted along the anterior, middle, and posterior internodal tracts to the atrioventricular node. Due to the slow conduction velocity of the atrioventricular node, the PR segment on the electrocardiogram is formed, also known as the PR interval. The normal PR interval is between 0.12 and 0.20 seconds.

[0008] When there is a block in the conduction from the atrium to the ventricle, it is manifested as an extension of the PR interval or the disappearance of the ventricular wave after the P wave.

[0009] 3. QRS Complex: It represents the electrical activity generated during ventricular depolarization and excitation. The normal duration is 0.06 - 0.10 seconds, and the amplitude is relatively large on the electrocardiogram, significantly larger than the P wave. After atrial excitation, the electrical signal continues to conduct downward, passes through the atrioventricular node and conducts to the ventricle, causing depolarization and excitation of the ventricular muscle. When the ventricle is excited, ventricular contraction occurs, and at the same time, electrical activity occurs. At this time, when the electrical activity conducts to the body surface, it will be recorded by the electrocardiogram recorder. The waveform of the QRS complex on the electrocardiogram is relatively large, may have more than 1 main peak, and may have 2 peaks, up and down. This waveform is called the QRS complex.

[0010] 4. ST Segment: A period of time when all ventricular muscles have completed depolarization and repolarization has not yet started. Under normal circumstances, the ST segment should be on the isoelectric line.

[0011] When there is ischemia or necrosis in a certain part of the myocardium, there is still a potential difference in the ventricle after depolarization is completed. At this time, it is manifested as a deviation of the ST segment on the electrocardiogram.

[0012] 5. T Wave: It represents the process of cardiac repolarization, that is, the process from ventricular depolarization and excitation to returning to the quiet state, usually the process of ventricular diastole. The normal duration is 0.05 - 0.25 seconds. On the electrocardiogram, the T wave is slightly higher than the P wave, but lower than the peak of the QRS complex.

[0013] 6. U Wave: The U wave can be seen after the T wave in some leads. Currently, it is considered to be closely related to ventricular repolarization.

[0014] In addition, because the QT interval is affected by the heart rate, the concept of corrected QT interval (QTC) is introduced. The QT interval is the time period from ventricular depolarization (starting from the QRS segment) to repolarization (ending with the T wave). The normal length of the QT interval is 0.44 seconds. The prolongation of the QT interval is often related to the occurrence of malignant arrhythmias.

[0015] When the human heart has congenital structural defects or abnormalities occur during life experiences, various arrhythmias will occur, such as Figure 2 as shown. Each arrhythmia has specific electrocardiogram waveform characteristics.

[0016] Among them, atrial fibrillation (abbreviated as AF) is the most common arrhythmia. With the increase of age, the incidence of AF continues to increase, reaching 10% in people over 75 years old. When in AF, the atrial activation frequency reaches 300 - 600 beats per minute, and the heart rate is often fast and irregular, sometimes reaching 100 - 160 beats per minute. It is not only much faster than the normal heart rate but also completely irregular, and the atrium loses its effective contraction function. The prevalence of AF is also closely related to diseases such as coronary heart disease, hypertension, and heart failure.

[0017] According to different sources, up to 50% of stroke incidents are caused by persistent or paroxysmal AF. Anticoagulant therapies such as warfarin can reduce the risk of stroke and death. However, continuous anticoagulation increases the risk of bleeding. Novel anticoagulants (NOACs) (such as Pradaxa, Xarelto, or Eliquis) have a faster effect compared to, for example, warfarin. NOACs reduce the bleeding risk by replacing continuous anticoagulants with anticoagulants when needed (also known as the "self - medication" method). The criteria for taking this drug are determined by the threshold level of the paroxysmal AF (PAF) burden. The AF burden is defined as the total duration of all AF events during the monitoring period (usually 24 hours or longer).

[0018] The self - administered anticoagulation treatment method faces many challenges. AF events, especially short - term PAF events, are difficult to detect. Patients with AF symptoms may be asymptomatic in most cases. It is necessary to monitor AF patients continuously for 24 / 7 for many years to automatically detect AF events. Most current AF detection devices are designed for diagnosing AF rather than for evaluating the PAF burden. They can be well used to detect at least one AF event during long - term monitoring, but cannot detect short - term PAF events with the required accuracy to evaluate the PAF burden, or they are too prominent for long - term continuous monitoring.

[0019] To solve this problem, Medtronic has developed and marketed an implantable cardiac monitor (ICM) with AF detection function. The ICM device is a miniature electrocardiogram (ECG) monitor inserted under the patient's skin. The latest version of such a monitor, Reveal LINQ TM can provide continuous monitoring for up to 3 years. LINQ has shown good results in detecting PAF events lasting 2 minutes or longer, but it cannot detect short - term PAF. Since short - term PAF events have been observed in many AF patients and can significantly increase the overall PAF burden, the application of ICM devices in anticoagulation treatment based on the "self - medication" AF burden is limited. Summary of the Invention

[0020] An embodiment of the present invention provides a portable electrocardiogram (ECG) acquisition device and an ECG signal processing solution therefor, which can accurately monitor specific arrhythmia events through simple operations using limited computing resources.

[0021] According to one aspect of an embodiment of the present invention, there is provided an ECG signal processing method for a small ECG acquisition device, including: acquiring ECG signals continuously collected from a human body; sampling the ECG signals in a target time period to obtain ECG sampling data; dividing the ECG sampling data in the target time period into sub-period ECG sampling data sets respectively corresponding to multiple sub-periods; determining the average value of data changes within each sub-period of the ECG sampling data corresponding to each sub-period; obtaining a first data set of inter-sub-period data change values by calculating the inter-sub-period data change values between the average values of data changes within each adjacent sub-period; calculating the root mean square of differences (RMSSD) for the items in the first data set; determining whether an event conforming to specific arrhythmia characteristics has occurred in the target time period according to the first data set and the RMSSD; if it is determined that the event conforming to the specific arrhythmia characteristics has occurred, recording the information of the event and / or sending the information of the event conforming to the specific arrhythmia characteristics.

[0022] Optionally, the method further includes: determining the proportion of the number of items in the first data set where the inter-sub-period data change value is greater than the RMSSD in the first data set; if the proportion is between a predetermined upper proportion limit value and a lower proportion limit value, stopping the processing of the ECG signals in the target time period.

[0023] Optionally, the determining whether an event conforming to specific arrhythmia characteristics has occurred in the target time period according to the first data set and the RMSSD includes: obtaining a second data set including the inter-sub-period data change values and the corresponding sub-periods by selecting the inter-sub-period data change values in the first data set that meet the following conditions: the inter-sub-period data change value is not less than the RMSSD and the inter-sub-period data change value corresponding to the adjacent sub-period after it is less than the RMSSD; obtaining a third data set of time intervals that meet the following conditions according to the time intervals between adjacent sub-periods in the second data set: the time intervals between adjacent sub-periods exceed a preset duration threshold; determining whether an event conforming to atrial fibrillation characteristics has occurred in the target time period according to the values of the items in the third data set.

[0024] Optionally, determining whether an event conforming to the characteristics of atrial fibrillation has occurred within the target time period according to the values of the items in the third data set includes: obtaining the total number m of the items in the third data set and the number p of the item values in the third data set; calculating the complexity index C of the electrocardiogram signal in the target time period according to the total number m and the number p of the item values; if the calculated complexity index C > a preset PAF threshold, it can be determined that an event conforming to the characteristics of atrial fibrillation has occurred within the target time period.

[0025] Optionally, the method further includes: storing the data of the electrocardiogram signals continuously collected from the human body, and / or when it is detected that the small electrocardiogram acquisition device has established a network connection, sending the stored data of the electrocardiogram signals and the information of the events conforming to the characteristics of specific cardiac arrhythmias that have been recorded.

[0026] According to another aspect of the embodiments of the present invention, there is provided a portable electrocardiogram acquisition device, including: an ECG acquisition electrode for continuously acquiring the electrocardiogram signal of the human body, the electrocardiogram signal being an electrocardiogram voltage signal; and a microcontroller unit including an ECG processing firmware, wherein the ECG processing firmware includes:

[0027] A preprocessing module for sampling the electrocardiogram signal in the target time period acquired by the ECG acquisition electrode to obtain electrocardiogram sampling data, and dividing the electrocardiogram sampling data in the target time period into sub-period electrocardiogram sampling data sets corresponding to multiple sub-periods respectively;

[0028] An averaging module for determining the average value of the data change within the sub-period corresponding to each sub-period of the electrocardiogram sampling data processed by the preprocessing module, obtaining a first data set of the data change values between sub-periods by calculating the data change values between the average values of the data changes within each adjacent sub-period, and calculating the root mean square value of the differences RMSSD for the items in the first data set;

[0029] An abnormal event detection module for determining whether an event conforming to the characteristics of a specific cardiac arrhythmia has occurred within the target time period according to the first data set and the root mean square value of the differences RMSSD;

[0030] An abnormal event processing module includes: a memory unit for recording the information of the event if it is determined that the event conforming to the characteristics of the specific cardiac arrhythmia has occurred; and / or a communication unit for sending the information of the events conforming to the characteristics of the specific cardiac arrhythmia that have been recorded.

[0031] Optionally, the ECG processing firmware further includes: a noise filtering unit configured to: determine the proportion of the number of items in the first data set whose inter-sub-period data change value is greater than the RMSSD in the first data set; if the proportion is between a predetermined upper proportion limit value and a lower proportion limit value, stop processing the electrocardiogram signal of the target time period; if the proportion is not between the predetermined upper proportion limit value and the lower proportion limit value, notify the abnormal event detection module to perform detection processing on the first data set.

[0032] Optionally, the abnormal event detection module includes:

[0033] A first processing unit configured to obtain a second data set including inter-sub-period data change values and corresponding sub-time periods by selecting inter-sub-period data change values in the first data set processed by the averaging module that meet the following conditions: the inter-sub-period data change value is not less than the root mean square of the differences RMSSD and the inter-sub-period data change value corresponding to the adjacent sub-time period thereafter is less than the root mean square of the differences RMSSD;

[0034] A second processing unit configured to obtain a third data set of time intervals that meet the following conditions according to the time intervals between adjacent sub-time periods in the second data set obtained by the first processing unit: the time intervals between adjacent sub-time periods exceed a preset duration threshold;

[0035] A detection unit configured to determine whether an event conforming to the characteristics of atrial fibrillation has occurred in the target time period according to the values of the items in the third data set.

[0036] Optionally, the detection unit is configured to obtain the total number m of items in the third data set and the number p of item values in the third data set, calculate the complexity index C of the electrocardiogram signal of the target time period according to the total number m and the number p of item values, and if the calculated complexity index C > a preset PAF threshold, it can be determined that an event conforming to the characteristics of atrial fibrillation has occurred in the target time period.

[0037] Optionally, the memory unit is further configured to store electrocardiogram signals continuously collected from the human body by the ECG acquisition electrode, and / or, the communication unit is further configured to send the data of the electrocardiogram signals stored in the memory and the information of the events conforming to the characteristics of specific cardiac rhythm abnormalities recorded when it is detected that the small electrocardiogram acquisition device has established a network connection.

[0038] According to another aspect of the embodiments of the present invention, there is provided a computer-readable storage medium having computer program instructions stored thereon, wherein when the program instructions are executed by a processor, the steps of any of the foregoing electrocardiogram signal processing methods for a small electrocardiogram acquisition device are implemented.

[0039] A portable electrocardiogram (ECG) acquisition device, an ECG signal processing method for a small-sized ECG acquisition device, and a computer-readable storage medium according to an embodiment of the present invention divide the continuously acquired ECG signals from a human body into detection units in time windows (target time periods), sample the divided ECG signals, and then divide the acquired ECG sampling data into sub-period ECG sampling data sets corresponding to multiple sub-time periods, calculate the average value of data changes within each sub-time period, and calculate the data change value between adjacent sub-periods to obtain a first data set reflecting the changes in ECG data between multiple sub-periods, and calculate the root mean square of the differences RMSSD of the items in the first data set. Then, for a specific arrhythmia event to be detected, the first data set and the RMSSD are analyzed and processed, so as to determine whether an event conforming to the characteristics of the specific arrhythmia has occurred within the target time period, and the information of the recorded events conforming to the characteristics of the specific arrhythmia can be recorded and sent. The above processing uses the ECG waveform characteristics of specific arrhythmias and can detect specific arrhythmia events with the least CPU usage and sufficient accuracy through simple arithmetic operations suitable for small devices with limited computing power, so that the device can provide long-term monitoring for up to several years without replacing the battery; and by continuously executing the ECG signal processing method on the continuously acquired ECG signals, continuous detection can be performed for a long time to obtain the abnormal rule information of the ECG signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic diagram showing the changes in cardiac electrical activity in different time periods of an ECG waveform;

[0041] Figure 2 shows the classification of various arrhythmias;

[0042] Figure 3 is a schematic diagram showing the ECG waveform of premature ventricular contractions;

[0043] Figure 4 is a schematic diagram showing the ECG waveforms of normal sinus rhythm with regular heartbeats and atrial fibrillation events with irregular heartbeats;

[0044] Figure 5 shows a flowchart of an ECG signal processing method for a small-sized ECG acquisition device according to an embodiment of the present invention;

[0045] Figure 6 shows a flowchart of noise detection and filtering processing for an ECG signal according to an embodiment of the present invention;

[0046] Figure 7 shows Figure 5Flow chart of an exemplary process of step S270 in

[0047] Figure 8 Shows a part 410 of the originally acquired electrocardiogram signal within a 33 - second sliding window and a part 420 of the derived data obtained through the process according to an embodiment of the present invention.

[0048] Figure 9 Shows a histogram of the complexity index C of the electrocardiogram waveform signal within a 33 - second sliding window according to the process of an embodiment of the present invention.

[0049] Figure 10 Shows a structural block diagram of a portable electrocardiogram acquisition device according to an embodiment of the present invention. Detailed implementation manners

[0050] The following further elaborates on the detailed implementation manners of the embodiments of the present disclosure in conjunction with the accompanying drawings (the same reference numerals in several accompanying drawings represent the same elements) and embodiments. The following embodiments are used to illustrate the present invention but are not used to limit the scope of the present invention.

[0051] Those skilled in the art can understand that terms such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor indicate an inevitable logical order between them.

[0052] As mentioned above, normal heart rhythm and each type of arrhythmia have specific electrocardiogram waveform characteristics.

[0053] For example, premature beats of the heart are called premature contractions. The normal pacemaker of the heart is the sinoatrial node, which has a regular rhythm of firing 60 - 100 times per minute. However, if an ectopic pacemaker, such as in the atrium, ventricle or other parts, prematurely generates a cardiac impulse, it is called a premature contraction or premature beat. Therefore, the electrocardiogram manifestations of premature contractions are different according to the location of occurrence. Premature contractions can occur in the atrium, ventricle, or at the atrioventricular junction of the atrium. Figure 3 Shows a schematic diagram of the ECG waveform of a cardiac premature contraction.

[0054] In atrial premature contractions, the electrocardiogram mainly shows that the P wave representing atrial contractions appears prematurely, and the QS waveform of the subsequent sinus rhythm is the same as that of the sinus rhythm, which is the manifestation of atrial premature contractions. Premature contractions in the atrioventricular junction area are characterized by a premature ventricular contraction, and a retrograde P wave can be seen before, during or after it. The characteristics of ventricular premature contractions are more unique. On the electrocardiogram, a wide and abnormal QRS complex appears prematurely, that is, the widest wave group on the electrocardiogram, and there is no atrial premature beat before it.

[0055] Taking atrial fibrillation as an example again, atrial fibrillation (AF) is defined as an arrhythmia lasting for 30 seconds or longer. Its main electrocardiogram manifestations include the following aspects: First, the P wave disappears, replaced by small and irregular baseline fluctuations, with varying morphology and amplitude, called small F waves, with a frequency of about 350 - 600 beats per minute. Second, the ventricular rate is extremely irregular. In patients with atrial fibrillation who have not received drug treatment and have normal atrioventricular conduction, the ventricular rate is usually between 100 - 160 beats per minute. Drugs, exercise, fever, hyperthyroidism, etc. can all shorten the refractory period of the atrioventricular node, thereby accelerating the ventricular rate. On the contrary, digitalis prolongs the refractory period of the atrioventricular node and slows down the ventricular rate. Third, the QRS complex morphology is usually normal. When the ventricular rate is too fast and there is intraventricular aberrant conduction, the QRS complex can widen and deform. These are the main manifestations of the electrocardiogram of atrial fibrillation. Figure 4 Shows the ECG waveforms of a normal sinus rhythm with regular heartbeats and the ECG waveforms of an atrial fibrillation event with irregular heartbeats. The ECG waveforms plot the electrical activity of the heart, usually expressed in microvolts or millivolts.

[0056] The embodiment of the present invention provides an electrocardiogram signal processing method applicable to small electrocardiogram acquisition devices (such as portable electrocardiogram acquisition devices), which uses relatively simple arithmetic operations and utilizes the waveform characteristics presented by the electrocardiogram waveforms to detect events that meet specific cardiac rhythm abnormalities.

[0057] This electrocardiogram signal processing method is based on the measurement of heart rate irregularity, which is carried out by quantitatively evaluating the complexity of the time series of selected data points in a sliding window of ECG data.

[0058] Figure 6 Shows a flowchart of an electrocardiogram signal processing method for a small electrocardiogram acquisition device according to an embodiment of the present invention. Of course, this electrocardiogram signal processing method can also be executed by software or firmware in a general-purpose computer or a dedicated computing device.

[0059] According to an exemplary embodiment of the present invention, the electrocardiogram signals continuously collected by the ECG acquisition electrodes are divided and processed according to a sliding window corresponding to the time length, and this sliding window is defined as the target time period. The electrocardiogram signal processing method according to the embodiment of the present invention processes the electrocardiogram signals corresponding to the target time period. In the following text of this application, no literal distinction is made between the target time period and the sliding window.

[0060] Refer to Figure 6 , in step S210, obtain the electrocardiogram signals continuously collected from the human body.

[0061] For example, a small electrocardiogram (ECG) acquisition device can continuously acquire the ECG signals of the human body through ECG acquisition electrodes. Alternatively, a dedicated computing device can receive the continuously acquired ECG signals from the ECG acquisition device. The continuously acquired ECG signals are ECG voltage signals acquired through the acquisition electrodes. It can be understood that, according to the processing requirements, ECG current signals or impedance signals can also be used.

[0062] In step S220, the ECG signals in the target time period are sampled to obtain ECG sampling data.

[0063] The duration of the target time period of the ECG signals can be determined according to the waveform characteristics, waveform distribution, and duration of the arrhythmia to be detected. Taking atrial fibrillation as an example, since atrial fibrillation is usually characterized by arrhythmia lasting for 30 seconds or longer, the target time period can be determined to be a duration exceeding 30 seconds, such as 33 seconds, 36 seconds, etc. Further, the value of the target time period (sliding window) can be appropriately adjusted according to the sampling frequency.

[0064] For example, the target time period (sliding window) can be determined to be around 33 seconds according to prior knowledge. Assuming that the ECG signals are sampled at a sampling frequency of 250 samples per second, then through the processing in step S210, 8250 ECG sampling data can be obtained. For the convenience of calculation and processing, the target time period can be adjusted to the duration corresponding to 8192 (a multiple of 2) ECG sampling data, that is, 8192 / 250 = 32.768 seconds, and each ECG sampling data contains ECG signal data with a duration of 4 ms = 1,000 / 250.

[0065] It should be noted that the sampling frequency, the size of the sliding window, and the average number of samples may vary according to the parameters of the ECG signals, device specifications, and applications.

[0066] In step S230, the ECG sampling data within the target time period is divided into sub-period ECG sampling data sets corresponding to multiple sub-periods (corresponding to SubWindow), and the sub-time periods corresponding to the multiple sub-period ECG sampling data sets reflect the time sequence in the target time period.

[0067] That is to say, based on step S210, the sliding window (corresponding to the target time period) is divided into equal-length sub-windows (corresponding to sub-time periods), and the ECG sampling data corresponding to the sliding window W is divided into sub-period ECG sampling data sets corresponding to multiple equal-length sub-time periods, and each sub-period ECG sampling data set has the same number of ECG sampling data.

[0068] For example, the aforementioned 32.768-second sliding window (target time period) can be divided into 16 sub-windows (sub-time periods), so that 8,192 electrocardiogram sampling data are divided into 16 sub-time period electrocardiogram sampling data sets, and each sub-time period electrocardiogram sampling data set contains 512 electrocardiogram sampling data.

[0069] In step S240, for multiple sub-time period electrocardiogram sampling data sets, determine the average value of data changes within the sub-time period of the electrocardiogram sampling data corresponding to each sub-time period.

[0070] Specifically, assume that the sliding window W is divided into equal-length sub-windows w i , and each sub-window w i has Save electrocardiogram sampling data. For each sub-time period electrocardiogram sampling data set of the sub-window (sub-time period) w i , calculate the average value of data changes within the sub-time period Vi between the values of consecutive electrocardiogram sampling data through Equation 1.

[0071]

[0072] where i represents the i-th sub-time period, j is the electrocardiogram sampling data index of the i-th sub-time period, Save is the number of electrocardiogram sampling data in each sub-time period, j = {1, 2, 3,..., Save}, v i,j , is the value of the j-th electrocardiogram sampling data in the electrocardiogram sampling data of the i-th sub-time period, v i,j-1 , is the value of the (j - 1)-th electrocardiogram sampling data in the electrocardiogram sampling data of the i-th sub-time period, and Vi is the average value of data changes within the sub-time period among the Save consecutive electrocardiogram sampling data in the sub-time period electrocardiogram sampling data set corresponding to the i-th sub-time period.

[0073] Through the aforementioned processing, obtain the average value of data changes within N sub-time periods V = {V1, V2, V3,..., V n}.

[0074] In step S250, by calculating the data change values between adjacent sub-time periods of the average value of data changes within the sub-time periods, obtain a first data set of data change values between sub-time periods. The data change values between sub-time periods in the first data set also correspond to the sub-time periods in the target time period respectively, and also reflect the time sequence in the target time period.

[0075] Specifically, convert the set V = {V1, V2, V3,..., V n} of the average value of data changes within the sub-time periods into a first data set D = {d1, d2, d3,..., d i ,..., d N-1} of the data change values between sub-time periods, where d i = |Vi+1 –V i |. In addition, each d i is associated with a timestamp Td i , where Td i is the average value V of the data change within a sub-period i corresponding to the time of sub-period i, such as the start time, end time or mid-time point of sub-period i.

[0076] In step S260, for the items of the first data set D = {d1, d2, d3, … d i , … d N-1}, the root mean square of the differences RMSSD is calculated.

[0077] Specifically, the root mean square of the successive differences RMSSD is calculated by equation (2).

[0078]

[0079] where d i = |V i+1 –V i |.

[0080] The root mean square of the differences RMSSD is the root mean square of the differences between adjacent normal cardiac cycles, and the normal value range is (27 ± 12) ms, which is an index of heart rate variability.

[0081] Thereafter, in step S270, based on the first data set obtained in step S250 and the root mean square of the differences RMSSD calculated in step S260, it is determined whether an event conforming to the characteristics of a specific cardiac arrhythmia has occurred within the target time period.

[0082] The data change values between the respective sub-periods in the first data set reflect the changes in electrocardiogram data between multiple sub-periods, while the root mean square of the differences RMSSD characterizes the sample standard deviation of the data change values between the respective sub-periods in the first data set. For a specific cardiac arrhythmia to be detected, this two data (the first data set and the root mean square of the differences RMSSD) obtained by arithmetic operations can be used for detection processing, and accordingly, an event conforming to the characteristics of a specific cardiac arrhythmia is detected.

[0083] If in step S270, it is determined that no event conforming to the characteristics of a specific cardiac arrhythmia has occurred, then steps S220 to S260 can be returned to for execution, and the sampling and detection processing of the electrocardiogram signal for the next target time period is continued, and thus the detection of cardiac arrhythmia events is continuously performed.

[0084] If, in step S270, it is determined that an event conforming to specific arrhythmia characteristics has occurred, then step S280 is executed. In step S280, information on the event conforming to specific arrhythmia characteristics is recorded and / or the information on the event conforming to specific arrhythmia characteristics is sent, for example, but not limited to, the type of the event (such as premature beats, atrial fibrillation, etc.) and the time information of the occurrence of the event, etc.

[0085] The information on the recorded abnormal event can be provided to the user or uploaded to the control device or server according to the timing or operation mode suitable for the small electrocardiogram acquisition device. For example, in order to save the power consumption of the small electrocardiogram acquisition device, the small electrocardiogram acquisition device is usually set / designed to establish a network connection only when necessary or at a predetermined time period. For this reason, according to an optional embodiment of the present invention, when the small electrocardiogram acquisition device establishes a network connection, the information on the recorded event conforming to specific arrhythmia characteristics is sent.

[0086] In addition, according to an exemplary embodiment of the present invention, the data of the electrocardiogram signals continuously collected from the human body can also be stored, and / or when the small electrocardiogram acquisition device establishes a network connection, the stored electrocardiogram signal data and the information on the recorded event conforming to specific arrhythmia characteristics are sent.

[0087] According to the electrocardiogram signal processing method for a small electrocardiogram acquisition device according to an embodiment of the present invention, the electrocardiogram signals continuously collected from the human body are divided into detection units with a time window (target time period), the divided electrocardiogram signals are sampled, and then the sampled electrocardiogram sampling data is divided into sub-period electrocardiogram sampling data sets corresponding to multiple sub-time periods. The average value of the data change within each sub-time period is calculated and the data change value between adjacent sub-periods is calculated to obtain a first data set reflecting the electrocardiogram data change between multiple sub-periods, and the root mean square of the differences of the items in the first data set, RMSSD, is calculated. Then, for the specific arrhythmia event to be detected, the first data set and the root mean square of the differences RMSSD are analyzed and processed, so as to determine whether an event conforming to the specific arrhythmia characteristics has occurred within the target time period, and the information on the recorded event conforming to the specific arrhythmia characteristics can be recorded and sent. The above processing can detect specific arrhythmia events with the least CPU usage rate and sufficient accuracy by using the electrocardiogram waveform characteristics of specific arrhythmia and performing simple arithmetic operations suitable for small devices with limited computing power, so that the device can provide long-term monitoring for several years without replacing the battery; and by continuously executing the electrocardiogram signal processing method on the continuously collected electrocardiogram signals, the detection can be continuously performed for a long time to obtain the abnormal rule information of the electrocardiogram signals.

[0088] It should be noted that the information of the events obtained by detection that conform to the characteristics of specific cardiac arrhythmias cannot be used as the final diagnosis result, but only for medical reference. This electrocardiogram signal processing method is used to perform preliminary detection of electrocardiogram signals collected from the human body that conform to specific signal characteristics, so that users (or medical staff) can monitor and prompt their own (or the monitored object's) status in real time, rather than for disease diagnosis. Medical staff can conduct a comprehensive examination and evaluation of the monitored object based on the aforementioned prompts received, and then perform necessary treatments. This electrocardiogram signal processing method can achieve long-term continuous detection, prompt users or medical staff in a timely manner, and save the monitoring time of users and improve the monitoring efficiency.

[0089] In daily life, due to the monitored object may be in a large-scale movement state or due to signal interference, the collected electrocardiogram signals may carry relatively large noise, and this noise will affect the accuracy of electrocardiogram signal processing. Therefore, it is necessary to perform noise detection and skip the electrocardiogram signals containing noise.

[0090] Correspondingly, according to an optional embodiment of the present invention, after performing step S260 and before performing step S270, perform the noise detection and filtering processing of the following operations S262 and S265.

[0091] Refer to Figure 6 , in operation S262, determine that among the first data set D = {d1, d2, d3,... d i ,... d N-1}, the proportion P(RMSSD) of the number of items with a data change value greater than RMSSD between sub-periods in the first data set.

[0092] Calculate the probability P(x) through formula (3):

[0093] P(x) = Pr(X > x) (3)

[0094] Where, Pr(X>x) represents the probability that the variable X takes a value greater than x. In the method of the embodiment of the present invention, the random variable X in formula (7) comes from the first data set D = {d1, d2, d3,... d i ,... d N-1}, and x is the root mean square RMSSD. In this case, equation (3) is converted to

[0095] P(RMSSD) = Pr(X>RMSSD) (4)

[0096] In operation S265, determine whether this proportion is between a predetermined upper proportion limit value and a lower proportion limit value.

[0097] Specifically, determine whether the following formula holds: F min≤P(RMSSD)<F max where, F max is the upper limit value of the proportion, and F min is the lower limit value of the proportion. The values of F max and F min can be determined according to experience. For example, F min = 25% - 30%, and F max = 60% - 65%. It is obvious to those skilled in the art that depending on the ECG signal characteristics, the analysis window, the application, and other factors, F min and F max can have different values.

[0098] If in operation S265, it is determined that the proportion value is between the predetermined upper limit value F max and the lower limit value F min of the proportion, it can be determined that the ECG signal in the target time period has relatively large noise, and the processing of the ECG signal in the target time period is stopped. That is to say, if P(RMSSD) is within the aforementioned F max and F min range, it can be determined that the first data set D stops further analysis of the ECG signal in the target time period due to excessive noise, and instead returns to execute step S220 to process and analyze the ECG data of the next target time period.

[0099] On the other hand, if in operation S265, it is determined that the proportion value is not between the predetermined upper limit value F max and the lower limit value F min of the proportion, it can be determined that the ECG signal in the target time period meets the processing requirements, and the processing of step S270 is continued.

[0100] The following will refer to Figure 7 to describe in detail the processing of detecting events conforming to the characteristics of atrial fibrillation in step S270. Among them, the complexity of the sequence of the inter-sub-period data change values in the first data set is analyzed, the complexity index indicating the existence of atrial fibrillation events is calculated, and whether an event conforming to the characteristics of atrial fibrillation has occurred is determined according to the calculated complexity index.

[0101] Referring to Figure 7 in step S271, by selecting the inter-sub-period data change values that meet the following conditions in the first data set, a second data set including the inter-sub-period data change values and the corresponding sub-time periods is obtained: the inter-sub-period data change value is not less than RMSSD and the inter-sub-period data change value corresponding to the adjacent sub-time period behind it is less than the RMSSD.

[0102] Specifically, first, from the first data set D = {d1, d2, d3,... di ,…d N-1 From the items d in}, select the item d that satisfies the following condition (5) i :

[0103] d i+1 < RMSSD ≤ d i (5)

[0104] That is, select the data change value d between sub - time periods whose data change value between its own sub - time periods is not less than the RMSSD value, but the data change value between the adjacent subsequent sub - time periods is less than the RMSSD value i .

[0105] After that, according to the item d selected by condition (5) i The time stamp Td of the corresponding sub - time period i constitute the second data set Td = {Td1, Td2, … Td i ,… Td k}, where Td1, Td2, … Td i ,… Td k correspond to the selected data change values d1, d2, … d between sub - time periods respectively. The second data set contains the items of the sub - time periods when the electrocardiogram signal changes from higher than (and equal to) RMSSD to lower than RMSSD k .

[0106] By filtering out the second data set that contains the data change value between sub - time periods and the corresponding sub - time periods, where the data change values between adjacent sub - time periods meet the above - mentioned filtering conditions, the data of the sub - time periods corresponding to overly large data change values between sub - time periods in a short time (such as QRS waves) and the data of the sub - time periods with overly small data change values between sub - time periods (without waveform change rules) can be filtered out. Therefore, the items (the time stamps Td of the sub - time periods) in the second data set Td do not all correspond to adjacent sub - time periods, but only to the sub - time periods of the waveform part of the baseline fluctuation i ).

[0107] In step S272, according to the time intervals between adjacent sub - time periods in the second data set, obtain the third data set of time intervals that meet the following conditions: the time intervals between adjacent sub - time periods exceed a preset duration threshold

[0108] Specifically, first obtain the time interval TT between each adjacent sub - time period through equation (6) i :

[0109] TT i = |Td i+1 – Td i | (6)

[0110] Thus, a set of time intervals between adjacent sub-time periods is obtained.

[0111] Then, a third data set of time intervals that meet condition (7) is selected from the set of the aforementioned time intervals:

[0112] |Td i+1 – Td i | > B (7)

[0113] where B is a determined threshold for the blank time period after a heartbeat. This threshold is set to 256 milliseconds. However, it is obvious to those skilled in the art that the threshold B may have different values based on the application.

[0114] Thus, a third data set TT = {TT1, TT2, TT3, …, TT i , …, TT m} is obtained. The third data set TT contains items where the time intervals between sub-time periods when the electrocardiogram signal changes from being higher than RMSSD to being lower than RMSSD are greater than the threshold for the blank time period of a heartbeat. The time points corresponding to these items (such as the Td i of TT i or the time stamp of Tdi+1) constitute the timing feature points of the time intervals that meet condition (7).

[0115] Figure 8 shows a part 410 of the originally acquired electrocardiogram signal within a 33 - second sliding window and a part 420 of the derived data obtained through the processing according to an embodiment of the present invention. Among them, the straight line 430 shows the RMSSD value calculated by equation (2). The small black dots 440 point to the d i value of the set D of consecutive differences from the average signal value V i . The large black dots 450 point to the differences {d1, d2, … d k} related to the time points 470 {Td1, Td2, … Td k} selected according to condition (7). The large black dots 460 point to the same points as the large black dots 450, but are set on the original ECG signal 410. All the differences are greater than B, so n = k - 1.

[0116] In step S273, according to the values of the items in the third data set, it is determined whether an event conforming to the characteristics of atrial fibrillation has occurred within the target time period.

[0117] First, the total number m of the items in the third data set and the number p of the values of the items in the third data set are obtained. In the third data set TT = {TT1, TT2, TT3, …, TT i , …, TT m} contains m time interval items, and these time interval items can have p different values, where 1 ≤ p ≤ m. As examples of two extreme cases, if all time intervals are the same, then p = 1, and if all time intervals are different, then p = m. More likely, p is between 1 and m. The number p of items with different time intervals characterizes the change characteristics of an electrocardiogram (ECG) signal having specific cardiac rhythm characteristics.

[0118] Secondly, calculate the complexity index of the ECG signal in the target time period according to the aforementioned values of p and m of the third data set TT. For example, the complexity index C can be calculated by Equation (8):

[0119] C = p * m (8)

[0120] If the calculated complexity index C > the preset PAF threshold, it can be determined that an event conforming to the characteristics of atrial fibrillation has occurred in the target time period; if C ≤ the preset PAF threshold, it can be determined that no event conforming to the characteristics of atrial fibrillation has occurred in the target time period. This value can be determined empirically. For example, analyze the receiver operating characteristic (ROC) of an ECG database of hundreds of patients with and without PAF events, and determine the PAF threshold empirically. For example, in the method of the embodiment of the present invention, the PAF threshold is set to 356. It is obvious to those skilled in the art that the PAF threshold can have different values based on the parameters of the ECG signal.

[0121] Figure 9 Shows a histogram of the complexity index C of the ECG waveform signal within a 33 - second sliding window, and these ECG waveform signals are collected from 150 subjects. Among them, 510 represents the complexity index part of the ECG waveform signal without atrial fibrillation, 520 represents the complexity index part of the ECG waveform signal with atrial fibrillation, and 530 represents the preset PAF threshold.

[0122] According to the processing of the foregoing steps S210 - S280, it is possible to accurately detect short - term and long - term events conforming to the characteristics of paroxysmal atrial fibrillation (PAF) from the continuously collected human ECG signals in a sliding - window manner through relatively simple arithmetic operations. This ECG signal processing method occupies less computing and storage resources and can be well embedded in small - sized ECG signal acquisition devices with limited computing power to monitor the continuously collected ECG signals in real time. The long - term monitoring of the ECG signal according to the ECG signal processing method of the embodiment of the present application is also helpful for monitoring persistent atrial fibrillation with a duration exceeding 7 days and permanent atrial fibrillation with a duration exceeding 1 year, and is helpful for monitoring the burden of atrial fibrillation events.

[0123] An embodiment of the present invention further provides a computer-readable storage medium storing the steps of performing any of the foregoing electrocardiogram signal processing methods, and this computer-readable storage medium has the same beneficial effects as the foregoing electrocardiogram signal processing method.

[0124] An embodiment of the present invention further provides a portable electrocardiogram acquisition device, as Figure 10 shown.

[0125] Referring to Figure 10 , the portable electrocardiogram acquisition device 300 includes an ECG acquisition electrode 310, a microcontroller unit 320 electrically connected to the ECG acquisition electrode 310, and an ECG processing firmware 330 included in the microcontroller unit 320.

[0126] The ECG acquisition electrode 310 is used to continuously acquire the electrocardiogram signal of the human body, and the electrocardiogram signal is an electrocardiogram voltage signal.

[0127] The microcontroller unit 320 can be, for example, a microcontroller unit (MCU), an ARM microprocessor, an ASIC (application specific integrated circuit) chip, an FPGA (field programmable gate array) chip, etc., but is not limited thereto.

[0128] An ECG processing firmware 330 for processing and analyzing the electrocardiogram signal is provided in the microcontroller unit 320. Among them, the ECG processing firmware 330 includes:

[0129] A preprocessing module 331, which is used to sample the electrocardiogram signal in the target time period (corresponding to Window) collected by the ECG acquisition electrode 310 to obtain electrocardiogram sampling data, and divide the electrocardiogram sampling data in the target time period into sub-period electrocardiogram sampling data sets corresponding to multiple sub-periods (corresponding to Sub Window) respectively;

[0130] An averaging module 332, which is used to determine the average value of the data change within the sub-period corresponding to the electrocardiogram sampling data corresponding to each sub-period obtained by the processing of the preprocessing module 331, obtain a first data set of the data change values between sub-periods by calculating the data change values between the average values of the data changes within each adjacent sub-period, and calculate the root mean square difference value RMSSD for the items of the first data set;

[0131] An abnormal event detection module 333, which is used to determine whether an event conforming to the specific arrhythmia characteristics has occurred in the target time period according to the first data set and the root mean square difference value RMSSD;

[0132] An abnormal event processing module 334, comprising: a memory unit for recording information of the event if it is determined that an event conforming to the specific arrhythmia characteristics has occurred; and / or a communication unit for transmitting information of the event conforming to the specific arrhythmia characteristics.

[0133] Optionally, the ECG processing firmware 330 further comprises: a noise filtering unit for: determining the proportion of the number of items in the first data set whose inter-sub-period data change value is greater than the RMSSD in the first data set; if the proportion is between a predetermined upper proportion value and a lower proportion value, stopping the processing of the electrocardiogram signal in the target time period; if the proportion is not between the predetermined upper proportion value and the lower proportion value, notifying the abnormal event detection module 333 to perform detection processing on the first data set.

[0134] Optionally, the abnormal event detection module 333 specifically comprises:

[0135] A first processing unit (not shown) for obtaining a second data set containing the inter-sub-period data change value and the corresponding sub-time period by selecting the inter-sub-period data change value that satisfies the following conditions from the first data set processed by the averaging module: the inter-sub-period data change value is not less than the root mean square of the differences value RMSSD and the inter-sub-period data change value corresponding to the subsequent adjacent sub-time period is less than the root mean square of the differences value RMSSD;

[0136] A second processing unit (not shown) for obtaining a third data set of time intervals that satisfy the following conditions according to the time intervals between adjacent sub-time periods in the second data set obtained by the first processing unit: the time intervals between adjacent sub-time periods exceed a preset duration threshold;

[0137] A detection unit (not shown) for determining whether an event conforming to the atrial fibrillation characteristics has occurred in the target time period according to the values of the items in the third data set.

[0138] Further optionally, the detection unit is used to obtain the total number m of items in the third data set and the number p of item values in the third data set, calculate the complexity index C of the electrocardiogram signal in the target time period according to the total number m and the number p of item values, and if the calculated complexity index C > a preset PAF threshold, it can be determined that an event conforming to the atrial fibrillation characteristics has occurred in the target time period.

[0139] Optionally, the memory unit is further configured to store the electrocardiogram signals continuously collected by the ECG acquisition electrodes from the human body, and / or the communication unit is further configured to send the data of the electrocardiogram signals stored in the memory and the information of the events that have been recorded and conform to specific arrhythmia characteristics when it is detected that the small electrocardiogram acquisition device has established a network connection.

[0140] The portable electrocardiogram acquisition device according to the embodiment of the present invention has beneficial effects similar to those of the foregoing electrocardiogram signal processing method.

[0141] It should be noted that, according to the needs of implementation, each component / step described in the present application can be split into more components / steps, or two or more components / steps or partial operations of the components / steps can be combined into new components / steps to achieve the purpose of the embodiment of the present invention.

[0142] The methods and apparatuses, electronic devices, and storage media of the present disclosure can be implemented in many ways. For example, the methods and apparatuses, electronic devices, and storage media of the embodiments of the present invention can be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is only for illustration, and the steps of the method of the embodiments of the present invention are not limited to the specific order described above unless otherwise specifically stated. In addition, in some embodiments, the present disclosure can also be implemented as a program recorded in a recording medium, and these programs include machine-readable instructions for implementing the method according to the embodiments of the present invention. Therefore, the present disclosure also covers a recording medium storing a program for executing the method according to the embodiments of the present invention.

[0143] The description of the embodiments of the present invention is given for the purposes of illustration and description, and is not intended to be exhaustive or to limit the disclosure to the forms disclosed. Many modifications and variations are obvious to those of ordinary skill in the art. The embodiments are chosen and described in order to best explain the principles and practical applications of the disclosure, and to enable those of ordinary skill in the art to understand the disclosure and design various embodiments with various modifications suitable for specific purposes.

Claims

1. A method for processing electrocardiogram signals for a small electrocardiogram acquisition device, comprising: Sampling the ECG signal in the target time period to obtain ECG sampling data; Dividing the ECG sampling data within the target time period into sub-period ECG sampling data sets corresponding to each of the plurality of sub-periods; Determine the average value of the data change within the sub-period of the electrocardiogram sampling data corresponding to each sub-period; Obtain a first data set of data change values between sub-periods by calculating data change values between sub-periods between average values of data change within adjacent sub-periods; For the items of the first data set, calculating the root mean square difference (RMSSD); determining, based on the first data set and the root mean square difference (RMSSD), whether an event meeting a specific abnormal heart rhythm characteristic occurs within the target time period; If it is determined that the event that meets the specific abnormal heart rhythm characteristics has occurred, then information about the event is recorded, and / or information about the event that meets the specific abnormal heart rhythm characteristics is sent, The step of determining whether an event meeting a specific abnormal heart rhythm characteristic occurs within the target time period based on the first data set and the root mean square difference (RMSSD) includes: A second data set including the inter-sub-period data change values and the corresponding sub-time periods is obtained by selecting, from the first data set, inter-sub-period data change values that satisfy the following conditions: the inter-sub-period data change value is not less than a root mean square difference value RMSSD and the inter-sub-period data change values corresponding to the subsequent adjacent sub-time periods are less than the root mean square difference value RMSSD; Acquire, based on the time intervals between adjacent sub-time periods in the second data set, a third data set of time intervals that meet the following conditions: the time intervals between adjacent sub-time periods exceed a preset time threshold; According to the values of the items in the third data set, it is determined whether an event meeting the characteristics of atrial fibrillation occurs within the target time period.

2. The method according to claim 1, further comprising: Determine, in the first data set, a proportion of the number of items whose data change values between sub-periods are greater than the root mean square difference RMSSD in the first data set; If the proportion value is between a predetermined upper limit value and a predetermined lower limit value, the processing of the electrocardiogram signal in the target time period is stopped.

3. The method according to claim 1 or 2, wherein: Determining whether an event meeting the characteristics of atrial fibrillation has occurred within the target time period based on the values of the items in the third data set includes: Get the total number of items in the third data set m and the number of item values in the third data set p ; According to the total number m and the number of item values p Calculate the complexity index of the ECG signal in the target time period C ; If the calculated complexity index C > Preset PAF If the threshold is set, it can be determined that an event consistent with the characteristics of atrial fibrillation has occurred within the target time period.

4. The method according to claim 3, further comprising: storing data of electrocardiogram signals continuously collected from a human body, and / or When it is detected that the small ECG acquisition device has established a network connection, the stored ECG signal data and the recorded information of events that meet the characteristics of specific abnormal heart rhythm are sent.

5. A portable electrocardiogram acquisition device comprising: ECG acquisition electrodes are used to continuously collect the human body's electrocardiogram signals, wherein the electrocardiogram signals are electrocardiogram voltage signals; and A microcontroller unit including ECG processing firmware, The ECG processing firmware includes: a preprocessing module, configured to sample the ECG signals collected by the ECG acquisition electrodes during the target time period to obtain ECG sampling data, and to divide the ECG sampling data within the target time period into sub-time period ECG sampling data sets corresponding to each of the plurality of sub-time periods; an averaging module, configured to determine an average value of data variation within a sub-period of the ECG sampling data corresponding to each sub-period processed by the preprocessing module, obtain a first data set of inter-sub-period data variation values by calculating inter-sub-period data variation values between the average values of data variation within adjacent sub-periods, and calculate a root mean square difference (RMSSD) for each item in the first data set; an abnormal event detection module, configured to determine whether an event meeting a specific abnormal heart rhythm characteristic has occurred within the target time period based on the first data set and the root mean square difference value RMSSD; Abnormal event processing module, including: a memory unit, configured to record information about the event if it is determined that the event meeting the characteristics of the specific abnormal heart rhythm has occurred; and / or, a communication unit for transmitting information on recorded events meeting the characteristics of a specific abnormal heart rhythm, The abnormal event detection module includes: a first processing unit, configured to obtain a second data set including the inter-sub-period data change values and corresponding sub-time periods by selecting, from the first data set obtained by processing the averaging module, inter-sub-period data change values that meet the following conditions: the inter-sub-period data change value is not less than a root mean square difference (RMSSD) and the inter-sub-period data change value corresponding to the subsequent adjacent sub-time period is less than the root mean square difference (RMSSD); a second processing unit, configured to obtain, based on the time intervals between adjacent sub-time periods in the second data set obtained by the first processing unit, a third data set of time intervals that meets the following condition: the time intervals between adjacent sub-time periods exceed a preset time threshold; The detection unit is configured to determine whether an event meeting the characteristics of atrial fibrillation has occurred within the target time period based on the values of the items in the third data set.

6. The portable electrocardiogram acquisition device according to claim 5, wherein the ECG processing firmware further comprises: Noise filtering unit, used for: Determine, in the first data set, a proportion of items whose data change values between sub-periods are greater than the RMSSD in the first data set; If the proportion value is between a predetermined upper limit value and a predetermined lower limit value, stopping processing the electrocardiogram signal in the target time period; If the proportion value is not between a predetermined upper limit value and a predetermined lower limit value, the abnormal event detection module is notified to perform detection processing on the first data set.

7. The portable electrocardiogram acquisition device according to claim 5 or 6, wherein: The detection unit is used to obtain the total number of items in the third data set m and the number of item values in the third data set p , according to the total number m and the number of item values p Calculate the complexity index of the ECG signal in the target time period C , and if the calculated complexity index is C > Preset PAF If the threshold is set, it can be determined that an event consistent with the characteristics of atrial fibrillation has occurred within the target time period.

8. The portable electrocardiogram acquisition device according to claim 5, wherein: The memory unit is further used to store the electrocardiogram signals continuously collected from the human body by the ECG collection electrodes, and / or, The communication unit is further configured to send the ECG signal data stored in the memory and the information of the recorded events meeting the specific characteristics of abnormal heart rhythm when it is detected that the portable ECG acquisition device has established a network connection.

9. A computer-readable storage medium having computer program instructions stored thereon, wherein: When the program instructions are executed by the processor, the steps of the electrocardiogram signal processing method for a small electrocardiogram acquisition device described in any one of claims 1 to 4 are implemented.

Images