A method and system for detecting baseline noise of electrocardiogram signals

Abstract

The present invention discloses a method and system for detecting baseline noise of electrocardiogram signals. After preprocessing and signal segmentation of the collected electrocardiogram signals of non-pregnant adults, R-wave detection is performed on the segmented electrocardiogram signals, and then upper and lower envelope line detection is performed on the segmented electrocardiogram signals. The time periods on the upper and lower envelope lines corresponding to the QRS waveforms are deleted to obtain the upper envelope line and the lower envelope line without QRS waveforms. The mean value obtained by subtracting the lower envelope line without QRS waveforms from the upper envelope line without QRS waveforms at the corresponding time points is used as the baseline noise. It will not affect the evaluation result of the baseline noise due to the QRS waveform, and can correctly quantify the baseline noise of the electrocardiogram signal, solving the technical problems that the existing baseline noise evaluation method of electrocardiogram signals cannot correctly quantify the baseline noise of electrocardiogram signals and is difficult to be used for evaluating the baseline noise of IN-FECG signals.

Description

Technical Field

[0001] The present invention relates to the technical field of electrocardiogram monitoring, and particularly to a method and a system for detecting baseline noise of electrocardiogram signals. Background Art

[0002] ECG (Electrocardiogram) signals are the main choice for various health practitioners to determine important information about the human heart, and are studied to diagnose and detect heart abnormalities, such as cardiac cavity enlargement, detection of cardiovascular diseases, detection of ischemia, measurement of heart rate, biometric identification, and so on. Weak electrocardiogram signals are vulnerable to various noises, such as: power supply noise, environmental noise, and human body noise, etc. The presence of noise may damage the morphological characteristics of electrocardiogram signals, resulting in information errors and misdiagnosis. Therefore, it is necessary to evaluate the noise of electrocardiogram signal monitoring devices.

[0003] Non-invasive fetal electrocardiogram (IN-FECG) signals are collected by electrodes placed on the abdomen of pregnant women. IN-FECG signals are weaker than adult ECG signals, usually only 10-20 microvolts, and are easily submerged in noise. Since the measured object of IN-FECG signals is a pregnant woman and cannot be measured in real time at any time, it is necessary to change the measured object to a non-pregnant adult to evaluate the noise of the ECG monitoring device and quantify the circuit noise of the ECG monitoring device. The existing electrocardiogram signal quality evaluation methods use analysis signals that contain the QRS waveforms of electrocardiogram signals, and the changes of electrocardiogram signals will seriously affect the evaluation results and cannot correctly quantify the baseline noise of electrocardiogram signals. For this reason, the present invention provides a method and a system for detecting baseline noise of electrocardiogram signals. On the one hand, it does not require the measured object to be a pregnant woman, and on the other hand, it can correctly quantify the baseline noise of electrocardiogram signals and can be used to accurately evaluate the baseline noise of IN-FECG signals. Summary of the Invention

[0004] The present invention provides a method and a system for detecting baseline noise of electrocardiogram signals, which are used to solve the technical problems that the existing electrocardiogram signal baseline noise evaluation methods cannot correctly quantify the baseline noise of electrocardiogram signals and are difficult to be used for evaluating the baseline noise of IN-FECG signals.

[0005] In view of this, the first aspect of the present invention provides a method for detecting baseline noise of electrocardiogram signals, including:

[0006] Collecting electrocardiogram signals of non-pregnant adults;

[0007] Preprocessing the electrocardiogram signals, and the preprocessing includes analog-to-digital conversion and filtering processing;

[0008] The preprocessed electrocardiogram (ECG) signal is segmented using a fixed window length and a sliding length to obtain multiple segmented ECG signals, where the window length is equal to the sliding length.

[0009] Detect the R-wave sites for each segmented ECG signal, mark the R-wave sites and their corresponding intensities, and take a segmented ECG signal of a fixed length for marking a single QRS waveform.

[0010] Calculate the peak envelope of the preprocessed ECG signal to obtain the upper envelope and the lower envelope of the window signal.

[0011] Delete the time periods on the upper envelope and the lower envelope corresponding to the QRS waveforms to obtain the upper envelope and the lower envelope without QRS waveforms.

[0012] Subtract the lower envelope without QRS waveforms from the upper envelope without QRS waveforms at the corresponding time points, calculate the mean value, and use the mean value calculation result as the baseline noise.

[0013] Optionally, calculating the peak envelope of the preprocessed ECG signal to obtain the upper envelope and the lower envelope of the window signal includes:

[0014] Find the sampling points that satisfy the first constraint condition in the preprocessed ECG signal to obtain the first sampling point data set.

[0015] Use cubic spline interpolation to supplement the first sampling point data set into a first interpolation data set with the number of elements equal to the window length, and use the first interpolation data set as the upper envelope.

[0016] Find the sampling points that satisfy the second constraint condition in the preprocessed ECG signal to obtain the second sampling point data set.

[0017] Use cubic spline interpolation to supplement the second sampling point data set into a second interpolation data set with the number of elements equal to the window length, and use the second interpolation data set as the lower envelope.

[0018] The first constraint condition is:

[0019]

[0020] where t i is the i-th sampling time point, k is the minimum peak interval, is the segmented ECG signal corresponding to the i-th sampling time point t i ;

[0021] The second constraint condition is:

[0022]

[0023] where tj is the j-th sampling time point, is the j-th sampling time point t i corresponding segmented electrocardiogram signal.

[0024] Optionally, multiple QRS waveforms are included within the window length.

[0025] Optionally, for each segmented electrocardiogram signal, R-wave site detection is performed, the R-wave sites and corresponding intensities are marked, and a segmented electrocardiogram signal of a fixed length is taken for the marking of a single QRS waveform, including:

[0026] Use the Pan & Tompkins algorithm to perform R-wave site detection on each segmented electrocardiogram signal, mark the R-wave sites and corresponding intensities, and take a segmented electrocardiogram signal of a fixed length for the marking of a single QRS waveform.

[0027] Optionally, when preprocessing the electrocardiogram signal, analog-to-digital conversion samples the electrocardiogram signal at a fixed sampling frequency.

[0028] Optionally, the value range of the fixed sampling frequency is 250 Hz - 1000 Hz.

[0029] Optionally, the filtering process includes high-pass filtering, low-pass filtering, and notch filtering;

[0030] The high-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 1 Hz - 5 Hz;

[0031] The low-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 100 Hz - 150 Hz;

[0032] The notch filtering process includes filtering the electrocardiogram signal using a single-notch type infinite impulse response digital filter with a cut-off frequency of 50 Hz or 60 Hz and a bandwidth of 5 - 10 Hz.

[0033] The second aspect of the present invention provides an electrocardiogram signal baseline noise detection system, including:

[0034] A signal acquisition module for acquiring the electrocardiogram signal of non-pregnant adults;

[0035] A preprocessing module for preprocessing the electrocardiogram signal, and the preprocessing includes analog-to-digital conversion and filtering;

[0036] A signal segmentation module for cutting the preprocessed electrocardiogram signal using a fixed window length and a sliding length to obtain multiple segmented electrocardiogram signals, where the window length and the sliding length are equal;

[0037] The R-wave detection module is used to detect the R-wave positions of each segmented electrocardiogram signal, mark the R-wave positions and the corresponding intensities, and take a segmented electrocardiogram signal of a fixed length for marking a single QRS waveform;

[0038] The envelope calculation module is used to calculate the peak envelope of the preprocessed electrocardiogram signal to obtain the upper envelope and the lower envelope of the window signal;

[0039] The envelope processing module is used to delete the time periods corresponding to the QRS waveforms on the upper envelope and the lower envelope to obtain the upper envelope and the lower envelope without QRS waveforms;

[0040] The baseline noise evaluation module is used to subtract the lower envelope without QRS waveforms from the upper envelope without QRS waveforms at the corresponding time points, calculate the mean value, and use the mean value calculation result as the baseline noise.

[0041] Optionally, the envelope calculation module is specifically used for:

[0042] Finding sampling points that meet the first constraint condition in the preprocessed electrocardiogram signal to obtain a first sampling point data set;

[0043] Using cubic spline interpolation to supplement the first sampling point data set to a first interpolation data set with the number of elements equal to the window length, and using the first interpolation data set as the upper envelope;

[0044] Finding sampling points that meet the second constraint condition in the preprocessed electrocardiogram signal to obtain a second sampling point data set;

[0045] Using cubic spline interpolation to supplement the second sampling point data set to a second interpolation data set with the number of elements equal to the window length, and using the second interpolation data set as the lower envelope;

[0046] The first constraint condition is:

[0047]

[0048] where t i is the i-th sampling time point, k is the minimum peak interval, is the segmented electrocardiogram signal corresponding to the i-th sampling time point t i ;

[0049] The second constraint condition is:

[0050]

[0051] where t j is the j-th sampling time point, is the segmented electrocardiogram signal corresponding to the j-th sampling time point t i ;

[0052] Optionally, the R-wave detection module is specifically configured to:

[0053] Use the Pan & Tompkins algorithm to detect the R-wave sites for each segmented electrocardiogram signal, mark the R-wave sites and the corresponding intensities, and take a segmented electrocardiogram signal of a fixed length for marking a single QRS waveform.

[0054] As can be seen from the above technical solutions, the electrocardiogram signal baseline noise detection method and system provided by the present invention have the following advantages:

[0055] For the electrocardiogram signal baseline noise detection method provided by the present invention, after preprocessing and signal segmentation of the collected electrocardiogram signals of non-pregnant adults, R-wave detection is performed on the segmented electrocardiogram signals, and then upper and lower envelope line detections are performed on the segmented electrocardiogram signals. The time periods corresponding to the QRS waveforms on the upper and lower envelope lines are deleted to obtain the upper envelope line and the lower envelope line without QRS waveforms. The mean value obtained by subtracting the lower envelope line without QRS waveforms from the upper envelope line without QRS waveforms at the corresponding time points is used as the baseline noise. It will not affect the evaluation result of the baseline noise due to the QRS waveform, and can correctly quantify the baseline noise of the electrocardiogram signal, solving the technical problem that the existing electrocardiogram signal baseline noise evaluation method cannot correctly quantify the baseline noise of the electrocardiogram signal and is difficult to be used for evaluating the baseline noise of IN-FECG signals.

[0056] For the electrocardiogram signal baseline noise detection method provided by the present invention, the electrocardiogram signal is sampled using a fixed sampling frequency, which can avoid the problem that different sampling frequencies will produce different results due to spectral aliasing when the hardware of the electrocardiogram monitoring device is under the same interference source, thus affecting the evaluation result of the baseline noise. It can also avoid the problem that different sampling frequencies will produce different results due to frequency leakage when using different window lengths in the software of the electrocardiogram monitoring device under the same interference source, thus affecting the evaluation result of the baseline noise.

[0057] For the electrocardiogram signal baseline noise detection method provided by the present invention, multiple QRS waveforms are included within the window length, which can prevent the problem that subsequent processing cannot proceed normally due to the lack of QRS waveforms within the window caused by too small a window length.

[0058] The electrocardiogram signal baseline noise detection system provided by the present invention is used to execute the electrocardiogram signal baseline noise detection method provided by the present invention. Its principle and effect are the same as those of the electrocardiogram signal baseline noise detection method provided by the present invention, and will not be elaborated here. Description of the Drawings

[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0060] Figure 1 It is a schematic flowchart of a method for detecting baseline noise of electrocardiogram signals provided in the present invention;

[0061] Figure 2 It is a schematic structural diagram of a system for detecting baseline noise of electrocardiogram signals provided in the present invention. Detailed implementation manners

[0062] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0063] For ease of understanding, please refer to Figure 1 , and an embodiment of a method for detecting baseline noise of electrocardiogram signals provided in the present invention includes:

[0064] Step 101: Collect electrocardiogram signals of non-pregnant adults.

[0065] It should be noted that in the embodiments of the present invention, the measured object is selected as a non-pregnant adult, and the electrocardiogram signals are collected through electrodes placed on the abdominal skin surface of the non-pregnant adult, denoted as the original signal X(t).

[0066] Step 102: Preprocess the electrocardiogram signals, and the preprocessing includes analog-to-digital conversion and filtering.

[0067] It should be noted that the original signal X(t) is preprocessed, and the original signal X(t) is sampled through an analog-to-digital converter to obtain a discrete signal S(t). Among them, for the discrete signal S(t), it satisfies: f s is the sampling frequency, that is, f s is a fixed value, and the value range is 250 Hz - 1000 Hz. The filtering mainly filters out known interference signals through a filter. This part of the filtering mainly includes high-pass filtering, low-pass filtering, and notch filtering;

[0068] The high-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 1 Hz - 5 Hz;

[0069] The low-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 100 Hz - 150 Hz;

[0070] The notch filtering process includes filtering the electrocardiogram signal using a single-notch type infinite impulse response digital filter with a cut-off frequency of 50 Hz or 60 Hz and a bandwidth of 5 - 10 Hz.

[0071] Step 103: Cut the preprocessed electrocardiogram signal using a fixed window length and a sliding length to obtain multiple segmented electrocardiogram signals, where the window length and the sliding length are equal.

[0072] It should be noted that in the embodiments of the present invention, the electrocardiogram signal S(t) is cut using a fixed window length and a sliding length to obtain multiple segmented electrocardiogram signals.

[0073] The overall segmented electrocardiogram signal cut by the window length w can be expressed as:

[0074] S w (t) = [s1, s2, …, s y

[0075] where S w (t) is the preprocessed electrocardiogram signal, and s1, s2, …, s y are the 1st to yth segmented electrocardiogram signals obtained after cutting the electrocardiogram signal S(t) by the window length w.

[0076] The window length and the sliding length for cutting the electrocardiogram signal S(t) should be equal, that is, the window length and the signal between windows do not overlap, so as to avoid inaccurate noise evaluation results caused by signal overlap. And to prevent subsequent processing from not being able to proceed normally, the window length should not be too small, and it should be ensured that multiple QRS waveforms are included within the window, and the window time length should be greater than 4 seconds, that is, w > 4f s .

[0077] Step 104: Detect the R-wave sites for each segmented electrocardiogram signal, mark the R-wave sites and the corresponding intensities, and take a segmented electrocardiogram signal of a fixed length for marking a single QRS waveform.

[0078] It should be noted that the Pan & Tompkins algorithm can be used to detect the R-wave sites for each segmented electrocardiogram signal. Denote the set of sampling time points corresponding to the detected R-waves as T m = [t m1 , t m2 , …, t mn ​, where t m1 , t m2 , …, t mn are the sampling time points corresponding to the 1st to nth R waves. The entire QRS complex is marked using a fixed time length. The time length before the R wave peak is set as t rleft , and the time length after the R wave peak is set as t rright . Then, for the R wave at t m1 , the QRS time period is from t m1 - t rleft to t m1 + t rright . And so on, for the R wave at t mn , the QRS time period is from t mn - t rleft to t mn + t rright .

[0079] Step 105: Calculate the peak envelope of the preprocessed ECG signal to obtain the upper envelope and lower envelope of the window signal.

[0080] It should be noted that by calculating the peak envelope of the preprocessed ECG signal S w (t), the upper envelope E up and the lower envelope E lo of the window signal can be obtained. Specifically, the calculation process of the upper envelope E up is as follows:

[0081] Find the sampling points that satisfy the first constraint condition in the preprocessed ECG signal to obtain the first sampling point data set;

[0082] Use the cubic spline interpolation method to supplement the first sampling point data set into a first interpolation data set with the number of elements equal to the window length, and use the first interpolation data set as the upper envelope.

[0083] The first constraint condition is:

[0084]

[0085] where t i is the i-th sampling time point, k is the minimum peak interval, and its value range is 10 - 20, is the segmented ECG signal corresponding to the i-th sampling time point t i .

[0086] The t i that satisfies the first constraint condition is denoted as the data set {t i}. This data set is discontinuous within the entire window length, and the number of elements is less than the entire window length w. The data set {t i}Supplement it to a set with the number of elements equal to the length w of the entire window, denoted as E up = [e u1 , e u2 , …, e uw , E up is the upper envelope of the signal S w (t).

[0087] The calculation process of the lower envelope E lo is as follows:

[0088] Find the sampling points that satisfy the second constraint condition in the preprocessed electrocardiogram signal to obtain the second sampling point data set;

[0089] Use the cubic spline interpolation method to supplement the second sampling point data set to a second interpolation data set with the number of elements equal to the window length, and use the second interpolation data set as the lower envelope.

[0090] The second constraint condition is:

[0091]

[0092] where t j is the j-th sampling time point, is the segmented electrocardiogram signal corresponding to the j-th sampling time point t i .

[0093] The t j that satisfies the second constraint condition is denoted as the data set {t j}, and this data set is discontinuous within the entire window length and the number of elements is less than the entire window length w. Supplement the data set {t j} to a set with the number of elements equal to the entire window length w, denoted as E lo = [e l1 , e l2 , …, e lw , E lo is the lower envelope of the signal S w (t).

[0094] Step 106, Delete the time periods corresponding to the QRS waveforms on the upper envelope and the lower envelope to obtain the upper envelope E up without QRS waveforms and the lower envelope.

[0095] It should be noted that delete the signals in the time periods corresponding to the QRS waveforms on the upper envelope and the lower envelope to obtain the upper envelope and the lower envelope without QRS waveforms, denoted as E u and E l .

[0096] Step 107: Subtract the lower envelope line without QRS waveform at the corresponding time point from the upper envelope line without QRS waveform, calculate the mean value, and use the mean value calculation result as the baseline noise.

[0097] It should be noted that subtracting the lower envelope line without QRS waveform at the corresponding time point from the upper envelope line without QRS waveform, calculating the mean value, and obtaining the baseline noise. The expression formula is:

[0098]

[0099] Among them, I is the baseline noise. The unit of I is the same as the unit after the analog-to-digital converter conversion in Step 101, namely volts, microvolts, millivolts, etc. The larger I is, the greater the noise generated by the hardware and the surrounding environment on the device, and the worse the corresponding signal quality; the smaller I is, the smaller the noise generated by the hardware and the surrounding environment on the device, and the better the corresponding signal quality.

[0100] The method for detecting the baseline noise of electrocardiogram signals provided by the present invention preprocesses and segments the collected electrocardiogram signals of non-pregnant adults, then detects the R waves of the segmented electrocardiogram signals, and then detects the upper and lower envelope lines of the segmented electrocardiogram signals. Delete the time periods corresponding to QRS waveforms on the upper and lower envelope lines to obtain the upper envelope line and lower envelope line without QRS waveforms. Take the mean value of the upper envelope line without QRS waveform minus the lower envelope line without QRS waveform at the corresponding time point as the baseline noise, which will not affect the evaluation result of the baseline noise due to the QRS waveform, and can correctly quantify the baseline noise of the electrocardiogram signal, solving the technical problems that the existing methods for evaluating the baseline noise of electrocardiogram signals cannot correctly quantify the baseline noise of electrocardiogram signals and are difficult to be used for evaluating the baseline noise of IN-FECG signals.

[0101] The method for detecting the baseline noise of electrocardiogram signals provided by the present invention samples the electrocardiogram signals at a fixed sampling frequency, which can avoid the problem that different sampling frequencies will produce different results due to spectral aliasing when the hardware of the electrocardiogram monitoring device is under the same interference source, thus affecting the evaluation result of the baseline noise. It can also avoid the problem that different sampling frequencies will produce different results due to frequency leakage when using different window lengths in the software of the electrocardiogram monitoring device under the same interference source, thus affecting the evaluation result of the baseline noise.

[0102] The method for detecting the baseline noise of electrocardiogram signals provided by the present invention includes multiple QRS waveforms within the window length, which can prevent the problem that subsequent processing cannot proceed normally due to the lack of QRS waveforms in the window caused by too small a window length.

[0103] For ease of understanding, please refer to Figure 2 , and an embodiment of a system for detecting the baseline noise of electrocardiogram signals provided in the present invention includes:

[0104] A signal acquisition module for acquiring the electrocardiogram (ECG) signals of non-pregnant adults;

[0105] A preprocessing module for preprocessing the ECG signals, where the preprocessing includes analog-to-digital conversion and filtering;

[0106] A signal segmentation module for cutting the preprocessed ECG signals using a fixed window length and a sliding length to obtain multiple segmented ECG signals, where the window length and the sliding length are equal;

[0107] An R-wave detection module for detecting the R-wave positions in each segmented ECG signal, marking the R-wave positions and the corresponding intensities, and taking a segmented ECG signal of a fixed length for marking a single QRS waveform;

[0108] An envelope calculation module for calculating the peak envelopes of the preprocessed ECG signals to obtain the upper envelope and the lower envelope of the window signals;

[0109] An envelope processing module for deleting the time periods corresponding to the QRS waveforms on the upper envelope and the lower envelope to obtain the upper envelope and the lower envelope without QRS waveforms;

[0110] A baseline noise assessment module for subtracting the lower envelope without QRS waveforms from the upper envelope without QRS waveforms at the corresponding time points and calculating the mean value, and taking the mean value calculation result as the baseline noise.

[0111] Specifically, the envelope calculation module is used for:

[0112] Finding sampling points that meet the first constraint condition in the preprocessed ECG signals to obtain a first sampling point data set;

[0113] Using the cubic spline interpolation method to supplement the first sampling point data set to a first interpolation data set with the number of elements equal to the window length, and taking the first interpolation data set as the upper envelope;

[0114] Finding sampling points that meet the second constraint condition in the preprocessed ECG signals to obtain a second sampling point data set;

[0115] Using the cubic spline interpolation method to supplement the second sampling point data set to a second interpolation data set with the number of elements equal to the window length, and taking the second interpolation data set as the lower envelope;

[0116] The first constraint condition is:

[0117]

[0118] where t i is the i-th sampling time point, and k is the minimum peak interval, For the i-th sampling time point t i The corresponding segmented electrocardiogram signal;

[0119] The second constraint is:

[0120]

[0121] where t j is the j-th sampling time point, and is the segmented electrocardiogram signal corresponding to the j-th sampling time point t i The corresponding segmented electrocardiogram signal.

[0122] The R-wave detection module is specifically used for:

[0123] Using the Pan & Tompkins algorithm to detect the R-wave sites of each segmented electrocardiogram signal, mark the R-wave sites and the corresponding intensities, and take a segmented electrocardiogram signal of a fixed length for the marking of a single QRS waveform.

[0124] Multiple QRS waveforms are included within the window length.

[0125] The value range of the fixed sampling frequency is 250Hz - 1000Hz.

[0126] The filtering process includes high-pass filtering, low-pass filtering, and notch filtering;

[0127] The high-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 1Hz - 5Hz;

[0128] The low-pass filtering process includes filtering the electrocardiogram signal using a Butterworth-type infinite impulse response digital filter with a cut-off frequency of 100Hz - 150Hz;

[0129] The notch filtering process includes filtering the electrocardiogram signal using a single-notch type infinite impulse response digital filter with a cut-off frequency of 50Hz or 60Hz and a bandwidth of 5 - 10Hz.

[0130] The electrocardiogram (ECG) signal baseline noise detection system provided by the present invention preprocesses and segments the collected ECG signals of non-pregnant adults. After detecting the R waves in the segmented ECG signals, it then detects the upper and lower envelope lines of the segmented ECG signals, deletes the time periods corresponding to the QRS waveforms on the upper and lower envelope lines, and obtains the upper and lower envelope lines without QRS waveforms. The mean value obtained by subtracting the lower envelope line without QRS waveforms from the upper envelope line without QRS waveforms at the corresponding time points is used as the baseline noise. This method will not be affected by the QRS waveforms when evaluating the baseline noise, and can correctly quantify the baseline noise of the ECG signals, solving the technical problems that the existing methods for evaluating the baseline noise of ECG signals cannot correctly quantify the baseline noise of ECG signals and are difficult to be used for evaluating the baseline noise of IN-FECG signals.

[0131] The electrocardiogram (ECG) signal baseline noise detection system provided by the present invention samples the ECG signals at a fixed sampling frequency, which can avoid the problem that different sampling frequencies will produce different results due to spectral aliasing when the hardware of the ECG monitoring device is under the same interference source, thus affecting the evaluation result of the baseline noise. It can also avoid the problem that different sampling frequencies will produce different results due to frequency leakage when using different window lengths in the software of the ECG monitoring device under the same interference source, thus affecting the evaluation result of the baseline noise.

[0132] The electrocardiogram (ECG) signal baseline noise detection system provided by the present invention includes multiple QRS waveforms within the window length, which can prevent the problem that subsequent processing cannot proceed normally due to the lack of QRS waveforms in the window caused by an overly small window length.

[0133] The electrocardiogram (ECG) signal baseline noise detection system provided in the embodiments of the present invention is used to execute the electrocardiogram (ECG) signal baseline noise detection method in the foregoing embodiments of the electrocardiogram (ECG) signal baseline noise detection method. Its principle is the same as that of the electrocardiogram (ECG) signal baseline noise detection method in the foregoing embodiments of the electrocardiogram (ECG) signal baseline noise detection method, and will not be elaborated here.

[0134] As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting baseline noise of electrocardiogram signals, characterized in that, Including: Collecting the electrocardiogram (ECG) signals of non-pregnant adults; Preprocessing the ECG signals, where the preprocessing includes analog-to-digital conversion and filtering; Cutting the preprocessed ECG signals using a fixed window length and a sliding length to obtain multiple segmented ECG signals, where the window length and the sliding length are equal; Detecting the R-wave sites for each segmented ECG signal, marking the R-wave sites and the corresponding intensities, and taking a segmented ECG signal of a fixed length for marking a single QRS waveform; Calculating the peak envelope of the preprocessed ECG signals to obtain the upper envelope and the lower envelope of the window signals; Deleting the time periods on the upper envelope and the lower envelope corresponding to the QRS waveforms to obtain the upper envelope and the lower envelope without QRS waveforms; Subtracting the lower envelope without QRS waveforms from the upper envelope without QRS waveforms at the corresponding time points and calculating the mean value, and taking the mean value calculation result as the baseline noise; 2. The electrocardiogram signal baseline noise detection method according to claim 1, wherein Calculating the peak envelope of the preprocessed ECG signals to obtain the upper envelope and the lower envelope of the window signals, including: Searching for sampling points that meet the first constraint condition in the preprocessed ECG signals to obtain a first sampling point data set; Using the cubic spline interpolation method to supplement the first sampling point data set into a first interpolation data set with the number of elements equal to the window length, and taking the first interpolation data set as the upper envelope; Searching for sampling points that meet the second constraint condition in the preprocessed ECG signals to obtain a second sampling point data set; Using the cubic spline interpolation method to supplement the second sampling point data set into a second interpolation data set with the number of elements equal to the window length, and taking the second interpolation data set as the lower envelope; The first constraint condition is: where t i is the i-th sampling time point, k is the minimum peak interval, and s ti is the segmented electrocardiogram signal corresponding to the i-th sampling time point t i ; The second constraint condition is: where t j is the j-th sampling time point, and i is the segmented electrocardiogram signal corresponding to the j-th sampling time point t.

3. The electrocardiogram signal baseline noise detection method according to claim 1, characterized in that There are multiple QRS waveforms within the window length.

4. The electrocardiogram signal baseline noise detection method according to claim 1, characterized in that Detecting the R-wave sites for each segmented ECG signal, marking the R-wave sites and the corresponding intensities, and taking a segmented ECG signal of a fixed length for marking a single QRS waveform, including: Using the Pan&Tompkins algorithm to detect the R-wave sites for each segmented ECG signal, marking the R-wave sites and the corresponding intensities, and taking a segmented ECG signal of a fixed length for marking a single QRS waveform.

5. The electrocardiogram signal baseline noise detection method according to claim 1, characterized in that When preprocessing the ECG signals, the analog-to-digital conversion samples the ECG signals using a fixed sampling frequency.

6. The electrocardiogram signal baseline noise detection method according to claim 5, characterized in that, The value range of the fixed sampling frequency is 250 Hz - 1000 Hz.

7. The electrocardiogram signal baseline noise detection method according to claim 1, characterized in that The filtering includes high-pass filtering, low-pass filtering, and notch filtering; The high-pass filtering includes filtering the ECG signals using a Butterworth type infinite impulse response digital filter with a cut-off frequency of 1 Hz - 5 Hz; The low-pass filtering includes filtering the ECG signals using a Butterworth type infinite impulse response digital filter with a cut-off frequency of 100 Hz - 150 Hz; The notch filtering includes filtering the ECG signals using a single notch type infinite impulse response digital filter with a cut-off frequency of 50 Hz or 60 Hz and a bandwidth of 5 - 10 Hz.

8. An electrocardiogram signal baseline noise detection system, characterized in that Including: A signal acquisition module for collecting the ECG signals of non-pregnant adults; A preprocessing module for preprocessing the ECG signals, where the preprocessing includes analog-to-digital conversion and filtering; A signal segmentation module, which is used to cut the preprocessed electrocardiogram (ECG) signal using a fixed window length and a sliding length to obtain multiple segmented ECG signals, where the window length is equal to the sliding length; An R-wave detection module, which is used to detect the R-wave sites of each segmented ECG signal, mark the R-wave sites and the corresponding intensities, and take a segmented ECG signal of a fixed length to mark a single QRS waveform; An envelope calculation module, which is used to calculate the peak envelope of the preprocessed ECG signal to obtain the upper envelope and the lower envelope of the window signal; An envelope processing module, which is used to delete the time periods corresponding to the QRS waveforms on the upper envelope and the lower envelope to obtain the upper envelope and the lower envelope without QRS waveforms; A baseline noise evaluation module, which is used to subtract the lower envelope without QRS waveforms from the upper envelope without QRS waveforms at the corresponding time points, calculate the mean value, and use the mean value calculation result as the baseline noise.

9. The electrocardiogram signal baseline noise detection system according to claim 8, characterized in that, The envelope calculation module is specifically used for: finding sampling points that meet the first constraint condition in the preprocessed ECG signal to obtain a first sampling point data set; using the cubic spline interpolation method to supplement the first sampling point data set into a first interpolation data set with the number of elements equal to the window length, and using the first interpolation data set as the upper envelope; finding sampling points that meet the second constraint condition in the preprocessed ECG signal to obtain a second sampling point data set; using the cubic spline interpolation method to supplement the second sampling point data set into a second interpolation data set with the number of elements equal to the window length, and using the second interpolation data set as the lower envelope; The first constraint condition is: where t i is the i-th sampling time point, k is the minimum peak interval, is the segmented electrocardiogram signal corresponding to the i-th sampling time point t i ; The second constraint condition is: where t j is the j-th sampling time point, is the segmented electrocardiogram signal corresponding to the j-th sampling time point t i .

10. The electrocardiogram signal baseline noise detection system according to claim 8, characterized in that, The R-wave detection module is specifically used for: using the Pan & Tompkins algorithm to detect the R-wave sites of each segmented ECG signal, mark the R-wave sites and the corresponding intensities, and take a segmented ECG signal of a fixed length to mark a single QRS waveform.

Images