Search

EP-4437940-B1 - METHOD AND DEVICE FOR EVALUATING ELECTROCARDIOSIGNAL QUALITY

EP4437940B1EP 4437940 B1EP4437940 B1EP 4437940B1EP-4437940-B1

Inventors

  • YIN, Haibo

Dates

Publication Date
20260506
Application Date
20230817

Claims (15)

  1. A computer implemented electrocardiogram signal quality evaluation method using an electronic device (100) for evaluating whether an electrocardiogram signal is a normal signal, comprising: obtaining (S201), by a functional module in the electronic device (100), a first electrocardiogram signal; and calculating (S202), by the functional module in the electronic device (100), a peak variability feature and a peak number variability feature of the first electrocardiogram signal, wherein the peak variability feature is obtained by calculating a ratio between a standard deviation and a mean value of peaks in a plurality of sample segments of the first electrocardiogram signal obtained by performing a signal segmentation in time domain using a pre-set duration on the first electrocardiogram signal; wherein the peak number variability feature is obtained by calculating a ratio of the difference between the largest value of the peak numbers in a sample segment of the plurality of sample segments and the smallest value of the peak numbers in a sample segment of the plurality of sample segments to a median value of the peak numbers in a sample segment of the plurality of sample segments; wherein if the peak variability feature of the first electrocardiogram signal is less than a first threshold, and the peak number variability feature of the first electrocardiogram signal is less than a second threshold, determining (S2034), by the functional module in the electronic device (100), that the first electrocardiogram signal is a normal signal; wherein the first threshold represents an abnormal dividing point of the peak variability feature, the first threshold is determined based on peak variability features of a plurality of electrocardiogram signal samples, wherein the second threshold represents an abnormal dividing point of the peak number variability feature, and the second threshold is determined based on peak number variability features of the plurality of electrocardiogram signal samples.
  2. The method according to claim 1, wherein the obtaining a first electrocardiogram signal comprises: obtaining an original electrocardiogram signal with first duration; and performing filtering processing on the original electrocardiogram signal to obtain the first electrocardiogram signal.
  3. The method according to claim 2, wherein the performing filtering processing on the original electrocardiogram signal comprises: performing low-pass filtering processing and high-pass filtering processing on the original electrocardiogram signal.
  4. The method according to any one of claims 1 to 3, wherein the calculating a peak variability feature and a peak number variability feature of the first electrocardiogram signal comprises: performing segmentation processing on the first electrocardiogram signal to obtain the plurality of signal segments; calculating a peak list and a peak number of each signal segment, wherein the peak list comprises N peaks, and N represents the peak number; obtaining the peak variability feature of the first electrocardiogram signal based on peak lists of the plurality of signal segments; and obtaining the peak number variability feature of the first electrocardiogram signal based on peak numbers of the plurality of signal segments.
  5. The method according to claim 4, wherein the signal segment comprises a plurality of discrete data points, and the calculating a peak list of each signal segment comprises: calculating an absolute value of a difference between each two adjacent data points in a first signal segment, wherein the first signal segment is any signal segment in the signal segments, a peak list of the first signal segment comprises one or more first peaks, the first peak is an absolute value of a difference that is greater than an absolute value of a previous difference, greater than an absolute value of a next difference, and greater than the peak searching threshold, and the peak searching threshold is determined based on absolute values of differences respectively corresponding to the plurality of signal segments of the first electrocardiogram signal.
  6. The method according to claim 5, wherein a manner for determining the peak searching threshold comprises: calculating a first product of a first constant and a first percentile in a set of the absolute values of the differences corresponding to the plurality of signal segments of the first electrocardiogram signal; calculating a second product of a second constant and a second percentile in the set of the absolute values of the differences corresponding to the plurality of signal segments of the first electrocardiogram signal; and calculating a difference between the first product and the second product to obtain the peak searching threshold.
  7. The method according to any one of claims 4 to 6, wherein the obtaining the peak variability feature of the first electrocardiogram signal based on peak lists of the plurality of signal segments comprises: calculating the standard deviation and the mean value of peaks of the plurality of signal segments of the first electrocardiogram signal; and calculating the ratio of the standard deviation to the mean value of the peaks of the plurality of signal segments to obtain the peak variability feature of the first electrocardiogram signal.
  8. The method according to any one of claims 4 to 6, wherein the obtaining the peak number variability feature of the first electrocardiogram signal based on peak numbers of the plurality of signal segments comprises: calculating a first difference between the largest value and the smallest value of the peak numbers of the plurality of signal segments of the first electrocardiogram signal; calculating the median value of the peak numbers of the plurality of signal segments of the first electrocardiogram signal; and calculating the ratio of the first difference value to the median value to obtain the peak number variability feature of the first electrocardiogram signal.
  9. The method according to any one of claims 1 to 8, wherein a manner of determining the first threshold and the second threshold comprises: obtaining the plurality of electrocardiogram signal samples, wherein duration of each of the plurality of electrocardiogram signal samples is the first duration; calculating a peak variability feature and a peak number variability feature of each electrocardiogram signal sample; obtaining the first threshold based on the peak variability features of the plurality of electrocardiogram signal samples; and obtaining the second threshold based on the peak number variability features of the plurality of electrocardiogram signal samples.
  10. The method according to claim 9, wherein the obtaining the first threshold based on the peak variability features of the plurality of electrocardiogram signal samples comprises: calculating a third product of a third constant and a third percentile in the peak variability features of the plurality of electrocardiogram signal samples; calculating a fourth product of a fourth constant and a fourth percentile in the peak variability features of the plurality of electrocardiogram signal samples; and calculating a difference between the third product and the fourth product to obtain the first threshold.
  11. The method according to claim 9, wherein the obtaining the second threshold based on the peak number variability features of the plurality of electrocardiogram signal samples comprises: calculating a fifth product of a fifth constant and a fifth percentile in the peak number variability features of the plurality of electrocardiogram signal samples; calculating a sixth product of a sixth constant and a sixth percentile in the peak number variability features of the plurality of electrocardiogram signal samples; and calculating a difference between the fifth product and the sixth product to obtain the second threshold.
  12. The method according to any one of claims 1 to 11, further comprising: if the peak variability feature of the first electrocardiogram signal is greater than the first threshold, and the peak number variability feature of the first electrocardiogram signal is greater than the second threshold, determining (S2031) that the first electrocardiogram signal is an abnormal signal; if the peak variability feature of the first electrocardiogram signal is greater than or equal to the first threshold, and the peak number variability feature of the first electrocardiogram signal is less than or equal to the second threshold, determining (S2032) that the first electrocardiogram signal is a noise signal; or if the peak variability feature of the first electrocardiogram signal is less than or equal to the first threshold, and the peak number variability feature of the first electrocardiogram signal is greater than or equal to the second threshold, determining (S2033) that the first electrocardiogram signal is a noise signal.
  13. The method according to any one of claims 9 to 12, wherein the obtaining the plurality of electrocardiogram signal samples comprises: obtaining a plurality of original electrocardiogram signals; performing data segmentation on the plurality of original electrocardiogram signals to obtain a plurality of original electrocardiogram signals with the first duration; and performing low-pass filtering processing and high-pass filtering processing on the original electrocardiogram signals with the first duration to obtain the plurality of electrocardiogram signal samples.
  14. The method according to any one of claims 9 to 12, wherein a manner of calculating the peak variability feature of each electrocardiogram signal sample is the same as a manner of calculating the peak variability feature of the first electrocardiogram signal; and a manner of calculating the peak number variability feature of each electrocardiogram signal sample is the same as a manner of calculating the peak number variability feature of the first electrocardiogram signal.
  15. An electronic device (100), comprising a processor (110), wherein the processor is configured to run a computer program stored in a memory (121), to cause the electronic device to implement the method according to any one of claims 1 to 14.

Description

TECHNICAL FIELD This disclosure relates to the field of signal evaluation technologies, and in particular, the invention relates to an electrocardiogram signal quality evaluation method and an electronic device. BACKGROUND An electrocardiogram signal is a weak signal of non-linearity, nonstationarity, and randomness, usually with a maximum magnitude of mV. Therefore, electrocardiogram signal monitoring usually requires an electrocardiogram monitoring apparatus to be closely attached to a surface of a living body to collect an electrocardiogram signal of the living body. It is clear that, due to features of the electrocardiogram signal and a scenario of the electrocardiogram signal monitoring, the electrocardiogram signal is highly susceptible to interference from within the living body (for example, electromyographic interference and respiratory interference) or outside the living body (for example, power frequency interference and a signal pickup process). At present, a collected electrocardiogram signal may be accompanied by a significant number of noise or abnormal signals (resulting from an abnormal acquisition process), resulting in the electrocardiogram signal being swamped with the noise and/or abnormal signals and losing some original features. In particular, when the electrocardiogram signal has some pathological features, these pathological features may be disturbed by the noise and/or abnormal signals, resulting in a reduced accuracy of an early warning algorithm associated with a pathological signal. US 2012/016249 A1 relates to a method and system for assessing the quality of ECG signals by analyzing a variety of signal features and classifying the signals accordingly. US 5,842,997 A relates to a method and apparatus for identifying artifact-contaminated ECG data segments. SUMMARY In view of this, an object of the present invention is to provides an electrocardiogram signal quality evaluation method and an electronic device which can detect a normal signal in an electrocardiogram signal. This object is solved by the attached independent claims and further embodiments and improvements of the invention are listed in the attached dependent claims. Hereinafter, up to the "brief description of the drawings", expressions like "...aspect according to the invention", "according to the invention", or "the present invention", relate to technical teaching of the broadest embodiment as claimed with the independent claims. Expressions like "implementation", "design", "optionally", "preferably", "scenario", "aspect" or similar relate to further embodiments as claimed, and expressions like "example", "...aspect according to an example", "the disclosure describes", or "the disclosure" describe technical teaching which relates to the understanding of the invention or its embodiments, which, however, is not claimed as such. To achieve the foregoing objective, the following technical solutions are used in this application: According to a first aspect according to the invention, this disclosure provides an electrocardiogram signal quality evaluation method. The method includes: obtaining a first electrocardiogram signal; andcalculating a peak variability feature and a peak number variability feature of the first electrocardiogram signal, whereif the peak variability feature of the first electrocardiogram signal is less than a first threshold, and the peak number variability feature of the first electrocardiogram signal is less than a second threshold, the first electrocardiogram signal is a normal signal, where the first threshold represents an abnormal dividing point of the peak variability feature, the first threshold is determined based on peak variability features of a plurality of electrocardiogram signal samples, the second threshold represents an abnormal dividing point of the peak number variability feature, and the second threshold is determined based on the peak number variability features of the plurality of electrocardiogram signal samples. In this disclosure, the peak variability feature and the peak number variability features of the plurality of electrocardiogram signal samples are counted, based on the peak variability feature of the plurality of electrocardiogram signal samples as one abnormal dividing point, the peak number variability features of the plurality of electrocardiogram signal samples is used as another abnormal dividing point; and a type of a to-be-evaluated electrocardiogram signal is determined based on a relationship between the peak variability feature and the peak number variability feature of the to-be-evaluated electrocardiogram signal and the two abnormal dividing points. A normal signal may be accurately distinguished by this method. In an implementation of the first aspect, obtaining a first electrocardiogram signal includes: obtaining an original electrocardiogram signal with first duration; andperforming filtering processing on the original electrocardiogram signal to obt