QRS complex identification in electrocardiogram signals

• US 9,414,761 B2
• Filed: 12/17/2014
• Issued: 08/16/2016
• Est. Priority Date: 06/02/2014
• Status: Active Grant

• Current Assignee: Indian Institute of Technology Delhi
• Original Assignee: Indian Institute of Technology Delhi
• Inventors: Bhaumik, Basabi, Jain, Sanjeev Kumar
• Primary Examiner(s): Schaetzle, Kennedy
• Application Number:
  US14/573,037
• Publication Number:
  US 20150342489A1
• Time in Patent Office: 608 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: A61B 5/349 Detecting specific parameters of the electrocardiograph cycle
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/366 Detecting abnormal QRS complex, e.g. widening
• CPC Class Codes: A61B 5/7225 Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation A61B5/30, A61B5/316 take precedence; transformers for specific applications H01F38/00; amplifiers H03F

Abstract

The present subject matter relates to processing of Electrocardiogram (ECG) signals, and in particular, relates to identifying a QRS complex in an ECG signal. The method includes receiving, and filtering the ECG signal by passing through at least one of a first low-pass filter and a high-pass filter to obtain a filtered ECG signal. The filtered ECG signal is processed based on a moving average technique. Further, a search region is identified in the processed ECG signal, and a maximum amplitude peak is identified in a time interval of the filtered ECG signal that corresponds to a time span of the search region of the processed ECG signal. The maximum amplitude peak is an R peak of the QRS complex. Subsequently, a Q peak and an S peak of the QRS complex is identified based on the R peak.

Claims

1. A method for identifying a QRS complex in an Electrocardiogram (ECG) signal, the method comprising:

receiving the ECG signal captured by an ECG recorder, wherein the ECG signal is indicative of an electrical activity of heart of an individual;

filtering the ECG signal by passing the ECG signal through at least one of a first low-pass filter and a high-pass filter to obtain a filtered ECG signal;

squaring the filtered ECG signal to obtain a squared filtered ECG signal;

generating an envelope ECG signal by capturing peaks of the squared filtered ECG signal;

determining a moving average ECG signal based on a moving average of sample points associated with the envelope ECG signal, wherein a sample point is indicative of an amplitude of the envelope ECG signal at a time instance; and

filtering the moving average ECG signal to obtain a processed ECG signal, by passing the moving average ECG signal through a second low-pass filter;

identifying a search region in the processed ECG signal, wherein the search region is indicative of a time span between two consecutive peaks of the processed ECG signal;

identifying a maximum amplitude peak in a time interval of the filtered ECG signal that corresponds to the time span of the search region of the processed ECG signal, wherein the maximum amplitude peak is an R peak of an R wave of the QRS complex; and

identifying a Q peak of a Q wave, and an S peak of an S wave, of the QRS complex based on the R peak of the R wave.

2. The method as claimed in claim 1 further comprising down-sampling the ECG signal for passing the sampled ECG signal through the at least one of the first low-pass filter and the high-pass filter, wherein the ECG signal is down-sampled based on a predefined reduction factor.

3. The method as claimed in claim 1, wherein the identifying the Q peak of the Q wave of the QRS complex comprises:

identifying a first sub-search region of the filtered ECG signal, wherein the first sub-search region is a region comprising a predefined set of sample points of the filtered ECG signal before a time instance of the R peak;

determining a first peak point in a segment of the filtered ECG signal, within the first sub-search region, wherein the segment comprises a predefined number of consecutive sample points with slope greater than a predefined slope threshold, and wherein the first peak point is one of the predefined number of consecutive sample points in the segment;

determining a second peak point that corresponds to a peak of the filtered ECG signal in the first sub-search region, wherein the second peak point is determined as a zero crossing point of a slope signal, wherein the slope signal is obtained from three-point derivatives of filtered ECG signal in the first sub-search region; and

wherein the Q peak of the Q wave of the QRS complex is identified as one of the first peak point and the second peak point which is closer to the R peak of the R wave of the QRS complex.

4. The method as claimed in claim 3, further comprising

determining a third peak point of the filtered ECG signal, based on a minimum value of the filtered ECG signal in the first sub-search region, when the first peak point and the second peak point are not determined, wherein the Q peak of the Q wave of the QRS complex is identified as the third peak point.

5. The method as claimed in claim 1, wherein the identifying the S peak of the S wave of the QRS complex comprises:

identifying a second sub-search region of the filtered ECG signal, wherein the second sub-search region is a region comprising a predefined set of sample points of the filtered ECG signal after a time instance of the R peak;

determining a first peak point in a segment of the filtered ECG signal, within the second sub-search region, wherein the segment comprises a predefined number of consecutive sample points with slope greater than a predefined slope threshold, and wherein the first peak point is one of the predefined number of consecutive sample points in the segment;

determining a second peak point that corresponds to a peak of the filtered ECG signal in the second sub-search region, wherein the second peak point is determined as a zero crossing point of a slope signal, wherein the slope signal is obtained from three-point derivatives of filtered ECG signal in the second sub-search region; and

wherein the S peak of the S wave of the QRS complex is identified as one of the first peak point and the second peak point which is closer to the R peak of the R wave of the QRS complex.

6. The method as claimed in claim 5, further comprising

determining a third peak point of the filtered ECG signal, based on a minimum value of the filtered ECG signal in the second sub-search region, when the first peak point and the second peak point are not determined, wherein the S peak of the S wave of the QRS complex is identified as the third peak point.

7. A QRS identification system for identifying a QRS complex in an Electrocardiogram (ECG) signal, the QRS identification system comprising:

a processing unit;

a signal receiving block, coupled to the processing unit, to

receive the ECG signal captured by an ECG recorder, wherein the ECG signal is indicative of an electrical activity of heart of an individual; and

filter the ECG signal by passing the ECG signal through at least one of a first low-pass filter and a high-pass filter to obtain a filtered ECG signal;

a signal processing block, coupled to the processing unit, to

square the filtered ECG signal to obtain a squared filtered ECG signal;

generate an envelope ECG signal by capturing peaks of the squared filtered ECG Signal;

determine a moving average ECG signal based on a moving average of sample points associated with the envelope ECG signal, wherein a sample point is indicative of an amplitude of the envelope ECG signal at a time instance; and

filter the moving average ECG signal to obtain a processed ECG signal, by passing the moving average ECG signal through a second low-pass filter; and

a QRS identification block, coupled to the processing unit, to determine a search region in the processed ECG signal, wherein the search region is indicative of a time span between two consecutive peaks of the processed ECG signal;

detect a maximum amplitude peak in a time interval of the filtered ECG signal that corresponds to the time span of the search region of the processed ECG signal, wherein the maximum amplitude peak is an R peak of an R wave of the QRS complex; and

identify a Q peak of a Q wave and an S peak of an S wave of the QRS complex, based on the R peak of the R wave.

8. The QRS identification system as claimed in claim 7, wherein the signal receiving block is coupled to the processing unit to down-sample the ECG signal for passing the sampled ECG signal through the at least one of the first low-pass filter and the high-pass filter, wherein the ECG signal is down-sampled based on a predefined reduction factor.

9. The QRS identification system as claimed in claim 7, wherein the QRS identification block is coupled to the processing unit to identify the two consecutive peaks of the processed ECG signal, for the determination of the search region, based on a peak threshold, the two consecutive peaks being indicative of peaks with an amplitude above the peak threshold.

10. The QRS identification system as claimed in claim 9, wherein the QRS identification block is coupled to the processing unit to determine the peak threshold for a predefined number of search regions, based on average of peaks occurred in a previous set of the predefined number of search regions.

11. The QRS identification system as claimed in claim 7, wherein the QRS identification block is coupled to the processing unit to,

identify a first sub-search region of the filtered ECG signal, wherein the first sub-search region is a region comprising a predefined set of sample points of the filtered ECG signal before a time instance of the R peak;

determine a first peak point in a segment of the filtered ECG signal, within the first sub-search region, wherein the segment comprises a predefined number of consecutive sample points with slope greater than a predefined slope threshold, and wherein the first peak point is one of the predefined number of consecutive sample points in the segment; and

determine a second peak point that corresponds to a peak of the filtered ECG signal in the first sub-search region, wherein the second peak point is determined as a zero crossing point of a slope signal, wherein the slope signal is obtained from three-point derivatives of filtered ECG signal in the first sub-search region;

wherein the Q peak of the Q wave of the QRS complex is identified as one of the first peak point and the second peak point which is closer to the R peak of the R wave of the QRS complex.

12. The QRS identification system as claimed in claim 11, wherein the QRS identification block is coupled to the processing unit to

determine a third peak point of the filtered ECG signal, based on a minimum value of the filtered ECG signal in the first sub-search region, when the first peak point and the second peak point are not determined, wherein the Q peak of the Q wave of the QRS complex is identified as the third peak point.

13. The QRS identification system as claimed in claim 7, wherein the QRS identification block is coupled to the processing unit to

identify a second sub-search region of the filtered ECG signal, wherein the second sub-search region is a region comprising a predefined set of sample points of the filtered ECG signal after a time instance of the R peak;

determine a first peak point in a segment of the filtered ECG signal, within the second sub-search region, wherein the segment comprises a predefined number of consecutive sample points with slope greater than a predefined slope threshold, and wherein the first peak point is one of the predefined number of consecutive sample points in the segment;

determine a second peak point that corresponds to a peak of the filtered ECG signal in the second sub-search region, wherein the second peak point is determined as a zero crossing point of a slope signal, wherein the slope signal is obtained from three-point derivatives of filtered ECG signal in the first sub-search region; and

wherein the S peak of the S wave of the QRS complex is identified as one of the first peak point and the second peak point which is closer to the R peak of the R wave of the QRS complex.

14. The QRS identification system as claimed in claim 13, wherein the QRS identification block is coupled to the processing unit to

determining a third peak point of the filtered ECG signal, based on a minimum value of the filtered ECG signal in the second sub-search region, when the first peak point and the second peak point are not determined, wherein the S peak of the S wave of the QRS complex is identified as the third peak point.

15. The QRS identification system as claimed in claim 7, wherein the first low-pass filter has a cut-off frequency within a range of 15 Hz-40 Hz, and wherein the high-pass filter has a cut-off frequency within a range of 0.5 Hz-5 Hz, and wherein the second low-pass filter has a cut-off frequency within a range of 8 Hz-16 Hz.

16. A cardiovascular disease detection apparatus for detection of cardiovascular diseases using QRS complex in Electrocardiogram (ECG) signals, the cardiovascular disease detection apparatus comprising:

a clock generator;

an ECG recorder, coupled to the clock generator, to record an ECG signal, the ECG signal being indicative of an electrical activity of heart of an individual; and

a QRS identification system, coupled to the clock generator, to

receive the ECG signal captured by the ECG recorder;

filter the ECG signal by passing the ECG signal through at least one of a first low-pass filter and a high-pass filter to obtain a filtered ECG signal;

square the filtered ECG signal to obtain a squared filtered ECG signal;

generate an envelope ECG signal by capturing peaks of the squared filtered ECG signal;

determine a moving average ECG signal based on a moving average of sample points associated with the envelope ECG signal, wherein a sample point is indicative of an amplitude of the envelope ECG signal in a time-amplitude curve at a time instance; and

filter the moving average ECG signal to obtain a processed ECG signal, by passing the moving average ECG signal through a second low-pass filter;

identify a search region in the processed ECG signal, wherein the search region is indicative of a time span between two consecutive peaks of the processed ECG signal;

identify a maximum amplitude peak in a time interval of the filtered ECG signal that corresponds to the time span of the search region of the processed ECG signal, wherein the maximum amplitude peak is an R peak of an R wave of the QRS complex; and

identify a Q peak of a Q wave and an S peak of an S wave of the QRS complex, based on the R peak of the R wave; and

a disease detector for detecting a cardiovascular disease based on the QRS complex of the ECG signal.

17. The cardiovascular disease detection apparatus as claimed in claim 16, wherein the clock generator comprises:

a device to,

receive an external clock signal (CLK); and

generate an inverted clock signal (CLKB), based on the external clock signal (CLK); and

a counter to

receive the inverted clock signal (CLKB) and a chip enabled signal (CE), wherein a high chip enabled signal (CE) is provided to reset the counter; and

generate an internal generated signal (CLK_CHIP), based on the external clock signal, wherein the internal generated clock signal (CLK_CHIP) is a fraction of the external clock signal.