Method and apparatus for detecting heartbeats in an ECG waveform using an activity function and on-time search

• US 5,715,829 A
• Filed: 11/29/1995
• Issued: 02/10/1998
• Est. Priority Date: 11/29/1995
• Status: Expired due to Fees

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Post, William L., Arand, Patricia A.
• Primary Examiner(s): Kamm, William E.
• Assistant Examiner(s): LAYNO, CARL HERNANDZ
• Application Number:
  US08/564,889
• Time in Patent Office: 804 Days
• Field of Search: 128/696, 128/706, 128/700, 128/708
• US Class Current: 600/509
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/7207 of noise induced by motion artifacts

Abstract

A method and apparatus for detecting heartbeats obtains ECG data from a plurality of ECG waveforms, which are in turn obtained from signals received from a plurality of ECG electrodes. QRS detection logic detects heartbeats in the ECG data. This logic calculates an activity function from a subset of the plurality of ECG waveforms determined to be least noisy, and uses this activity function to search for heartbeats. The QRS detection logic allows for true heartbeats to be detected while discarding false "noise" beats.

Claims

1. A method for searching for a heartbeat in a plurality of ECG waveforms comprising the steps of:

determining a preliminary interval time period between preliminary peaks of the ECG waveforms;

selecting a subset of a plurality of ECG waveforms that have optimal noise characteristics;

calculating an activity function from said subset of said plurality of ECG waveforms, wherein said activity function emphasizes heartbeat characteristics in said subset of said plurality of ECG waveforms and minimizes noise characteristics in said subset of said plurality of ECG waveforms;

first searching said activity function for peaks using an on-time search, wherein said on-time search searches for peaks within a predetermined window of time around the lapse of said preliminary interval time period; and

storing any peaks found by said on-time search.

2. The method of claim 1, further comprising the steps of:

choosing one peak from said plurality of peaks as said heartbeat, said one peak being closest to the lapse of said preliminary interval time period; and

storing said heartbeat chosen in said choosing step.

3. The method of claim 1, further comprising the steps of:

determining that no peaks were detected by said on-time search during said first searching step; and

second searching said activity function using a modified on-time search.

4. The method of claim 3, further comprising the steps of:

finding a plurality of peaks, responsive to said second searching step;

choosing one peak from said plurality of peaks as said heartbeat, said one peak being closest to the lapse of said preliminary interval time period; and

storing said heartbeat chosen in said choosing step.

5. The method of claim 3, further comprising the steps of:

determining that no peaks were detected by said on-time search during said first searching step or by said modified on-time search during said second searching step;

third searching said activity function using a late beat search;

finding a plurality of peaks, responsive to said third searching step;

choosing one peak from said plurality of peaks as said heartbeat, said one peak being the first of said plurality of peaks; and

storing said heartbeat chosen in said choosing step.

6. The method of claim 1, wherein said selecting step selects the three ECG waveforms with the least noise.

7. The method of claim 1, wherein said calculating an activity function step further comprises the steps of:

decimating data contained on each of said subset of said ECG waveforms, thereby creating decimated data;

filtering said decimated data, thereby creating filtered decimated data;

taking the absolute value of the first difference of said filtered decimated data, thereby creating absolute difference data;

summing said absolute difference data for each of said subset of ECG waveforms, thereby creating summed data; and

thresholding and smoothing said summed data to create said activity function.

8. The method of claim 1, wherein said first searching step further comprises the steps of:

selecting an interval range greater than said preliminary interval time period;

setting a first threshold at a first threshold percentage of an average peak value;

linearly decreasing said first threshold to a second threshold having a second threshold percentage of said average peak value; and

noting any peaks that are above the linearly decreasing threshold between said first threshold and said second threshold.

9. The method of claim 3, wherein said second searching step further comprises the steps of:

selecting an interval range greater than said preliminary interval time period;

setting a third threshold at a third threshold percentage of an average peak value, said third threshold being constant over said interval range; and

noting any peaks that are above said third threshold.

10. The method of claim 5, wherein said third searching step further comprises the steps of:

selecting an interval range of three preliminary interval time periods of said subset of ECG waveforms;

setting a first threshold at a first threshold percentage of an average peak value;

linearly decreasing said first threshold to a Second threshold having a second threshold percentage of said average peak value; and

noting any peaks that are between said first threshold and said second threshold.

11. A medical device for searching for a heartbeat in a plurality of ECG waveforms, comprising:

means for determining a preliminary interval time period between preliminary peaks of the ECG waveforms;

means for selecting a subset of a plurality of ECG waveforms that have optimal noise characteristics;

means for calculating an activity function from said subset of said plurality of ECG waveforms, wherein said activity function emphasizes heartbeat characteristics in said subset of said plurality of EGG waveforms and minimizes noise characteristics in said subset of said plurality of ECG waveforms;

first searching means for searching said activity function for peaks using an on- time search, wherein said on-time search searches for peaks within a predetermined window of time around the lapse of said preliminary interval time period; and

means for storing any peaks found by said on-time search.

12. The medical device of claim 11, further comprising:

means for detecting a plurality of peaks, responsive to said first searching means;

means for choosing one peak from said plurality of peaks as said heartbeat, said one peak being closest to the lapse of said preliminary interval time period; and

means for storing said heartbeat chosen by said choosing means.

13. The medical device of claim 11, further comprising:

means for determining that no peaks were detected by said first searching means;

second searching means for searching said activity function using a modified on-time search.

14. The medical device of claim 13, further comprising:

means for finding a plurality of peaks, responsive to said second searching means;

means for choosing one peak from said plurality of peaks as said heartbeat, said one peak being closest to the lapse of said preliminary interval time period; and

means for storing said heartbeat chosen in said choosing step.

15. The medical device of claim 13, further comprising:

means for determining that no peaks were detected by said first searching means or by said second searching means;

third searching means for searching said activity function using a late beat search;

means for finding a plurality of peaks, responsive to said third searching means;

means for choosing one peak from said plurality of peaks as said heartbeat, said one peak being the first of said plurality of peaks; and

means for storing said heartbeat chosen by said choosing means.

16. The medical device of claim 11, wherein said subset of plurality of ECG waveforms selected by said selecting means is the three ECG waveforms with the least noise.

17. The medical device of claim 11, wherein said means for calculating an activity function further comprises:

means for decimating data contained on each of said subset of said ECG waveforms, thereby creating decimated data;

means for filtering said decimated data, thereby creating filtered decimated data;

means for taking the absolute value of the first difference of said filtered decimated data, thereby creating absolute difference data;

means for summing said absolute difference data for each of said subset of ECG waveforms, thereby creating summed data; and

means for thresholding and smoothing said summed data to create said activity function.

18. The medical device of claim 11, wherein said first searching means further comprises:

means for selecting an interval range greater than

means for setting a first threshold at a first threshold percentage of an average peak value;

means for linearly decreasing said first threshold to a second threshold having a second threshold percentage of said average peak value; and

means for noting any peaks that are above the linearly decreasing threshold between said first threshold and said second threshold.

19. The medical device of claim 13, wherein said second searching means further comprises:

means for selecting an interval range greater than said preliminary interval time period;

means for setting a third threshold at a third threshold percentage of an average peak value, said third threshold being constant over said interval range; and

means for noting any peaks that are above said third threshold.

20. The medical device of claim 15, wherein said third searching means further comprises:

means for selecting an interval range of three preliminary interval time periods of said subset of ECG waveforms;

means for setting a first threshold at a first threshold percentage of an average peak value;

means for linearly decreasing said first threshold to a second threshold having a second threshold percentage of said average peak value; and

means for noting any peaks that are between said first threshold and said second threshold.