Method and apparatus for creating a representative heartbeat from an ECG waveform

• US 5,613,496 A
• Filed: 11/29/1995
• Issued: 03/25/1997
• Est. Priority Date: 11/29/1995
• Status: Expired due to Term

• Current Assignee: Koninklijke Philips N.V.
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Forbes, Alfred D., Arand, Patricia A., Post, William L.
• Primary Examiner(s): Kamm, William E.
• Application Number:
  US08/565,504
• Time in Patent Office: 482 Days
• Field of Search: 346/413.06, 128/696, 128/699, 128/202, 128/703, 128/708
• US Class Current: 600/509
• CPC Class Codes: A61B 5/0245 by using sensing means generating electric signals, i.e. ECG signals
• CPC Class Codes: A61B 5/35 by template matching
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

A method and apparatus for creating a representative heartbeat 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. Classification logic classifies heartbeats into categories based on shape and/or timing. Alignment logic aligns the heartbeats. This logic slides the heartbeats occurring at various times across a stationary alignment template to calculate when the heartbeats are best aligned, and performs adjustments to reduce the effects of noise or jitter on the different ECG waveforms. Representative heartbeat creation logic creates a representative heartbeat from the aligned heartbeats. This logic time slices through the aligned heartbeats, discarding a percentage of the smallest and largest magnitudes of the aligned heartbeats at each instance of time, and averaging the remaining magnitudes to produce a representative heartbeat. This trimmed averaging technique results in a high quality representative beat, since samples from noise and misclassified beats are discarded. The representative heartbeat is displayed, either alone or with head rate and/or other measurement information, to the cardiologist or medical professional for diagnosis of the condition of the patient'"'"'s head, such as a diagnosis of coronary artery disease, based on finding a depressed ST segment in the representative heartbeat of a patient undergoing a stress or exercise test.

Claims

1. A method of creating information about a representative heartbeat from an ECG waveform, said method comprising the steps of:

detecting a plurality of heartbeats in said ECG waveform;

classifying said plurality of heartbeats as dominant heartbeats and non-dominant heartbeats;

aligning said dominant heartbeats to create aligned dominant heartbeats;

creating a representative heartbeat from said plurality of aligned dominant heartbeats, said creating step further comprising the steps of;

time slicing through said plurality of aligned dominant heartbeats;

responsive to said time slicing step, determining the magnitudes of each of said plurality of aligned dominant heartbeats at a first slice of time, thereby creating first magnitudes;

sorting the first magnitudes from smallest first magnitude to largest first magnitude, thereby creating sorted first magnitudes;

first discarding a first percentage of the smallest first magnitudes;

second discarding a second percentage of the largest first magnitudes, wherein said first magnitudes remaining after said first discarding step and said second discarding step are remaining first magnitudes;

averaging said remaining first magnitudes, thereby creating an average first magnitude; and

storing said average first magnitude.

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

responsive to said time slicing step, storing the magnitudes of each of said plurality of aligned dominant heartbeats at a second slice of time, thereby creating second magnitudes;

sorting the second magnitudes from smallest second magnitude to largest second magnitude, thereby creating sorted second magnitudes;

third discarding said first percentage of the smallest second magnitudes;

fourth discarding said second percentage of the largest second magnitudes, wherein said second magnitudes remaining after said third discarding step and said fourth discarding step are remaining second magnitudes;

averaging said remaining second magnitudes, thereby creating an average second magnitude; and

storing said average second magnitude.

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

responsive to said time slicing step, storing the magnitudes of each of said plurality of aligned dominant heartbeats at 3rd-N slices of time, thereby creating 3rd-N magnitudes;

sorting the 3rd-N magnitudes from smallest 3rd-N magnitudes to largest 3rd-N magnitudes, thereby creating sorted 3rd-N magnitudes;

fifth discarding said first percentage of each of the smallest 3rd-N magnitudes;

sixth discarding said second percentage of each of the largest 3rd-N magnitudes, wherein said 3rd-N magnitudes remaining after said fifth discarding step and said sixth discarding step are remaining 3rd-N magnitudes;

averaging said remaining 3rd-N magnitudes, thereby creating average 3rd-N magnitudes; and

storing said average 3rd-N magnitudes.

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

generating said representative heartbeat from said first average magnitude, said second average magnitude, and said 3rd-N average magnitudes; and

displaying said representative heartbeat.

5. The method of claim 4, wherein said first discarding step discards 33% of the smallest magnitudes at said first, second, and 3rd-N slices of time.

6. The method of claim 5, wherein said second discarding step discards 33% of the largest magnitudes at said first, second, and 3rd-N slices of time.

7. The method of claim 6, further comprising the step of:

creating a plurality of representative heartbeats corresponding to a plurality of ECG waveforms.

8. The method of claim 1, further comprising the step of:

obtaining measurements for said representative heartbeat.

9. The method of claim 8, further comprising the step of:

displaying said representative heartbeat and said measurements for said representative heartbeat.

10. The method of claim 9, further comprising the steps of:

calculating a rate of said plurality of heartbeats; and

displaying said representative heartbeat and said rate of said plurality of heartbeats.

11. The method of claim 1, further comprising the step of:

displaying said representative heartbeat.

12. The method of claim 1, wherein said step of aligning said dominant heartbeats to create aligned dominant heartbeats further comprises the steps of:

creating an alignment template;

sliding a first heartbeat from said plurality of dominant heartbeats across said alignment template;

while performing said sliding step, computing an area difference between said first heartbeat and said alignment template;

determining a first position of said first heartbeat when said area difference is minimum; and

storing said first position of said first heartbeat as a best aligned position.

13. The method of claim 12, further comprising the step of:

repeating the creating, sliding, computing, determining and storing steps for the other heartbeats in said plurality of dominant heartbeats, thereby creating said aligned heartbeats.

14. The method of claim 13, further comprising the steps of:

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

selecting said plurality of dominant heartbeats from one of said subset of said plurality of ECG waveforms that have optimal noise characteristics.

15. The method of claim 13, further comprising the steps of:

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

determining a plurality of median alignment times for said subset of said plurality of ECG waveforms; and

correcting said aligned heartbeats with said plurality of median alignment times.

16. A medical device for creating information about a representative heartbeat from an ECG waveform, said medical device comprising:

detection logic for detecting a plurality of heartbeats in said ECG waveform;

classification logic for classifying said plurality of heartbeats as dominant heartbeats and non-dominant heartbeats;

alignment logic for aligning said dominant heartbeats to create aligned dominant heartbeats;

representative heartbeat logic for creating a representative heartbeat from said plurality of aligned dominant heartbeats, said representative heartbeat logic further comprising;

means for time slicing through said plurality of aligned dominant heartbeats;

means, responsive to said time slicing means, for determining the magnitudes of each of said plurality of aligned dominant heartbeats at a first slice of time, thereby creating first magnitudes;

means for sorting the first magnitudes from smallest first magnitude to largest first magnitude, thereby creating sorted first magnitudes;

first discarding means for discarding a first percentage of the smallest first magnitudes;

second discarding means for discarding a second percentage of the largest first magnitudes, wherein said first magnitudes remaining after said first discarding means and said second discarding means are remaining first magnitudes;

means for averaging said remaining first magnitudes, thereby creating an average first magnitude; and

means for storing said average first magnitude.

17. The medical device of claim 16, further comprising:

responsive to said time slicing means, means for storing the magnitudes of each of said plurality of aligned dominant heartbeats at a second slice of time, thereby creating second magnitudes;

means for sorting the second magnitudes from smallest second magnitude to largest second magnitude, thereby creating sorted second magnitudes;

third discarding means for discarding said first percentage of the smallest second magnitudes;

fourth discarding means for discarding said second percentage of the largest second magnitudes, wherein said second magnitudes remaining after said third discarding means and said fourth discarding means are remaining second magnitudes;

means for averaging said remaining second magnitudes, thereby creating an average second magnitude; and

means for storing said average second magnitude.

18. The medical device of claim 17, further comprising:

responsive to said time slicing means, means for storing the magnitudes of each of said plurality of aligned dominant heartbeats at 3rd-N slices of time, thereby creating 3rd-N magnitudes;

means for sorting the 3rd-N magnitudes from smallest 3rd-N magnitudes to largest 3rd-N magnitudes, thereby creating sorted 3rd-N magnitudes;

fifth discarding means for discarding said first percentage of each of the smallest 3rd-N magnitudes;

sixth discarding means for discarding said second percentage of each of the largest 3rd-N magnitudes, wherein said 3rd-N magnitudes remaining after said fifth discarding means and said sixth discarding means are remaining 3rd-N magnitudes;

averaging said remaining 3rd-N magnitudes, thereby creating average 3rd-N magnitudes; and

storing said average 3rd-N magnitudes.

19. The medical device of claim 18, further comprising:

means for generating said representative heartbeat from said first average magnitude, said second average magnitude, and said 3rd-N average magnitudes; and

a display for displaying said representative heartbeat.

20. The medical device of claim 18, further comprising:

means for generating said representative heartbeat from said first average magnitude, said second average magnitude, and said 3rd-N average magnitudes; and

a printer for printing said representative heartbeat.

21. The medical device of claim 18, wherein said first discarding means discards 33% of the smallest magnitudes at said first, second, and 3rd-N slices of time.

22. The medical device of claim 21, wherein said second discarding means discards 33% of the largest magnitudes at said first, second, and 3rd-N slices of time.

23. The medical device of claim 22, further comprising:

means for creating a plurality of representative heartbeats corresponding to a plurality of ECG waveforms.

24. The medical device of claim 18, further comprising:

measurement logic for obtaining measurements for said representative heartbeat.

25. The medical device of claim 24, further comprising:

a display for displaying said representative heartbeat.

26. The medical device of claim 24, further comprising:

a display for displaying said representative heartbeat and said measurements for said representative heartbeat.

27. The medical device of claim 24, further comprising:

calculation logic for calculating a rate of said plurality of heartbeats; and

a display for displaying said representative heartbeat and said rate of said plurality of heartbeats.

28. The medical device of claim 16, wherein said means for aligning said dominant heartbeats to create aligned dominant heartbeats further comprises:

means for creating an alignment template;

means for sliding a first heartbeat from said plurality of dominant heartbeats across said alignment template;

while performing said sliding means, means for computing an area difference between said first heartbeat and said alignment template;

means for determining a first position of said first heartbeat when said area difference is minimum; and

means for storing said first position of said first heartbeat as a best aligned position.

29. The medical device of claim 28, further comprising:

means for repeating the creating, sliding, computing, determining and storing means for the other heartbeats in said plurality of dominant heartbeats, thereby creating said aligned heartbeats.

30. The medical device of claim 29, further comprising:

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

means for selecting said plurality of dominant heartbeats from one of said subset of said plurality of ECG waveforms that have optimal noise characteristics.

31. The medical device of claim 29, further comprising:

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

means for determining a plurality of median alignment times for said subset of said plurality of ECG waveforms; and

means for correcting said aligned heartbeats with said plurality of median alignment times.