Heart rate detection utilizing autoregressive analysis

• US 4,616,659 A
• Filed: 05/06/1985
• Issued: 10/14/1986
• Est. Priority Date: 05/06/1985
• Status: Expired due to Term

• Current Assignee: Bell Telephone Laboratories, Inc. (Nokia Corporation)
• Original Assignee: AT&T, Inc.
• Inventors: Zimmerman, Gustavus H. III, Prezas, Dimitrios P.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US06/730,486
• Time in Patent Office: 526 Days
• Field of Search: 128/696, 128/706, 128/708, 128/731, 364/415, 364/417, 364/724
• US Class Current: 600/519
• CPC Class Codes: A61B 5/0245 by using sensing means generating electric signals, i.e. ECG signals

Abstract

A system for determining the heart rate in response to an electrocardiogram (EKG) signal that has been filtered by performing an autoregressive analysis, low-pass filtered by performing a digital filtering function, and the rate determined by utilizing a digital peak detection technique. In a heart rate detection system, the EKG signal from an EKG instrument is first digitized and then the colored noise component is removed from the digitized EKG samples by performing a linear predictive coding filtering function on the EKG samples. After the colored noise has been filtered from the digitized EKG samples, a low-pass filtering function is performed to remove high-frequency white noise by utilizing a digital finite impulse response filter. Finally, the heart rate is determined from the previously filtered EKG samples by a peak detection technique that performs a recursive algorithm to determine both the amplitude of the remaining peaks in the EKG samples and the period between these peaks.

Claims

1. A detection system for determining the cardiac cyclic rate by detecting a predetermined component of the cardiac cycle, comprising:

means for obtaining cardiac signals defining a plurality of cardiac cycles;

means for performing an autoregressive analysis filtering of said cardiac signals to remove color noise;

means for executing a low-pass filtering of the autoregressive filtered signals to remove white noise;

means for peak detecting the low-pass filtered signals to detect said predetermined component thereby determining said cyclic rate; and

means for displaying said cyclic rate in response to the detection of each predetermined component.

2. The system of claim 1 wherein said predetermined component is the R-peak of the QRS complex of said cardiac cycle.

3. The system of claim 1 wherein said performing means comprises:

means from calculating linear filter coefficients for said cardiac signals modeling the source producing said color noise;

means responsive to said linear filter coefficients and said cardiac signals for computing a plurality of residual signals representing the difference between the model defined by said linear filter coefficients and each of said cardiac signals with said residual signals representing the results of said autoregressive analysis filtering.

4. The system of claim 3 wherein said calculating means comprises:

means for dividing said cardiac signals into a plurality of channels each having a predefined number of cardiac signals;

means for computing reflection coefficients for each of said channels; and

means for stepping-up said reflection coefficients to obtain said linear filter coefficients for each of said channels.

5. The system of claim 4 wherein said obtaining means comprises means responsive to patient electrocardiogram signals for generating said cardiac signals.

6. The system of claim 5 wherein said executing means comprises:

means for storing a plurality of filter coefficients defining the low-pass filter;

means for storing a set of said autoregressive filtered signals;

means for multiplying said filter coefficients times said stored set of autoregressive filtered signals;

means for summing the results of each of said multiplication; and

means for generating one of said low-pass filtered signals from said summed results.

7. The system of claim 6 wherein said executing further comprises means for discarding one of said stored set of said autoregressive filtered signals;

means for storing another autoregressive filtered signal in said stored set of said autoregressive filtered signals; and

means for activating said multiplying, summing, and generating means to produce another one of said low-pass filtered signals.

8. The system of claim 7 wherein said peak detecting means comprises:

means for locating the main R-peak of maximum amplitude within said low-pass filtered signals;

means for locating peaks of lesser amplitude than that of said maximum amplitude spaced by not less than a minimum distance based on the highest expected heart rate from said peak of maximum amplitude and from each other within said low-pass filtered signals;

means for measuring one-by-one the distances between adjacent peaks using as a reference the location of said peak of maximum amplitude;

means for testing for periodicity by comparing successive distance measurements for substantial equality and rejecting peaks which are not peridically related to the peak of maximum amplitude; and

means for determining said heart rate from the periodicity.

9. The system of claim 8 wherein said locating means for peaks of lesser amplitude further comprises means for ordering peaks of acceptable amplitude by time of occurrence within said low-pass filtered signals rather than by amplitude.

10. The system of claim 8 wherein said distance-measuring means comprises means for measuring first the separation distance of a given peak from the main peak and then comparing the location of given peaks with assumed locations separated from the main peak by multiples of said separating distance, failure to match within a preselected tolerance the location of actual peaks with said assumed locations being an indication of lack of periodicity of such other peak with said main peak.

11. A system for determining heart beat rate by detecting the R-wave of the QRS complex of a patient'"'"'s electrocardiogram signal, comprising:

means for obtaining said patient'"'"'s electrocardiogram signals;

means for digitizing said electrocardiogram signals;

means for dividing said digitized electrocardiogram signals into a plurality of channels each having a predefined number of electrocardiogram signals;

means for computing reflection coefficients for each of said channels from said digitized electrocardiogram signals modeling noise contained within said electrocardiogram signals;

means for computing the residual representing the difference between said model and said electrocardiogram signals from said residual thereby representing the linear filtered output signals of said model;

means for storing a plurality of filtered coefficients defining a low-pass filter;

means for computing output signals of said low-pass filter in response to said linear filtered signals and said stored filtered coefficients;

means for locating the main R-peak of maximum amplitude within the low-pass filtered signals;

means for locating peaks of lesser amplitude than that of said peak of maximum amplitude by not less than a minimum distance based on the highest expected heart rate from said peak of maximum amplitude and from each other within said low-pass filtered signals;

means for measuring one-by-one the distances between adjacent peaks using as a reference the location of said peak of said maximum amplitude;

means for testing for periodicity by comparing successive distance measurements for substantial equality and rejecting peaks which are not periodically related to the peak of maximum amplitude; and

means for determining said heart rate from the periodicity.

12. A monitoring method for determining the cardiac cycle rate by detecting a predetermined component of the cardiac cycle, comprising:

reading the electrocardiogram signals of a patient for a plurality of cardiac cycles;

performing an autoregressive analysis filtering of said electrocardiogram signals to remove color noise;

removing white noise by low-pass filtering said autoregressive filtered signals;

locating the main peak of maximum amplitude within the low-pass filtered signals;

locating the main peak of maximum amplitude within the low-pass filtered signals;

locating peaks of lesser amplitude than said peak of maximum amplitude by not less than a minimum distance based on the highest expected heart rate from said peak of maximum amplitude and from each other within said low-pass filtered signals;

measuring the distance between adjacent peaks using as a reference the location of said peak of maximum amplitude;

testing for periodicity by comparing successive distance measurements for substantial equality and rejecting peaks which are not periodically related to the peak of maximum amplitude; and

determining said cardiac cycle rate from the periodicity.

13. The method of claim 12 wherein said performing step comprises regressively executing the following steps:

computing a forward error signal for each of said electrocardiogram signals representing the difference between each electrocardiogram signals and succeeding electrocardiogram signals;

computing a backward error signal representing the difference between each of said electrocardiogram signals and preceding electrocardiogram signals;

recursively computing reflection coefficients from said forward and backward error signals; and

setting said autoregressive filtered signals equal to said forward error signals of all of said electrocardiogram signals.

14. A method for measuring the heart rate of a patient in response to an electrocardiogram signal comprising the steps of:

digitizing samples of said electrocardiogram signal;

locating the peak of maximum amplitude in said digitized samples;

locating peaks of lesser amplitude spaced a distance apart greater than the highest expected heart rate and having an amplitude greater than a predefined percentage of said peak of largest amplitude;

measuring the distance between adjacent peaks;

testing for periodicity by comparing successive distance measurements for substantial equality; and

displaying the periodicity as an indication of said heart rate.