Intramyocardial anomalous activity detection by subtracting modeled respiratory effect

• US 6,132,381 A
• Filed: 08/14/1997
• Issued: 10/17/2000
• Est. Priority Date: 08/14/1997
• Status: Expired due to Fees

• Current Assignee: Agilent Technologies, Inc.
• Original Assignee: Agilent Technologies, Inc.
• Inventors: Forbes, A. Dean, Helfenbein, Eric D.
• Primary Examiner(s): O'Connor, Cary
• Assistant Examiner(s): Natnithithadha, Navin
• Application Number:
  US08/911,089
• Time in Patent Office: 1,160 Days
• Field of Search: 600/483, 600/508, 600/509, 600/513, 600/515, 600/519, 600/520, 600/521
• US Class Current: 600/483
• CPC Class Codes: A61B 5/0535 Impedance plethysmography for measuring blood flow A61B5/0295
• CPC Class Codes: A61B 5/349 Detecting specific parameters of the electrocardiograph cycle
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal

Abstract

A technique of detecting repeating patterns of anomalous intramyocardial Wenckebach electrical activity in a patient is provided. The method of the technique deals with determining the presence or absence of such intramyocardial anomaly by processing signals that are electrical potential signals which include a cardiac component and a respiratory component. The method includes estimating the respiratory component (preferably by a regression curve-fitting technique based on data of the electrical potential signals and data of respiratory signals), finding the difference between the electrical potential and the estimated respiratory component and based on this difference inferring the presence of absence of the intramyocardial anomaly. Preferably, a generalized inverse of a matrix representing Wenckebach basis functions reflecting Wenckebach phases and modes is used to act on a vector representing the difference between the electrical potential and the estimated respiratory component to determine whether any repeating intramyocardial electrical anomaly is present.

Claims

1. An apparatus for detecting repeating patterns of anomalous Intramyocardial electrical activity in the heart of a patient, comprising:

(a) means for obtaining data on an electrical potential that includes a cardiac component and a respiratory component;

(b) means for obtaining respiratory data from the body; and

(c) processor electrically associated with the means for obtaining electrical potential data and with the means for obtaining respiratory data, the processor regressing the electrical potential data on the respiratory data to estimate the respiratory component in the electrical potential, said processor aligning the electrical potential data by heart beats and statistically modeling the respiratory data with the electrical potential data in alignment across multiple heart beats the processor also determining whether anomalous Intramyocardial activity is present by finding the difference between the electrical potential data and the estimated respiratory component.

2. The apparatus according to claim 1 wherein the processor determines whether intramyocardial anomaly is present, including:

(a) in the processor storing data of a nonsingular Wenckebach matrix W corresponding to Wenckebach patterns of cardiac activity; and

(b) the processor calculating a vector of Wenckebach basis function strengths x based on the nonsingular Wenckebach matrix W and the electrical potential data, the Wenckebach basis function strengths indicating the presence or absence of periodically repeated anomalies in said difference.

3. The apparatus according to claim 2 wherein the processor calculates the Wenckebach basis fuinction strengths via a relationship describing the electrical potential data as including Wenckebach activity being additive to respiratory interference, in the relationship the nonsingular Wenckebach matrix W acts on the vector of Wenckebach basis function strengths x to describe said difference between the electrical potential data and the estimated respiratory data, in the relationship the Wenckebach basis function strengths are expressed in terms of said difference between the electrical potential data and the estimated respiratory data and in terms of a function of the nonsingular Wenckebach matrix W, the Wenckebach basis function strengths allowing determination of presence or absence of effect in the electrical potential data caused by repeating patterns of anomalous intramyocardial Wenckebach activity.

4. The apparatus according to claim 2 wherein the processor orthogonalizes a singular matrix representing Wenckebach patterns to remove singularity therefrom to obtain the nonsingular Wenckebach matrix W.

5. The apparatus according to claim 2 wherein the means for obtaining respiratory data comprises electrodes for sensing body surface potential (BSP) signal and deriving both the electrical potential and the respiratory data from the BSP signal.

6. The apparaus according to claim 2, wherein the processor calculates the vector of Wenckebach basis function strengths x by a relationship in which the nonsingular Wenckebach matrix W, a vector b of cardiac electrical potential data obtained from the electrical potential data after reducing respiratory effect by regressing, and the vector of Wenckebach basis function strengths x are related by x=W^I b, where the W^I represents the generalized inverse of the matrix W.

7. The apparatus according to claim 6, wherein the processor tests the respiratory data to determine the presence or absence of periodically repeated Wenckebach-like activity in the respiratory data to assess the risk of mistakenly inferring Wenckebach anomaly in said difference, wherein the processor for said testing performs calculation, including:

obtaining respiratory data points by extracting from the respiratory data according to heart beat times in the electrical potential data, calculating values of Wenckebach basis function strengths x based on the extracted respiratory data points and the nonsingular Wenckebach matrix W, and calculating a generalized likelihood ratio (GLR), wherein small values in the GLR will indicate the absence of periodically repeating Wenckebach-like activity in the respiratory data.

8. The apparatus according to claim 1, wherein the means for obtaining electrical potential data comprises electrodes for detecting body surface electrical potentials (BSP) from the body of the patient.

9. The apparatus according to claim 8, further comprising one or more filters to filter the BSP to remove noise, which includes non-respiratory physiologically generated electrical noise, to obtain the electrical potential data used for regressing.

10. The apparatus according to claim 8, wherein the processor has a digital filter embodied by a computer program having code means for filtering data from the BSP signal.

24. The apparatus according to claim 1 wherein the electrical potential data comprises data in time-series and the processor arranges by alignment of heart beats such that each heart beat has points in time in the time-series corresponding to points in time in the time-series of other heart beats and the respiratory data are modeled with data sets from the electrical potential data, each of said sets spanning multiple heart beats at a point in time on the aligned heart beats, each set having a different point in time.

25. The apparatus according to claim 24 wherein the means for obtaining respiratory data obtains two or more sets of respiratory data, each set spanning substantially the same time period and wherein the processor models each set of respiratory data with the electrical potential data to estimate the respiratory component of the electrical potential data.

11. A method for detecting repeating patterns of anomalous Intramyocardial activity, comprising:

(a) providing electrical potential data that include a cardiac component and a respiratory component;

(b) regressing the electrical potential data on respiratory data to estimate the respiratory component, by aligning the electrical potential data by heart beats and statistically modeling the respiratory data with the electrical potential data in alignment across multiple heart beats; and

(c) determining the presence or absence of intramyocardial anomaly by finding the difference between the electrical potential data and the estimated respiratory component and detecting periodically repeating intramyocardial anomaly in the difference.

12. The method according to claim 11 wherein the step of determining intramyocardial anomaly includes the steps of:

(a) generating a nonsingular Wenckebach matrix W describing patterns Wenckebach activity; and

(b) calculating a vector of Wenckebach basis finction strengths x based on the nonsingular Wenckebach matrix W and the regressed electrical potential data to determine the presence or absense of repeating patterns of anomalous intramyocardial activity, the vector of Wenckebach basis function strengths x indicating the intensity of Wenckebach activity.

13. The method according to claim 12 wherein the Wenckebach basis function strengths x are calculated via a relationship that describes the electrical potential data as including data due to Wenckebach activity being additive to data due to respiratory interference, in the relationship the nonsingular Wenckebach matrix W acts on the vector of Wenckebach basis function strengths x to describe said difference between the electrical potential data and the estimated respiratory interference, in the relationship expressing the Wenckebach basis function strengths in terms of a function of the nonsingular Wenckebach matrix W and said difference between the electrical potential data and the estimated respiratory interference.

14. The method according to claim 12, wherein the nonsingular Wenckebach matrix W is obtained by orthogonalizing a singular Wenckebach matrix describing phases of modes of Wenckebach activity.

15. The method according to claim 12, wherein the nonsingular Wenckebach matrix W, a vector b of values of the difference between the electrical potential data and the estimated respiratory component, and the Wenckebach basis function strengths in x are related by x=W^I b , where W^I represents generalized inverse of the matrix W.

16. The method according to claim 15 further comprising determining a generalized likelihood ratio (GLR), said GLR being the power in b due to Wenckebach activity divided by the power in b not due to Wenckebach activity, the value of said GLR being indicative of whether Wenckebach activity has been reliably detected and hence whether x is a reliable estimate of the amplitude of Wenckebach activity.

17. The method according to claim 12, wherein the estimated respiratory component is estimated using an equation having an indicator function indicating respiratory phases including inspiration, expiration, and rest of the corresponding respiratory data.

18. The method according to claim 12, further comprising performing singular value decomposition on a larger Wenckebach matrix W to obtain the nonsingular Wenckebach matrix W.

19. The method according to claim 12, wherein the step of regressing includes regressing the BSP data on more than one set of respiratory data obtained from different sites on the patient.

20. The method according to claim 11, further comprising filtering BSP signal detected from the body of the patient to remove electromagnetic interference and non-respiratory physiologically generated electrical interference.

21. The method according to claim 20, further comprising filtering BSP signal to remove at least one physiologically generated electrical interference selected from the group consisting of regular electrocardiographic signal, muscle tremor noise, and electrogastric signal.

22. The method according to claim 11 further comprising:

(a) measuring body surface potential (BSP) signal from the body of the patient and determining the electrical potential from the BSP signal;

(b) filtering the BSP signal to remove noise including a non-respiratory physiologically generated electrical noise to obtain the electrical potential data used for regressing; and

(c) obtaining the respiratory data.

23. The method according to claim 12 further comprising a step for testing the respiratory data to determine the presence or absence of periodic activity to indicate the risk of mistakenly inferring the presence of Wenckebach anomalies in said difference, the step of testing including obtaining data points by extracting from the respiratory data according to time of heart beats in the electrical potential data, acting on the extracted data points with the nonsingular Wenckebach matrix W to obtain the Wenckebach basis function strengths x, the step of testing further comprising calculating a generalized likelihood ratio (GLR) based on W and x wherein small values in the GLR will indicate the absence of Wenckebach-like interference in the respiratory data.

26. The method according to claim 11 wherein the electrical potential data comprises data in time-series and the regression-analyzing step comprises arranging the electrical potential data by alignment of heart beats such that each heart beat has points in time in the time-series corresponding to points in time in the time-series of other heart beats and the regression-analyzing step further comprises modeling the respiratory data with data sets from the electrical potential data, wherein each of said sets spans multiple heart beats at a point in time on the aligned heart beats, each set having a different point in time.

27. An apparatus for detecting repeating patterns of anomalous intramyocardial electrical activity in the heart of a patient, comprising:

(a) means for obtaining data on an electrical potential that includes a cardiac component and a respiratory component;

(b) means for obtaining respiratory data from the body; and

(c) processor electrically associated with the means for obtaining electrical potential data and with the means for obtaining respiratory data, the processor regressing the electrical potential data on the respiratory data to estimate the respiratory component in the electrical potential, the processor also determining whether anomalous intramyocardial activity is present by finding the difference between the electrical potential data and the estimated respiratory component, wherein the processor calculates the vector of Wenckebach basis function strengths x by a relationship in which the nonsingular Wenckebach matrix W, a vector b of cardiac electrical potential data obtained from the electrical potential data, and the vector of Wenckebach basis function strengths x are related by x=W^I b , where the W^I represents the generalized inverse of the matrix W.

28. An article of manufacture comprising a program storage medium, tangibly embodying a program code means readble by a computer to cause the computer to analyze for repeating patterns of anomalous intramyocardial activity of a patient, the computer readable program code means in the article of manufacture including:

(a) code means for regressing electrical potential data of the patient on respiratory data of the patient to estimate a respiratory component in the electrical potential data, the electrical potential data including a cardiac component and a respiratory component;

(b) code means for generating a vector b by determining the difference between the electrical potential data and the estimated respiratory component;

(c) code means for generating a nonsingular Wenckebach matrix W describing patterns of Wenckebach activity; and

(d) code means for determining a vector of Wenckebach basis function strengths x based on the nonsingular Wenckebach matrix W and the vector b of said difference to determine the presence or absence of repeating patterns of Wenckebach activity, wherein the vector of Wenckebach basis finction strengths x indicating the intensity of Wenckebach activity.