Arrhythmia detector using delta modulated turning point morphology of the ECG wave

• US 5,301,677 A
• Filed: 02/06/1992
• Issued: 04/12/1994
• Est. Priority Date: 02/06/1992
• Status: Expired due to Term

• Current Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Original Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Inventors: Hsung, Jean-Cheui
• Primary Examiner(s): Cohen, Lee S.
• Assistant Examiner(s): Jastrzab, Jeffrey R.
• Application Number:
  US07/833,843
• Time in Patent Office: 796 Days
• Field of Search: 128/702-705, 128/419 D, 607/5
• US Class Current: 600/518
• CPC Class Codes: A61B 5/361 Detecting fibrillation
• CPC Class Codes: A61B 5/363 Detecting tachycardia or bradycardia
• CPC Class Codes: A61B 5/364 Detecting abnormal ECG interval, e.g. extrasystoles, ectopic heartbeats
• CPC Class Codes: A61N 1/3702 Physiological parameters A61N1/365 takes precedence; evoked response A61N1/371

Abstract

An arrhythmia detector includes a delta modulator which digitizes the ECG signal and produces a serial digital signal that represents the input ECG signal. A microprocessor classifies the signal as having no slope, positive slope, or negative slope. A finite state machine model in the microprocessor uses the positive slope, negative slope, or no slope information to determine the number of times an isoelectric line segment is present in the ECG signal. By accumulating the line-counts or number of line segments over a two-second period, a determination can be made whether the ECG signal has been in isoelectric line less than a threshold. For two out of three 2 second periods, if the percentage of isoelectric time is less than the ventricular fibrillation threshold then a ventricular fibrillation signal can be generated. For ventricular tachycardia verses supraventricular tachycardia discrimination, a second threshold which relates to the average heart rate observed in a linear formula LC(TH)=b-a*(rate)/30 can be used. If the isoelectric line-count is less than the LC(TH), then VT is declared. The regression coefficient a and b can be found by measuring the line-count and rate pairs in 2 second periods during normal sinus rhythm in periodic basis, for example, 10 data pairs every hour. The delta modulator uses a comparator which receives the ECG signal and a tracking signal to output a series of binary digits. A digital-to-analog converter generates the tracking signal.

Claims

1. An arrhythmia detector for analyzing ECG signals obtained from a heart, said detector comprising:

means for delta modulating an input ECG signal for obtaining a digital representation of said input ECG signal from a heart;

means for periodically classifying said digital signal as representing no slope, positive slope, or negative slope segments of said input ECG signal and respectively generating a no slope, positive slope, or negative slope signal;

means for determining, based upon a sequence over a predetermined period of time of no slope, positive slope, and negative slope signals, the amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time; and

means for issuing a fibrillation signal when the amount of time an isoelectric line is present in said input ECG signal is less than a predetermined threshold.

2. The arrhythmia detector as claimed in claim 1, wherein said means for issuing generates said fibrillation signal when the amount of time an isoelectric line is present is less than said predetermined threshold during two out of three consecutive predetermined periods of time.

3. The arrhythmia detector as claimed in claim 1, wherein said means for delta modulating comprises an integrator, and said means for delta modulating outputs a digital signal based on the difference of said input ECG signal and a tracking signal provided by said integrator.

4. The arrhythmia detector as claimed in claim 1, wherein said means for periodically classifying includes means for generating a series of binary bits and storing said bits in a digital word, means for counting the number of zeros in said digital word, and means for generating said positive slope signal when said digital word has a greater number of zeros than ones, generating said negative slope signal when said digital word has a lesser number of zeros than ones, and generating said no slope signal when the number of zeros equals that of ones.

5. The arrhythmia detector as claimed in claim 1, wherein said means for determining generates an isoelectric line number corresponding to said amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time, and wherein said means for determining includes a finite state machine means for mapping sequences of said no slope, positive slope, and negative slope signals to a finite number of states and certain transitions between states are assigned line count value increments corresponding to said isoelectric line number.

6. An arrhythmia detector for detecting at lest fibrillation and tachycardia by analyzing ECG signals obtained from a heart, said detector comprising:

means for digitizing an input ECG signal for obtaining a digital signal representation of said input ECG signal;

means for periodically classifying said digital signal representation as representing no slope, positive slope, or negative slope segments of said input ECG signal and respectively generating a no slope, positive slope, or negative slope signal;

means for determining, based upon a sequence over a predetermined period of time of no slope, positive slope, and negative slope signals the amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time, and generating an isoelectric line number corresponding to said amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time;

means for issuing and storing a preliminary fibrillation signal when the isoelectric line number fails to exceed a first threshold during said predetermined period of time such that if for any set of three consecutive predetermined periods of time at least two preliminary fibrillation signals are issued, a declaration of fibrillation is made; and

means for issuing a tachycardia signal when said isoelectric line number fails to exceed a second threshold, said second threshold being greater than said first threshold during said predetermined period of time.

7. The arrhythmia detector of claim 6, wherein said means for digitizing is a delta modulator means for outputting a digital signal.

8. The arrhythmia detector of claim 6, wherein said means for periodically classifying said digital signal includes means for generating a series of binary bits and storing said bits in a digital word, means for counting the number of zeros in said digital word, and means for generating said positive slope signal when said digital word has a greater number of zeros than ones, generating said negative slope signal when said digital word has a lesser number of zeros than ones, and generating said no slope signal when the number of zeros equals that of ones.

9. The arrhythmia detector of claim 6, wherein said means for determining includes a finite state machine means for mapping sequences of said no slope, positive slope, and negative slope signals to a finite number of states and certain transitions between states are assigned line count values corresponding to said isoelectric line number.

10. The arrhythmia detector of claim 8, wherein said digital word has 8 binary bits representing an eight millisecond interval of said input ECG signal.

11. The arrhythmia detector of claim 6, wherein said predetermined period of time is two second and corresponds to 256 consecutive eight bit digital words representing consecutive eight millisecond intervals of said input ECG signal.

12. The arrhythmia detector as claimed in claim 6, including means for determining an average rate of the four last R--R intervals, wherein said means for issuing said tachycardia signal sets said second threshold according to a linear function of the heart rate as determined by said average rate of the four last R--R intervals within said predetermined period of time, wherein said linear function being determined by a linear regression from multiple data pairs collected from an extended period of time when the heart is in normal rhythm, said data pairs consisting of said isoelectric line number and said average rate of the last four R--R intervals within said predetermined period of time.

13. An arrhythmia detector for analyzing ECG signals obtained from a heart, said detector comprising:

means for digitizing an input ECG signal for obtaining a digital representation of said input ECG signal;

means for periodically classifying said digital signal as representing no slope, positive slope, and negative slope segments of said input ECG signal and respectively generating a no slope, positive slope, or negative slope signal;

means for determining, based upon a sequence over a predetermined period of time of no slope, positive slope, or negative slope signals, the amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time; and

means for issuing a fibrillation signal when the amount of time an isoelectric line is present in said input ECG signal is less than a first predetermined threshold.

14. The arrhythmia detector as claimed in claim 13, and further comprising means for issuing a tachycardia signal when the amount of time an isoelectric line segment is present fails to exceed a second threshold, said second threshold being greater than said first predetermined threshold during said predetermined period of time.

15. The arrhythmia detector as claimed in claim 14, including means for determining an average rate of the four last R--R intervals, wherein said means for issuing said tachycardia signal sets said second threshold according to a linear function of the heart rate as determined by said average rate of the four last R--R intervals within the predetermined period of time, wherein said linear function being determined by a linear regression from multiple data pairs collected from an extended period of time when the heart is in normal rhythm, said data pairs consisting of the amount of time an isoelectric line segment is present and average rate of the last four R--R intervals within said predetermined period of time.

16. An arrhythmia detector for detecting at least fibrillation by analyzing ECG signals obtained from a heart, said detector comprising:

means for digitizing an input ECG signal for obtaining a digital representation of said input ECG signal;

means for periodically classifying said digital signal as representing no slope, positive slope, and negative slope segments of said input ECG signal and respectively generating a no slope, positive slope, or negative slope signal;

means for determining, based upon a sequence over a predetermined period of time of no slope, positive slope, or negative slope signals the amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time, and generating an isoelectric line number corresponding to said amount of time an isoelectric line segment is present in said input ECG signal over said predetermined period of time; and

means for issuing and storing a preliminary fibrillation signal when the isoelectric line number fails to exceed a first threshold during said predetermined period of time such that if for any set of three consecutive predetermined periods of time at least two preliminary fibrillation signals are issued, a declaration of fibrillation is made.

17. The arrhythmia detector as claimed in claim 16, further comprising means for issuing a tachycardia signal when said isoelectric line number fails to exceed a second threshold, said second threshold being greater than said first threshold during said predetermined period of time.

18. The arrhythmia detector as claimed in claim 17, including means for determining an average rate of the four last R--R intervals, wherein said means for issuing said tachycardia signals sets said second threshold according to a linear function of the heart rate as determined by said average rate of the four last R--R intervals within the predetermined period of time, wherein said linear function being determined by a linear regression from multiple data pairs collected from an extended period of time when the heart is in normal rhythm, said data pairs consisting of said isoelectric line number and said average rate of the last R--R intervals within said predetermined period of time.