Systems and methods for adaptive filtering artifacts from composite signals

• US 5,630,425 A
• Filed: 02/17/1995
• Issued: 05/20/1997
• Est. Priority Date: 02/17/1995
• Status: Expired due to Term

• Current Assignee: EP Technology (Boston Scientific Corporation)
• Original Assignee: EP Technology (Boston Scientific Corporation)
• Inventors: Panescu, Dorin, Whayne, James G.
• Primary Examiner(s): Jastrzab, Jeffrey R.
• Application Number:
  US08/390,687
• Time in Patent Office: 823 Days
• Field of Search: 128/696, 128/697, 128/901, 128/642, 128/716, 128/733, 128/731, 128/736, 607/9, 607/122, 364/724.01, 364/724.19, 364/413.05, 364/413.06, 364/413.27
• US Class Current: 600/508
• CPC Class Codes: A61B 5/287 Holders for multiple electrodes, e.g. electrode catheters for electrophysiological study [EPS]
• CPC Class Codes: A61B 5/6855 with a distal curved tip
• CPC Class Codes: A61B 5/6858 with a distal basket, e.g. expandable basket
• CPC Class Codes: A61N 1/3702 Physiological parameters A61N1/365 takes precedence; evoked response A61N1/371
• CPC Class Codes: H03H 17/0263 Rank order filters
• CPC Class Codes: Y10S 128/901 Suppression of noise in electric signal

Abstract

Analog or digital systems or methods filter an artifact from a signal derived from a biological event comprising sample values arranged with respect to time. The systems and methods generate a variable expressing a template of the artifact and reduce the derived biological signal by the variable to derive a filtered derived biological signal. The systems and methods change the variable over time based upon the energy of the filtered derived biological signal to minimize the energy of the artifact remaining in the filtered derived biological signal over time. The systems and methods output the filtered derived biological signal. The systems and methods remove artifacts from signals derived from a biological event; for example, electrograms, electrocardiograms, electroencephalograms, electrogastrograms, electromyograms, and respiratory signals.

Claims

1. An element for filtering an artifact from a signal derived from a biological event comprising sample values arranged with respect to time, the element comprising

means for generating a variable expressing a template of an artifact having a substantially similar morphology to the signal,

means for reducing the derived biological signal by the variable to derive a filtered derived biological signal, and

means for changing the variable over time based upon the energy of the filtered derived biological signal to minimize the energy of the artifact remaining in the filtered derived biological signal over time and to substantially preserve the shape of the signal, and means for outputting the filtered derived biological signal.

2. An element according to claim 1

wherein the means for changing the variable changes the variable by iteration.

3. An element according to claim 2

wherein the iteration is based upon a Least-Means-Squares Method.

4. An element according to claim 2

wherein the iteration is based upon a Steepest-Descent Method.

5. An element according to claim 2

wherein the iteration is based upon a Recursive-Least-Squares Method.

6. A method for filtering an artifact from a signal derived from a biological event comprising sample values arranged with respect to time, the method comprising the steps of

generating a variable expressing a template of an artifact having a substantially similar morphology to the signal,

reducing the derived biological signal by the variable to derive a filtered derived biological signal, and

changing the variable over time based upon the energy of the filtered derived biological signal to minimize the energy of the artifact remaining in the filtered derived biological signal over time and to substantially preserve the shape of the signal.

7. A method according to claim 6

wherein the variable is changed by iteration.

8. A method according to claim 7

wherein the iteration is based upon a Least-Means-Squares Method.

9. A method according to claim 7

wherein the iteration is based upon a Steepest-Descent Method.

10. A method according to claim 7

wherein the iteration is based upon a Recursive-Least-Squares Method.

11. An element for filtering an artifact from an electrogram comprising

means for inputting the electrogram,

means for generating a variable correlated to the artifact,

means for reducing the electrogram by the variable to derive a filtered electrogram signal,

means for changing the variable over time based upon the energy of the filtered electrogram signal to minimize the energy of the artifact remaining in the filtered electrogram signal over time, and

means for outputting the filtered electrogram signal.

12. An element according to claim 11

wherein the means for changing the variable changes the variable by iteration.

13. An element according to claim 12

wherein the iteration is based upon a Least-Means-Squares Method.

14. An element according to claim 12

wherein the iteration is based upon a Steepest-Descent Method.

15. An element according to claim 12

wherein the iteration is based upon a Recursive-Least-Squares Method.

16. An element according to claim 11

wherein the variable is expressed as a template of the artifact signal portion.

17. A system for processing an electrogram comprising

electrode means for electrically communicating with a heart,

means coupled to the electrode means for pacing the heart while recording a paced electrogram IN(t) that changes with respect to time, the paced electrogram IN(t) comprising a useful electrogram portion that changes with respect to time EG(t) and that has an energy and a pacing artifact portion that changes with respect to time PACE(t) and that has an energy, and

a processing element for removing the pacing artifact portion including

filter means for generating an output TPACE(t) comprising a preselected function correlated to the pacing artifact portion,

an input for receiving IN(t),

means for subtracting TPACE(t) from IN(t) and generating the result as output EG(t), and

the filter means being further operative for changing TPACE(t) over time based upon the energy of EG(t) so as to minimize the energy of the pacing artifact within EG(t) over time.

18. A system according to claim 17

wherein the filter means changes TPACE(t) by iteration.

19. A system according to claim 18

wherein the iteration is based upon a Least-Means-Squares Method.

20. A system according to claim 18

wherein the iteration is based upon a Steepest-Descent Method.

21. A system according to claim 18

wherein the iteration is based upon a Recursive-Least-Squares Method.

22. A system according to claim 17

wherein TPACE(t) comprises a template expressing the pacing artifact portion.

23. A method for processing an electrogram having a periodic artifact signal portion and a useful signal portion, the method comprising the steps of

generating a variable correlated to the artifact signal portion,

reducing the electrogram by the variable to derive a filtered electrogram signal, and

changing the variable over time based upon the energy of the filtered electrogram signal to minimize the energy of the artifact signal portion within the filtered electrogram signal over time.

24. A method according to claim 23

wherein the variable is changed by iteration.

25. A method according to claim 24

wherein the iteration is based upon a Least-Means-Squares Method.

26. A method according to claim 24

wherein the iteration is based upon a Steepest-Descent Method.

27. A method element according to claim 24

wherein the iteration is based upon a Recursive-Least-Squares Method.

28. A method according to claim 23

wherein the variable is expressed as a template of the artifact signal portion.