Method and apparatus for determining a central aortic pressure waveform

• US 10,052,070 B2
• Filed: 11/19/2012
• Issued: 08/21/2018
• Est. Priority Date: 03/15/2006
• Status: Active Grant

• Current Assignee: The Board of Trustees of Michigan State University (Michigan State University)
• Original Assignee: The Board of Trustees of Michigan State University (Michigan State University)
• Inventors: Mukkamala, Ramakrishna, Swamy, Gokul, Ling, Qi, Li, Tongtong
• Primary Examiner(s): Jang, Christian
• Application Number:
  US13/680,404
• Publication Number:
  US 20130102909A1
• Time in Patent Office: 2,101 Days
• CPC Class Codes: A61B 5/021 Measuring pressure in heart or blood vessels A61B5/0205 takes precedence
• CPC Class Codes: A61B 5/02108 from analysis of pulse wave characteristics
• CPC Class Codes: A61B 5/7225 Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation A61B5/30, A61B5/316 take precedence; transformers for specific applications H01F38/00; amplifiers H03F
• CPC Class Codes: A61B 5/7246 using correlation, e.g. template matching or determination of similarity
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• 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: A61B 5/7278 Artificial waveform generation or derivation, e.g. synthesising signals from measured signals A61B5/327 takes precedence

Abstract

A method is provided for determining a central aortic pressure waveform. The method includes: measuring two or more peripheral artery pressure waveforms; analyzing the signals so as to extract common features in the measured waveforms; and determining an absolute central aortic pressure waveform based on the common features.

Claims

1. A method for determining an absolute central aortic pressure waveform comprising:

measuring, using a sensor, peripheral artery pressure signals or related signals at more than one peripheral location in an arterial tree of a subject;

modeling the arterial tree as a single input, multi-output system in which a common input corresponds to aortic blood pressure and each output corresponds to one of the measured signals;

analyzing, by a computing device having a processor, the measured signals so as to extract features common to the measured signals and thereby estimate the common input of the system to within an arbitrary scale factor; and

determining, by the computing device, an absolute central aortic pressure waveform by scaling the estimated input based on the measured signals.

2. The method of claim 1 wherein channels of the single-input, multi-output system model of the arterial tree are first estimated to within an arbitrary scale factor and then an inverse of the estimated channels is applied to the measured signals so as to reconstruct the common input to within an arbitrary scale factor.

3. The method of claim 1 wherein channels of the single-input, multi-output system model of the arterial tree are characterized by linear and time-invariant impulse responses that are coprime.

4. The method of claim 3 wherein analyzing the measured signals further comprises applying multi-channel blind system identification to the measured signals to estimate parameters and order of the impulse responses and thereby estimate the impulse responses to within an arbitrary scale factor;

and estimating the input of the system to within an arbitrary scale factor by deconvolution.

5. The method of claim 4 further comprising the representation of the impulse responses with a set of basis functions with unknown parameters prior to the step of applying multi-channel blind system identification.

6. The method of claim 5 wherein the basis functions are truncated exponential varying sinusoids or truncated polynomials for finite impulse responses.

7. The method of claim 5 wherein the basis functions are complex exponentials for infinite impulse responses.

8. The method of claim 5 wherein the number of basis functions takes on an assumed value or is determined.

9. The method of claim 8 wherein the number of basis functions is determined through mean squared error analysis.

10. The method of claim 4 wherein a subset of the unknown parameters of the impulse responses takes on assumed values.

11. The method of claim 4 wherein the parameters and order of the impulse responses are estimated based on cross relations between pairs of measured signals.

12. The method of claim 4 wherein the parameters and order of the impulse responses are estimated based on properties of a channel subspace.

13. The method of claim 4 wherein the parameters of the impulse responses are estimated using least squares methods.

14. The method of claim 4 wherein the parameters of the impulse responses are estimated with the eigenvector method, an iterative two-step maximum likelihood method, an adaptive neural network, or a numerical search.

15. The method of claim 4 wherein the orders of the impulse responses take on assumed values or are determined from the measured signals.

16. The method of claim 15 wherein a maximum order of the impulse responses is determined.

17. The method of claim 15 wherein the orders are determined from the signals by singular value analysis, cross validation, cross validation-based criteria, or information-based criteria.

18. The method of claim 4 wherein single channel deconvolution is applied to one or more of the estimated impulse responses and corresponding measured output signals so as to result in multiple versions of the common input to within an arbitrary scale factor.

19. The method of claim 18 wherein the single channel deconvolution is achieved with Fourier methods or least squares methods.

20. The method of claim 18 further comprising selecting one of the multiple versions as the common input to within an arbitrary scale factor.

21. The method of claim 18 further comprising using an average or median of at least some of the multiple versions as the common input to within an arbitrary scale factor.

22. The method of claim 4 wherein multi-channel deconvolution is applied to one or more of the estimated impulse responses and corresponding measured signals so as to result in a single common input to within an arbitrary scale factor.

23. The method of claim 22 wherein the multi-channel deconvolution is achieved based on Bezoult'"'"'s theorem or least squares methods.

24. The method of claim 3 wherein the common input of the system is estimated to within an arbitrary scale factor in one step by applying multi-channel blind system identification to the measured signals.

25. The method of claim 24 wherein the estimation is achieved with the input subspace method.

26. The method of claim 1 wherein the estimated input is scaled to have a mean value equal to that of one of the measured signals.

27. The method of claim 1 wherein the estimated input is scaled to have a mean value equal to that of the measured signal with the largest mean value.

28. The method of claim 1 wherein the estimated input is scaled to have a mean value equal to the mean or median of the mean values of select measured signals.

29. The method of claim 1 further comprising scaling the estimated impulse responses to have unity gain.

30. The method of claim 29 wherein the scaled estimated impulse responses are utilized to monitor local arterial functioning.