Method and apparatus for estimating a physiological parameter from a physiological signal

• US 6,594,512 B2
• Filed: 07/26/2001
• Issued: 07/15/2003
• Est. Priority Date: 11/21/2000
• Status: Expired due to Term

• Current Assignee: Dragerwerk AG & Company KGaA (FPS Vermögensverwaltung GmbH)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Huang, John W.
• Primary Examiner(s): Winakur, Eric F.
• Assistant Examiner(s): KREMER, MATTHEW J
• Application Number:
  US09/915,610
• Publication Number:
  US 20020099280A1
• Time in Patent Office: 719 Days
• Field of Search: 600/309-310, 600/322-326, 600/330, 600/336
• US Class Current: 600/324
• CPC Class Codes: A61B 5/14551 for measuring blood gases
• CPC Class Codes: A61B 5/7257 using Fourier transforms

Abstract

A system is disclosed which first identifies a plurality of characteristics of a physiological signal any one of which may represent a physiological parameter. A plurality of different techniques are used to provide respective likelihood factors for each such identified characteristic. The resulting likelihood factors are then analyzed to select the one characteristic of the physiological signal which most likely represents the desired physiological parameter. The physiological parameter is then calculated from the selected characteristic of the physiological signal.

Claims

1. Apparatus for estimating a physiological parameter from a physiological signal derived from patient blood, comprising:

a source of the physiological signal;

a characteristic identification processor, coupled to the physiological signal source, for identifying respective characteristics in the physiological signal;

at least one likelihood processor, coupled to the characteristic identification processor and deriving likelihood representative values from a function associating probability values with pulse rates, said values respectively corresponding to the identified characteristics and representing a probability that the corresponding identified characteristic is associated with said physiological parameter;

an arbitrator for selecting one of the identified characteristics in response to the likelihood representative values; and

a computational processor, for estimating the physiological parameter responsive to the selected one of the identified characteristics.

2. The apparatus of claim 1 wherein the function associates probability values with pulse rates and said probability values are adjusted in response to at least one of, (a) a previously computed pulse rate and (b) a previously computed blood oxygen saturation representative value.

3. The apparatus of claim 1 wherein the characteristic identification processor comprises a peak identifying processor for identifying one or more peaks at respective corresponding frequencies in the spectrum of the physiological signal as characteristics of the physiological signal.

4. The apparatus of claim 3 wherein the peak identification processor comprises:

a fast Fourier transform (FFT) circuit, coupled to the physiological signal source, for generating a frequency spectrum, having respective magnitudes at associated frequencies, representing the physiological signal;

a peak locating processor, responsive to the spectrum, for identifying peaks in the respective magnitudes and the frequency associated with the peaks.

5. The apparatus of claim 4 further comprising a window filter coupled to the peak identification processor for generating a plurality of time domain signals window filtered at frequencies respectively corresponding to the identified peaks.

6. The apparatus of claim 5 wherein the window filter comprises:

a circuit selecting a subset of magnitudes in the frequency spectrum in the neighborhood of the frequency of each identified peak; and

an inverse fast Fourier transform processor, for generating a time domain window filtered signal in response to the selected subset of magnitudes.

7. The apparatus of claim 3, wherein:

the frequency of each identified peak has an associated period; and

one of the plurality of likelihood processors comprises;

a circuit for generating a plurality of time domain signals, each signal being the physiological signal window filtered in the neighborhood of the frequency of a respective corresponding identified peak;

a processor for generating a plurality of time delayed signals respectively corresponding to the plurality of time domain signals delayed by respective time intervals equal to one-half the period associated with the time domain signal; and

a processor for generating a plurality of likelihood signals, each corresponding to a respective one of the identified peaks, representing degree of congruence of the time domain signal to the delayed time domain signal corresponding to the one of the identified peaks.

8. The apparatus of claim 7 wherein the time delayed signal generating processor delays the plurality of time domain signals by respective time intervals equal to (n+½

) times the period associated with the time domain signal where n is an integer greater than or equal to zero.

9. The apparatus of claim 1 wherein:

the estimated physiological parameter comprises a patient pulse rate and the identified characteristics comprise peaks in the physiological signal and the physiological signal comprises a blood oxygen representative signal.

10. The apparatus of claim 1

further comprising a bandpass filter coupled between the physiological signal source and the characteristic identification processor.

11. A method for estimating pulse rate from a physiological signal derived from patient blood, comprising the steps of:

receiving a physiological signal;

identifying respective characteristics in said received physiological signal;

deriving likelihood representative values from a function associating probability values with pulse rates, said values respectively corresponding to the identified characteristics and representing a probability that the corresponding identified characteristic is associated with said physiological parameter;

selecting one of said identified characteristics in response to said likelihood representative values; and

estimating said pulse rate using said selected one of said identified characteristics.

12. The method of claim 11, wherein

the function associates probability values with pulse rates and said probability values are adjusted in response to at least one of, (a) a previously computed pulse rate and (b) a previously computed blood oxygen saturation representative value.

13. The method of claim 11,

wherein the function is a probability density function.

14. The method of claim 11 wherein the identified characteristics comprise peaks in the physiological signal and the physiological signal comprises a blood oxygen representative signal.

15. The method of claim 11 wherein:

the physiological signal comprises first and second component signals; and

including the steps of;

generating a first plurality of time domain signals, each signal being the first component signal window filtered in the neighborhood of the frequency of a respective corresponding identified peak;

generating a second plurality of time domain signals, each signal being the second component signal window filtered in the neighborhood of the frequency of a respective corresponding identified peak; and

generating a plurality of likelihood signals, each corresponding to a respective one of the identified peaks, representing degree of congruence of the corresponding one of the first plurality of time domain signals, to the corresponding one of the second plurality of time domain signals.

16. The method of claim 11, wherein said step of deriving likelihood representative values comprises deriving a likelihood representative value corresponding to a particular identified characteristic from a function associating probability values with pulse rates, in response to an identified pulse rate corresponding to the particular identified characteristic.

17. The method of claim 11 further including the step of bandpass filtering data representing identified characteristics and deriving said likelihood representative values based on filtered identified characteristic representative data.

18. The method of claim 11 wherein:

the frequency of each identified peak has an associated period; and

further including the step of;

generating respective pluralities of time delayed signals respectively corresponding to the first and second pluralities of time domain signals delayed by respective time intervals.

19. The method of claim 11 wherein said step of deriving likelihood representative values includes the step of generating a plurality of signals, respectively representing the probability that corresponding peak represents the physiological parameter.

20. The method of claim 19 including the steps of extracting the probability that the frequency location of each identified peak in the physiological signal represents the physiological parameter;

determining a degree of spread of the previously estimated physiological parameters; and

generating a new probability density function.

21. The method of claim 20 including the step of scaling the probability density function.

22. The method of claim 11 wherein, said received physiological signal is derived from an SpO₂ sensor.

23. A method for estimating a physiological parameter from a physiological signal derived from patient blood, comprising the steps of:

receiving a physiological signal;

identifying respective characteristics in said received physiological signal;

deriving likelihood representative values from a function associating probability values with pulse rates, said values respectively corresponding to the identified characteristics and representing a probability that the corresponding identified characteristic is associated with said physiological parameter;

selecting one of said identified characteristics in response to said likelihood representative values; and

estimating said physiological parameter using said selected one of said identified characteristics.

24. A method according to claim 23, including the step of bandpass filtering said received physiological signal and wherein said identifying step comprises identifying respective characteristics in said bandpass filtered received physiological signal.