Method and apparatus for determining blood pressure on measured heart sounds

1. A method of determining blood pressure, comprising:

measuring heart sounds through transducers on an outer surface of a chest;

separating the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterizing S 1 and S 2 ;

determining a blood pressure based on the characterization;

denoising the heart sounds;

determining a heart rate; and

segmenting the heart sounds to identify S 1 and S 2 ;

wherein denoising S 1 and S 2 comprises determining a propagation speed of signal propagation within a patient by iteratively varying an assumed distance between a sensor and a heart valve; and

identifying S 1 and S 2 based on the determination of the propagation speed of signal propagation.

2. The method of claim 1 , wherein determining the propagation speed includes establishing an influence matrix that identifies an optimal influence matrix to separate source signals.

3. The method of claim 1 , wherein separating the measured heart sounds into S 1 and S 2 further comprises:

applying an autocorrelation function to the denoised heart sounds to generate correlated S 1 and S 2 sounds;

applying a moving window based on a normal heart beat to the correlated S 1 and S 2 sounds;

retaining S 1 peaks and discarding two peaks between the S 1 peaks; and

calculating an average interval between at least two of the S 1 peaks.

4. The method of claim 1 ,

wherein segmenting the heart sounds to identify S 1 and S 2 comprises:

selecting a resampling window based on the determined heart rate;

resampling the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determining a duration between the local maxima; and

determining which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .

5. The method of claim 1 , wherein mathematically characterizing S 1 and S 2 further comprises calculating growth rates and decay rates from S 1 and S 2 .

6. The method of claim 1 , wherein determining a blood pressure based on the characterization of S 1 and S 2 is continuous.

7. A non-transitory computer-readable medium tangibly embodying computer-executable instructions of a program being executable by a hardware processor of a computing device with a user interface to provide operations comprising:

measuring heart sounds through non-invasive transducers;

separating the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterizing S 1 and S 2 ;

determining a blood pressure based on the characterization;

denoising the heart sounds;

determining a heart rate; and

segmenting the heart sounds to identify S 1 and S 2 ;

wherein the denoising instructions for denoising S 1 and S 2 further comprise:

determining a propagation speed of signal propagation within a patient by iteratively varying an assumed distance between a sensor and a heart valve; and

identifying S 1 and S 2 based on the determination of the propagation speed of signal propagation.

8. The computer-readable medium of claim 7 , wherein the instructions for determining the propagation speed further comprise instructions to establish an influence matrix that identifies an optimal influence matrix to separate source signals.

9. The computer-readable medium of claim 7 , wherein the instructions separating the measured heart sounds into S 1 and S 2 further comprise instructions to:

apply an autocorrelation function to the denoised heart sounds to generate correlated S 1 and S 2 sounds;

apply a moving window based on a normal heart beat to the correlated S 1 and S 2 sounds;

retain S 1 peaks and discarding two peaks between the S 1 peaks; and

calculate an average interval between at least two of the S 1 peaks.

10. The computer-readable medium of claim 7 ,

wherein the instructions segmenting the heart sounds to identify S 1 and S 2 further comprise instructions to:

select a resampling window based on the determined heart rate;

resample the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determine a duration between the local maxima; and

determine which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .

11. The computer-readable medium of claim 7 , wherein the instructions mathematically characterizing S 1 and S 2 further comprise instructions to calculate growth rates and a decay rates from S 1 and S 2 .

12. The computer-readable medium of claim 7 , wherein determining a blood pressure based on the characterization of S 1 and S 2 is continuous.

13. A system for measuring blood pressure, comprising:

a sensor configured non-invasively and positioned on a chest area to acquire heart sounds; and

a computer coupled to the sensor, the computer configured to:

measure heart sounds;

separate the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterize S 1 and S 2 ;

determine a blood pressure based on the characterization;

denoise the heart sounds;

determine a heart rate; and

segment the heart sounds to identify S 1 and S 2 ;

wherein the denoising further comprises the computer configured to:

determine a propagation speed of signal propagation within a patient by iteratively varying an assumed distance between a sensor and a heart valve; and

identify S 1 and S 2 based on the determination of the propagation speed of signal propagation; and

wherein the instructions to determine the propagation speed include the computer configured to establish an influence matrix that identifies an optimal influence matrix to separate source signals.

14. The system of claim 13 , wherein separating the measured heart sounds into S 1 and S 2 further comprises the computer configured to:

apply an autocorrelation function to the denoised heart sounds to generate correlated S 1 and S 2 sounds;

apply a moving window based on a normal heart beat to the correlated S 1 and S 2 sounds;

retain S 1 peaks and discarding two peaks between the S 1 peaks; and

calculate an average interval between at least two of the S 1 peaks.

15. The system of claim 13 , wherein segmenting the heart sounds to identify S 1 and S 2 further comprises the computer configured to:

select a resampling window based on the determined heart rate;

resample the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determine a duration between the local maxima; and

determine which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .

16. The system of claim 13 , wherein mathematically characterizing S 1 and S 2 further comprises the computer configured to calculate growth rates and a decay rates from S 1 and S 2 .

17. The system of claim 13 , wherein determining a blood pressure based on the characterization of S 1 and S 2 is continuous.

18. A method of determining blood pressure, comprising:

measuring heart sounds through transducers on an outer surface of a chest;

separating the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterizing S 1 and S 2 ;

determining a blood pressure based on the characterization;

denoising the heart sounds;

determining a heart rate; and

segmenting the heart sounds to identify S 1 and S 2 ;

wherein segmenting the heart sounds to identify S 1 and S 2 comprises:

selecting a resampling window based on the determined heart rate;

resampling the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determining a duration between the local maxima; and

determining which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .

19. A non-transitory computer-readable medium tangibly embodying computer-executable instructions of a program being executable by a hardware processor of a computing device with a user interface to provide operations comprising:

measuring heart sounds through non-invasive transducers;

separating the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterizing S 1 and S 2 ;

determining a blood pressure based on the characterization;

denoising the heart sounds;

determining a heart rate; and

segmenting the heart sounds to identify S 1 and S 2 ; and

wherein the instructions segmenting the heart sounds to identify S 1 and S 2 further comprise instructions to:

select a resampling window based on the determined heart rate;

resample the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determine a duration between the local maxima; and

determine which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .

20. A system for measuring blood pressure, comprising:

a sensor configured non-invasively and positioned on a chest area to acquire heart sounds; and

a computer coupled to the sensor, the computer configured to:

measure heart sounds;

separate the measured heart sounds into a first heart sound (S 1 ) and a second heart sound (S 2 );

mathematically characterize S 1 and S 2 ;

determine a blood pressure based on the characterization;

denoise the heart sounds;

determine a heart rate; and

segment the heart sounds to identify S 1 and S 2 ;

wherein segmenting the heart sounds to identify S 1 and S 2 further comprises the computer configured to:

select a resampling window based on the determined heart rate;

resample the measured heart sounds to identify local maxima and a maximum peak in a cardiac cycle;

determine a duration between the local maxima; and

determine which of the local maxima represents S 1 and which represents S 2 based on a duration between the local maxima, based on an interval S 1 to S 2 that is known to be smaller than an interval of S 2 to S 1 .