Continuous cardiac output derived from the arterial pressure waveform using pattern recognition

• US 5,390,679 A
• Filed: 06/03/1993
• Issued: 02/21/1995
• Est. Priority Date: 06/03/1993
• Status: Expired due to Fees

• Current Assignee: Eli Lilly and Company
• Original Assignee: Eli Lilly and Company
• Inventors: Martin, James F.
• Primary Examiner(s): Sykes, Angela D.
• Application Number:
  US08/072,038
• Time in Patent Office: 628 Days
• Field of Search: 128/668, 128/672-683, 128/691, 128/692, 128/694, 128/713
• US Class Current: 600/486
• CPC Class Codes: A61B 5/0215 by means inserted into the body catheters A61M25/00
• CPC Class Codes: A61B 5/029 Measuring or recording blood output from the heart, e.g. minute volume A61B8/065 takes precedence
• CPC Class Codes: A61B 5/7239 using differentiation including higher order derivatives
• 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

Abstract

A system and method of determining cardiac output from a cardiac pressure waveform. The pressure waveform is continuously sensed and converted into a digital data stream. When a complete beat frame of data is sensed, the processor extracts a plurality of features from the sensed waveform which characterize the waveform. These features are compared to stored representative pressure waveforms (RPW'"'"'s) which have a known cardiac output. The RPW'"'"'s are obtained by applying a pattern recognition methodology to pressure waveforms of known cardiac output. In one embodiment the typical shapes function is used. The pattern recognition methodology results in classes and associated cardiac outputs which are stored as the RPW'"'"'s. The features extracted from waveforms representative of unknown cardiac output are thus classified by the comparison to the RPW'"'"'s. When a match is made between the extracted features and an RPW, the cardiac output of the selected RPW is output as the cardiac output of the patient.

Claims

1. A system for determining the value of a characteristic of the vascular system of a patient from the blood pressure existing in that vascular system, comprising:

a sensor which senses the blood pressure in the vascular system and which provides a pressure signal representative of that blood pressure;

a digitizer which receives the pressure signal from the sensor and converts the pressure signal into a digitized pressure signal;

a memory having stored therein a plurality of representative pressure waveforms, each representative pressure waveform comprising a set of values of selected features of a respective pressure signal having associated therewith a known value of the characteristic, the memory also storing the known characteristic value corresponding to each representative pressure waveform; and

a processor which;

receives the digitized pressure signal from the digitizer;

extracts at least one feature of the selected set of features from the digitized pressure signal and determines the value of each extracted feature;

compares the value of each extracted feature from the digitized pressure signal to the value of each corresponding feature of representative pressure waveforms stored in the memory;

based on the comparison, selects the representative pressure waveform most similar to the digitized pressure signal; and

provides the characteristic value corresponding to the selected representative pressure waveform as

the value of the characteristic of the vascular system of the patient.

2. The system of claim 1 wherein the processor extracts the at least one feature from the selected set of features from the digitized pressure signal of a single beat frame of a heart.

3. The system of claim 1 wherein the sensor senses the blood pressure continuously and the digitizer continuously converts the pressure signal into a digitized pressure signal.

4. The system of claim 1 further comprising a display device which receives the characteristic value from the processor and displays that value.

5. The system of claim 4 wherein the display device comprises a monitor which visually displays the characteristic value.

6. The system of claim 4 wherein the display device is a recorder which records the characteristic value.

7. The system of claim 1 wherein the processor further:

extracts all of the selected features from the digitized pressure signal and determines the value of each extracted feature; and

compares the values of the extracted features of the digitized pressure signal to the values of the corresponding features of representative pressure waveforms;

selects the representative pressure waveform having feature values closest to the corresponding extracted feature values of the digitized pressure signal and provides the value of the characteristic of the selected representative pressure waveform as the value of the characteristic of the vascular system of the patient.

8. The system of claim 1 wherein the memory has stored therein representative pressure waveforms each of which comprises a class of values of selected features of a respective pressure signal having associated therewith a known value of the characteristic of the vascular system, wherein the classes are determined from actual pressure signals generated by vascular systems having known values of the characteristic, the classes resulting from application of a pattern recognition method to the actual pressure signals.

9. The system of claim 8 wherein the classes stored in the memory result from application of the Typical Shape Function pattern recognition method applied to the actual pressure signals.

10. The system of claim 1 wherein at least one of the selected features is selected from the group consisting of:

the pressure at the end of diastole;

the systolic pressure;

or the pressure of the dicrotic notch.

11. The system of claim 1 wherein at least one of the selected features is selected from the group consisting of:

the time at the end of diastole;

the time of systolic pressure;

the time of the dicrotic notch;

the period of the cardiac cycle;

or the rate of the cardiac cycle.

12. The system of claim 1 wherein at least one of the selected features is a mathematical feature selected from the group consisting of:

the exponential decay time constant of the pressure signal;

the derivative of the pressure signal at the systolic pressure;

the coefficients of the Fourier Series of the pressure signal;

the coefficients of the Fourier Contours of the pressure signal;

the Fourier Transform of the pressure signal;

the Beta-spline description of the pressure signal;

the mean, the maximum, the minimum, the standard deviation, covariance or entropy of the pressure signal.

13. The system of claim 1 wherein the processor determines the arterial resistance from the pressure signal as the characteristic of the patient'"'"'s vascular system.

14. The system of claim 1 wherein the processor determines the arterial compliance from the pressure signal as the characteristic of the patient'"'"'s vascular system.

15. The system of claim 1 wherein the processor determines the ventricular contractility from the pressure signal as the characteristic of the patient'"'"'s vascular system.

16. The system of claim 1 wherein the processor determines the cardiac output of the cardiovascular system from the pressure signal as the characteristic of the patient'"'"'s vascular system.

17. A method for determining the value of a characteristic of the vascular system of a patient from the blood pressure existing in that vascular system, comprising the steps of:

sensing the blood pressure in the vascular system and providing a pressure signal representative of that blood pressure;

converting the pressure signal representative of sensed blood pressure into a digitized pressure signal;

storing a plurality of representative pressure waveforms, each representative pressure waveform comprising a set of values of selected features of a respective pressure signal having associated therewith a known value of the characteristic and storing the known characteristic value corresponding to each representative pressure waveform;

extracting at least one feature of the selected set of features from the digitized pressure signal and determining the value of each extracted feature;

comparing the value of each extracted feature from the digitized pressure signal to the value of each corresponding feature of the stored representative pressure waveforms;

based on the comparison, selecting the representative pressure waveform most similar to the digitized pressure signal; and

providing the characteristic value corresponding to the selected representative pressure waveform as the value of the characteristic of the vascular system of the patient.

18. The method of claim 17 wherein the step of extracting comprises the step of extracting at least one feature from the digitized pressure signal over a single beat frame of a heart.

19. The method of claim 17 wherein the step of sensing comprises the step of continuously sensing the blood pressure and the step of converting comprises continuously converting the pressure signal into a digitized pressure signal.

20. The method of claim 17 further comprising the step of displaying the characteristic value corresponding to the selected representative pressure waveform.

21. The method of claim 20 wherein the step of displaying comprises the step of visually displaying the characteristic value corresponding to the selected representative pressure waveform.

22. The method of claim 20 wherein the step of displaying comprises the step of recording the characteristic value corresponding to the selected representative pressure waveform.

23. The method of claim 17 wherein:

the step of extracting comprises the step of extracting all of the selected features from the digitized pressure signal and determining the value of each extracted feature;

the step of comparing comprises comparing the values of the extracted selected features of the digitized pressure signal to the values of the corresponding features of representative pressure waveforms;

the step of selecting comprises selecting the representative pressure waveform having feature values closest to the corresponding extracted feature values of the digitized pressure signal;

the step of providing comprises providing the value of the characteristic of the selected representative pressure waveform as the value of the characteristic of the vascular system of the patient.

24. The method of claim 17 wherein the step of storing representative pressure waveforms comprises the steps of:

sensing actual pressure signals generated by vascular systems having known values of the characteristic;

applying a pattern recognition method to the actual pressure signals to result in a plurality of classes each with a corresponding value of the characteristic; and

storing the classes and respective corresponding values of the characteristic.

25. The method of claim 24 wherein the step of applying a pattern recognition method comprises applying a Typical Shape Function pattern recognition method to the actual pressure signals to result in the classes.

26. The method of claim 17 wherein the step of extracting comprises extracting at least one of the selected features from the group consisting of:

the pressure at the end of diastole;

the systolic pressure;

or the pressure of the dicrotic notch.

27. The method of claim 17 wherein the step of extracting comprises extracting at least one of the selected features is selected from the group consisting of:

the time at the end of diasrole;

the time of systolic pressure;

the time of the dicrotic notch;

the period of the cardiac cycle;

or the rate of the cardiac cycle,

28. The method of claim 31 wherein the step of extracting comprises extracting at least one of the selected features is a mathematical feature selected from the group consisting of:

the exponential decay time constant of the pressure signal;

the derivative of the pressure signal at the systolic pressure;

the coefficients of the Fourier Series of the pressure signal;

the coefficients of the Fourier Contours of the pressure signal;

the Fourier Transform of the pressure signal;

the Beta-spline description of the pressure signal;

the mean, the median, the maximum, the minimum, the standard deviation, covariance or entropy of the pressure signal.

29. The method of claim 17 wherein the step of storing comprises storing the known value of the arterial resistance as the characteristic value.

30. The method of claim 17 wherein the step of storing comprises storing the known value of the arterial compliance as the characteristic value.

31. The method of claim 17 wherein the step of storing comprises storing the known value of the ventricular contractility as the characteristic value.

32. The method of claim 17 wherein the step of storing comprises storing the known value of the cardiac output as the characteristic value.