Pulse Rate Determination Using Gaussian Kernel Smoothing of Multiple Inter-Fiducial Pulse Periods

• US 20130137945A1
• Filed: 11/30/2011
• Published: 05/30/2013
• Est. Priority Date: 11/30/2011
• Status: Active Grant

• Current Assignee: Covidien LP (Medtronic Plc)
• Original Assignee: Nellcor Puritan Bennett Ireland
• Inventors: Addison, Paul S., Watson, James
• Granted Patent: US 8,870,783 B2
• Time in Patent Office: Days
• US Class Current: 600/323
• CPC Class Codes: A61B 5/024 Detecting, measuring or recording pulse rate or heart rate A61B5/0205, A61B5/021 take precedence
• CPC Class Codes: A61B 5/02416 using photoplethysmograph signals, e.g. generated by infrared radiation A61B5/14552 takes precedence
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence

Abstract

Systems and methods are provided for determining the pulse rate of a patient from multiple fiducial points using Gaussian kernel smoothing. Based on acquired pleth signals, each recorded fiducial pulse period is converted to a Gaussian kernel function. The Gaussian kernel functions for all recorded fiducial points are summed to generate a Gaussian kernel smoothed curve. The pulse rate of a patient may be determined from the Gaussian kernel smoothed curve. All acquired fiducial pulse periods contribute to generate the Gaussian kernel smoothing curve. The number of fiducial points utilized may change to improve pulse rate determination or provide additional functionality to the system.

Claims

1. A method for determining a physiological parameter from a physiological signal, comprising:

identifying at least one fiducial point on consecutive waves of the physiological signal, wherein corresponding fiducial points of the consecutive waves form sets of fiducial points;

determining a fiducial period for each set of fiducial points;

converting each of the fiducial periods to a respective Gaussian kernel function;

summing the Gaussian kernel functions to create a distribution curve; and

determining the physiological parameter from the distribution curve.

2. The method, as set forth in claim 1, wherein the at least one fiducial point comprises a peak of the physiological signal, a trough of the physiological signal, a dicrotic notch of the physiological signal, a point of maximum slope of the physiological signal, a point of maximum curvature of the physiological signal, a pulse mean crossing of the physiological signal, a point of zero crossing in the physiological signal, a point of zero crossing in a derivative of the physiological signal, or any combination thereof.

3. The method, as set forth in claim 1, wherein identifying one or more fiducial points comprises recording data for each fiducial point, wherein data comprises a fiducial point identifier, a time of the fiducial point, an amplitude of the fiducial point, a derivative of the physiological signal at the fiducial point, or any combination thereof.

4. The method, as set forth in claim 1, wherein identifying one or more fiducial points comprises identifying fiducial points based on a ranking of the fiducial points.

5. The method, as set forth in claim 4, wherein the ranking is based on a reliability factor of each fiducial point, a consistency factor of each fiducial point, a location factor of each fiducial point, a quality factor of each fiducial point, or any combination thereof.

6. The method, as set forth in claim 4, wherein the ranking of each fiducial point is dynamic and based on an effect the respective Gaussian kernel function has on the distribution curve.

7. The method, as set forth in claim 1, wherein identifying one or more fiducial points comprises identifying a number of fiducial points based on a user input number.

8. The method, as set forth in claim 1, wherein determining the physiological parameter from the distribution curve comprises determining the peak of the distribution curve.

9. The method, as set forth in claim 1, comprising filtering the fiducial periods to remove fiducial periods corresponding to a heart rate outside the range of about 20-350 BPM.

10. The method, as set forth in claim 1, comprising filtering the physiological signal to remove a non-physiological signal component.

11. The method, as set forth in claim 1, wherein a standard deviation of each Gaussian kernel function is about 0.5.

12. A monitor comprising:

a processor configured to;

identify at least one fiducial point on consecutive waves of a physiological signal;

determine a fiducial period between the at least one fiducial point for each consecutive wave;

convert each of the fiducial periods to a respective Gaussian kernel function;

sum the Gaussian kernel functions to create a distribution curve; and

determine a physiological parameter from the distribution curve;

a memory configured to store data associated with each identified fiducial points.

13. The monitor, as set forth in claim 12, wherein the at least one fiducial point comprises a peak of the physiological signal, a trough of the physiological signal, a dicrotic notch of the physiological signal, a point of maximum slope of the physiological signal, a point of maximum curvature of the physiological signal, a pulse mean crossing of the physiological signal, a point of zero crossing in the physiological signal, a point of zero crossing in a derivative of the physiological signal, or any combination thereof.

14. The monitor, as set forth in claim 12, wherein the processor is configured to identify at least one fiducial point based on a ranking of the fiducial points, wherein the ranking is based on a reliability factor of each fiducial point, a consistency factor of each fiducial point, a location factor of each fiducial point, a quality factor of each fiducial point, or any combination thereof.

15. The monitor, as set forth in claim 14, wherein the processor is configured to determine a physiological parameter from the distribution curve created by the sum of a subset of Gaussian kernel functions.

16. The monitor, as set forth in claim 12, wherein the data stored in memory comprises a fiducial point identifier for each fiducial point, a time the fiducial point was identified, an amplitude of the identified fiducial point, a derivative of the physiological signal at the identified fiducial point, or any combination thereof.

17. A pulse oximetry system, comprising:

a sensor comprising;

an emitter configured to transmit electromagnetic radiation; and

a detector configured to detect the electromagnetic radiation and generate electrical signals based upon the detected electromagnetic radiation; and

a monitor coupled to the sensor, wherein the monitor is configured to;

identify multiple fiducial points on consecutive waves of the electrical signal;

determine a fiducial period between the multiple fiducial points for each consecutive wave;

determine a pulse rate based on each fiducial period using Gaussian kernel smoothing.

18. The pulse oximetry system, as set forth in claim 17, comprising a display configured to display the electrical signal, calculated physiological parameters, notifications associated with the operation of the system, and combinations thereof.

19. The pulse oximetry system, as set forth in claim 17, comprising a memory configured to store identified fiducial points.

20. The pulse oximetry system, as set forth in claim 17, comprising a filter configured to remove electrical signals associated with extreme or unlikely physiologically conditions.