Systems and methods for physiological signal enhancement and biometric extraction using non-invasive optical sensors

• US 9,918,666 B2
• Filed: 01/13/2015
• Issued: 03/20/2018
• Est. Priority Date: 01/13/2014
• Status: Active Grant

• Current Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Original Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Inventors: Yousefi, Rasoul, Nourani, Mehrdad
• Primary Examiner(s): Winakur, Eric
• Assistant Examiner(s): Fardanesh, Marjan
• Application Number:
  US14/595,866
• Publication Number:
  US 20150196257A1
• Time in Patent Office: 1,162 Days
• Field of Search: None
• 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 5/0816 Measuring devices for examining respiratory frequency measuring frequency of electric signals G01R23/00
• CPC Class Codes: A61B 5/14551 for measuring blood gases
• CPC Class Codes: A61B 5/7246 using correlation, e.g. template matching or determination of similarity

Abstract

A system and method for signal processing to remove unwanted noise components including: (i) wavelength-independent motion artifacts such as tissue, bone and skin effects, and (ii) wavelength-dependent motion artifact/noise components such as venous blood pulsation and movement due to various sources including muscle pump, respiratory pump and physical perturbation. Disclosed are methods, analytics, and their uses for reliable perfusion monitoring, arterial oxygen saturation monitoring, heart rate monitoring during daily activities and in hospital settings and for extraction of physiological parameters such as respiration information, hemodynamic parameters, venous capacity, and fluid responsiveness. The system and methods disclosed are extendable to include monitoring platforms for perfusion, hypoxia, arrhythmia detection, airway obstruction detection and sleep disorders including apnea.

Claims

1. A method for enhancing a set of photoplethysmographic signals to calculate a blood oxygen level and a heart rate comprising the steps of:

removing motion noise from the set of photoplethysmographic signals to generate a set of enhanced signals;

extracting a fundamental period from the set of enhanced signals;

determining a venous noise reference from the set of enhanced signals and the fundamental period;

generating a set of clean signals from the set of enhanced signals and the venous noise reference; and

,

calculating the heart rate and the blood oxygen level from the set of clean signals.

2. The method of claim 1, wherein the step of removing further comprises the steps of:

generating a reference signal from the set of photoplethysmographic signals; and

,

filtering motion noise from the set of photoplethysmographic signals using the reference signal.

3. The method of claim 2, wherein the step of filtering further comprises the steps of:

determining a first set of filter criteria;

generating a first filter coefficient from the first set of filter criteria;

applying the first filter coefficient to the reference signal to generate the set of enhanced signals;

generating a first error estimation from the set of enhanced signals and the set of photoplethysmographic signals; and

,

updating the first filter coefficient based on the first error estimation.

4. The method of claim 2, wherein the step of extracting a fundamental period further comprises the steps of:

calculating an autocorrelation function of the set of enhanced signals;

calculating a set of time domain windows;

generating a modified autocorrelation function for each of the set of time domain windows;

combining each modified autocorrelation function at a subset of the set of time domain windows;

determining a maximum energy level from the subset of the set of time domain windows; and

,

determining the fundamental period from the maximum energy level.

5. The method of claim 1, wherein the set of enhanced signals further comprises an enhanced red signal and an enhanced infrared signal, wherein the step of determining a venous noise reference further comprises the steps of:

extracting an arterial scaled estimate from the set of enhanced signals;

generating a weight based on the arterial scaled estimate;

generating a first error signal from the weight and the arterial scaled estimate;

generating an optimized weight from the weight by minimizing a variance of the first error signal using the fundamental period; and

,

generating the venous noise reference from the enhanced red signal, the enhanced infrared signal, and the optimized weight.

6. The method of claim 1, wherein the step of generating a set of clean signals further comprises the steps of:

determining a second set of filter criteria;

generating a second filter coefficient from the second set of filter criteria;

applying the second filter coefficient to the venous noise reference to generate an output;

generating a second error estimation and the set of clean signals from the output and the set of enhanced signals; and

,

updating the second filter coefficient based on the second error estimation.

7. The method of claim 1, wherein the step of removing further comprises the steps of:

generating an estimated fundamental frequency from the set of photoplethysmographic signals;

generating a reference noise signal from the estimated fundamental frequency and the set of photoplethysmographic signals;

generating an estimated motion noise from the reference noise signal; and

,

generating a second error signal and the set of enhanced signals from the set of photoplethysmographic signals and the estimated motion noise.

8. The method of claim 7, further comprising the step of adjusting a filter coefficient to minimize the second error signal.

9. A method for motion-tolerant processing of a set of photoplethysmographic signals to calculate a blood oxygen level comprising the steps of:

generating a set of processed signals from the set of photoplethysmographic signals;

generating a set of enhanced signals from the set of processed signals;

calculating the blood oxygen level from the set of enhanced signals;

wherein the step of generating a set of enhanced signals from the set of processed signals further comprises the steps of;

removing a set of baseline changes from the set of processed signals to generate a set of baseline signals;

normalizing the set of baseline signals to generate a set of normalized signals;

generating an intermediate reference signal from the set of normalized signals;

filtering the set of normalized signals with the intermediate reference signal to generate a set of filtered signals;

receiving a fundamental period; and

,

generating the set of enhanced signals from the set of filtered signals and the fundamental period.

10. The method of claim 9, further comprising the steps of:

generating a photodetector gain and an LED intensity from the set of baseline signals;

driving a set of LEDs with the LED intensity; and

,

amplifying a photodetector signal with the photodetector gain.

11. The method of claim 9, wherein the set of enhanced signals further comprise an enhanced red signal and an enhanced infrared signal, and wherein the step of calculating the blood oxygen level from the set of enhanced signals further comprises the steps of:

generating a time window from the enhanced red signal and the enhanced infrared signal based on a set of conditions;

generating a windowed red signal from the time window and the enhanced red signal;

generating a windowed infrared signal from the time window and the enhanced infrared signal;

extracting the fundamental period from the windowed infrared signal;

generating a clean infrared signal from the windowed infrared signal and the fundamental period;

generating a clean red signal from the windowed red signal and the fundamental period;

calculating an optimum ratio from the clean infrared signal and the clean red signal; and

,

calculating the blood oxygen level from the optimum ratio.

12. The method of claim 11, wherein the step of extracting the fundamental period from the windowed infrared signal further comprises the steps of:

segmenting the windowed infrared signal into a set of segmented windowed infrared signals;

calculating an autocorrelation function for each of the set of segmented windowed infrared signals to generate a set of autocorrelation functions;

determining a peak for each of the set of autocorrelation functions;

removing a doubled frequency from the set of segmented windowed infrared signals using the peak to generate a modified signal; and

,

determining a maximum peak from the modified signal to generate the fundamental frequency.

13. The method of claim 11, wherein the step of calculating an optimum ratio further comprises the steps of:

calculating an estimated infrared signal from the clean infrared signal;

calculating an error estimation signal from the clean red signal;

calculating a ratio vector from the error estimation signal;

calculating a mean from the ratio vector; and

,

calculating the optimum ratio from the mean.

14. A system for enhancing a set of photoplethysmographic signals to calculate a blood oxygen level and a heart rate comprising:

a evaluation module;

a set of photoplethysmographic sensors, each photoplethysmographic sensor connected to the module;

a computing device connected to the module and programmed to carry out the steps of;

removing motion noise from the set of photoplethysmographic signals to generate a set of enhanced signals;

extracting a fundamental period from the set of enhanced signals;

determining a venous noise reference from the set of enhanced signals and the fundamental period;

generating a set of clean signals from the set of enhanced signals and the venous noise reference; and

,

calculating the heart rate and the blood oxygen level from the set of clean signals.

15. The system of claim 14, wherein the computing device is further programmed to carry out the steps of:

generating a reference signal from the set of photoplethysmographic signals; and

,

filtering motion noise from the set of photoplethysmographic signals using the reference signal.

16. The system of claim 15, wherein the computing device is further programmed to carry out the steps of:

determining a first set of filter criteria;

generating a first filter coefficient from the first set of filter criteria;

applying the first filter coefficient to the reference signal to generate the set of enhanced signals;

generating a first error estimation from the set of enhanced signals and the set of photoplethysmographic signals; and

,

updating the first filter coefficient based on the first error estimation.

17. The system of claim 15, wherein the computing device is further programmed to carry out the steps of:

calculating an autocorrelation function of the set of enhanced signals;

calculating a set of time domain windows;

generating a modified autocorrelation function for each of the set of time domain windows;

combining each modified autocorrelation function at a subset of the set of time domain windows;

determining a maximum energy level from the subset of the set of time domain windows; and

,

determining the fundamental period from the maximum energy level.

18. The system of claim 14, wherein the set of enhanced signals further comprises an enhanced red signal and an enhanced infrared signal, and wherein the computing device is further programmed to carry out the steps of:

extracting an arterial scaled estimate from the set of enhanced signals;

generating a weight based on the arterial scaled estimate;

generating a first error signal from the weight and the arterial scaled estimate;

generating an optimized weight from the weight by minimizing a variance of the first error signal using the fundamental period; and

,

generating the venous noise reference from the enhanced red signal, the enhanced infrared signal, and the optimized weight.

19. The system of claim 14, wherein the computing device is further programmed to carry out the steps of:

determining a second set of filter criteria;

generating a second filter coefficient from the second set of filter criteria;

applying the second filter coefficient to the venous noise reference to generate an output;

generating a second error estimation and the set of clean signals from the output and the set of enhanced signals; and

,

updating the second filter coefficient based on the second error estimation.

20. The system of claim 14, wherein the computing device is further programmed to carry out the steps of:

generating an estimated fundamental frequency from the set of photoplethysmographic signals;

generating a reference noise signal from the estimated fundamental frequency and the set of photoplethysmographic signals;

generating an estimated motion noise from the reference noise signal; and

,

generating a second error signal and the set of enhanced signals from the set of photoplethysmographic signals and the estimated motion noise.

21. The system of claim 20, wherein the computing device is further programmed to carry out the step of adjusting a filter coefficient to minimize the second error signal.