Heart rate monitor with time varying linear filtering
A system, method, and device for monitoring a cardiac signal of a user includes an emitter (LED) for transmitting light toward skin of the user, a receiver (photodiode) for receiving a reflection of the transmitted light and generating a light intensity signal, and a processor configured to generate a photoplethysmogram (PPG) signal based on the light intensity signal. The PPG signal includes a cardiac component, a motion component and a respiratory component. An inertial sensor provides a motion signal to the processor based on sensed movement of the device and one or more time-variant filters are configured by the processor to filter the PPG signal to isolate the cardiac component of the PPG signal based on determined filter coefficients.
1. A fitness monitor for monitoring a cardiac signal of a user, the device comprising:
an emitter configured to transmit light toward skin of the user;
a receiver configured to receive a reflection of the transmitted light and provide a photoplethysmogram (PPG) signal based on the received reflection of the transmitted light, the PPG signal including a cardiac component and a motion component;
an inertial sensor configured to provide a motion signal based on sensed movement of the fitness monitor;
a time-variant filter configured to isolate a component of the PPG signal based on filter coefficients corresponding to a desired component of the PPG signal, and
a processor coupled to the emitter, the receiver, the inertial sensor and the time-variant filter, the processor configured to:
determine one or more frequencies associated with the motion signal;
transform the PPG signal to the frequency domain;
identify, in the frequency domain, the cardiac component in the PPG signal based on the determined one or more frequencies of the motion signal and the transformed PPG signal;
determine filter coefficients based on the identified cardiac component in the PPG signal;
configure the time-variant filter with the determined filter coefficients;
filter, in the time domain via the time-variant filter, the motion component from the PPG signal to isolate the identified cardiac component; and
calculate, in the time domain, the time between successive cycles of the isolated cardiac component of the PPG signal.
2. The fitness monitor of claim 1 , wherein the time-variant filter is integrated within the processor.
3. The fitness monitor of claim 1 , wherein the filter coefficients are determined upon each occurrence of the cardiac component being identified based on the determined one or more frequencies of the motion signal and the transformed PPG signal.
4. The fitness monitor of claim 3 , wherein the processor configures the time-variant filter with the determined filter coefficients upon each occurrence of the cardiac component being identified.
5. The fitness monitor of claim 1 , wherein the processor is further configured to periodically reconfigure the time-variant filter based on the cardiac component identified in the frequency domain.
6. The fitness monitor of claim 1 , further comprising a bandpass filter, wherein the processor is further configured to configure the bandpass filter to pre-condition the PPG signal by attenuating frequency components of the PPG signal below a predetermined frequency band that includes the cardiac component and/or attenuates frequency components above the predetermined frequency band that includes the cardiac component, the predetermined frequency band spanning the full range of human heart rates.
7. The fitness monitor of claim 1 , wherein the cardiac component is associated with heart rate, and wherein the determined time between successive cycles of the cardiac component is the time between successive heartbeats.
8. The fitness monitor of claim 1 , wherein the processor is further configured to analyze the cardiac component of the PPG signal to determine heart-rate variability, stress, recovery time, VO2 max, and/or sleep quality of the user, based on the determined time between successive cycles of the cardiac component.
9. The fitness monitor of claim 1 , further comprising a user interface coupled to the processor to display information based on the isolated cardiac component.
10. The fitness monitor of claim 1 , further comprising a housing to enclose the processor, the housing including a means for attaching the fitness monitor to the user.
11. The fitness monitor of claim 1 , wherein the determined filter coefficients include a center frequency corresponding to a fundamental frequency of the identified cardiac component of the PPG signal, and wherein the time-variant filter isolates the identified cardiac component by passing portions of the PPG signal associated with the center frequency.
12. The fitness monitor of claim 1 , wherein the processor calculates the time between successive cycles of the isolated cardiac component by identifying similar peaks, troughs, or threshold-crossings of the isolated cardiac component in the successive cycles.
13. The fitness monitor of claim 1 , wherein the processor identifies the cardiac component in the frequency domain by correlating one or more peaks of a motion signal with temporally consistent spectral peaks of the PPG signal.
14. A fitness monitor for monitoring a cardiac signal of a user, the device comprising:
an emitter configured to transmit light toward skin of the user;
a receiver configured to receive a reflection of the transmitted light and provide a photoplethysmogram (PPG) signal based on the received reflection of the transmitted light, the PPG signal including a cardiac component and a motion component;
an inertial sensor configured to provide a motion signal based on sensed movement of the fitness monitor;
a time-variant filter configured to isolate a component of the PPG signal based on filter coefficients corresponding to undesired components of the PPG signal, and
a processor coupled to the emitter, the receiver, the inertial sensor and the time-variant filter, the processor configured to:
determine one or more frequencies associated with the motion signal and the PPG signal provided during a first period,
identify the motion component in the PPG signal for the first time period based on the determined one or more frequencies of the motion signal and the PPG signal,
filter, via the time-variant filter, the identified motion component from the PPG signal provided during a second time period to isolate the cardiac component of the PPG signal for the second time period, and
calculate, in the time domain, the time between successive cycles of the isolated cardiac component of the PPG signal for the second time period.
15. The fitness monitor of claim 14 , wherein the first and second time periods correspond to the same period of time such that the filtered motion component is from the same instance of the PPG signal in which the motion component is identified.
16. The fitness monitor of claim 14 , wherein the time-variant filter is integrated within the processor.
17. The fitness monitor of claim 14 , wherein the processor is further configured to determine filter coefficients based on the identified motion component in the PPG signal for the first time period and configure the time-variant filter with the determined filter coefficients.
18. The fitness monitor of claim 14 , wherein the cardiac component of the PPG signal for the second time period is isolated upon each occurrence of the motion component being identified based on the determined one or more frequencies of the motion signal and the motion component.
19. The fitness monitor of claim 14 , further comprising a bandpass filter, wherein the processor is further configured to configure the bandpass filter to pre-condition the PPG signal by attenuating frequency components of the PPG signal below a predetermined frequency band that includes the cardiac component and/or attenuates frequency components above the predetermined frequency band that includes the cardiac component, the predetermined frequency band spanning the full range of human heart rates.
20. The fitness monitor of claim 14 , wherein the cardiac component is associated with heart rate, and wherein the determined time between successive cycles of the cardiac component is the time between successive heartbeats.
21. The fitness monitor of claim 14 , wherein the determined filter coefficients include a center frequency corresponding to a fundamental frequency of the identified motion component of the PPG signal for the first time period, and wherein the time-variant filter isolates the identified cardiac component during the second time period by not passing undesired components of the PPG signal associated with the center frequency.
22. The fitness monitor of claim 14 , wherein the processor calculates the time between successive cycles of the isolated cardiac component by identifying similar peaks, troughs, or threshold-crossings of the isolated cardiac component in the successive cycles.
23. The fitness monitor of claim 14 , wherein the processor identifies the motion component in the PPG signal for the first time period by correlating one or more peaks of a motion signal for the first time period with temporally consistent spectral peaks of the PPG signal.
24. A fitness monitor for monitoring a cardiac signal of a user, the device comprising:
an emitter configured to transmit light toward skin of the user;
a receiver configured to receive a reflection of the transmitted light and provide a photoplethysmogram (PPG) signal based on the received reflection of the transmitted light, the PPG signal including a cardiac component and a motion component;
an inertial sensor configured to provide a motion signal based on sensed movement of the fitness monitor;
a time-variant filter configured to isolate a component of the PPG signal based on filter coefficients corresponding to a desired component of the PPG signal, and
a processor coupled to the emitter, the receiver, the inertial sensor and the time-variant filter, the processor configured to:
determine one or more frequencies associated with the motion signal and the PPG signal provided during a first time period,
identify the cardiac component in the PPG signal for the first time period based on the determined one or more frequencies of the motion signal and the PPG signal,
filter, via the time-variant filter, the PPG signal provided for a second time to isolate the cardiac component of the PPG signal for the second time period based on the identified cardiac component, and
calculate, in the time domain, the time between successive cycles of the isolated cardiac component of the PPG signal for the second time period.
25. The fitness monitor of claim 24 , wherein the time-variant filter is integrated within the processor.
26. The fitness monitor of claim 24 , wherein the processor is further configured to determine filter coefficients based on the identified cardiac component in the PPG signal for the first time period and configure the time-variant filter with the determined filter coefficients.
27. The fitness monitor of claim 24 , wherein the determined filter coefficients include a center frequency corresponding to a fundamental frequency of the identified cardiac component of the PPG signal, and wherein the time-variant filter isolates the identified cardiac component by passing portions of the PPG signal associated with the center frequency.
28. The fitness monitor of claim 24 , wherein the processor calculates the time between successive cycles of the isolated cardiac component by identifying similar peaks, troughs, or threshold-crossings of the isolated cardiac component in the successive cycles.
29. The fitness monitor of claim 24 , wherein the processor identifies the cardiac component in the PPG signal for the first time period by correlating one or more peaks of a motion signal for the first time period with temporally consistent spectral peaks of the PPG signal.