Method and apparatus for removing motion artifacts from biomedical signals

• US 10,278,647 B2
• Filed: 06/09/2016
• Issued: 05/07/2019
• Est. Priority Date: 06/09/2015
• Status: Active Grant

• Current Assignee: University of Connecticut (State of Connecticut)
• Original Assignee: University of Connecticut (State of Connecticut)
• Inventors: Salehizadeh, Seyed M. A., Chon, Ki H., Noh, Yeonsik
• Primary Examiner(s): Gedeon, Brian T
• Application Number:
  US15/178,089
• Publication Number:
  US 20160361021A1
• Time in Patent Office: 1,062 Days
• Field of Search: None
• CPC Class Codes: A61B 2503/10 Athletes
• CPC Class Codes: A61B 2505/07 Home care
• CPC Class Codes: A61B 2505/09 Rehabilitation or training
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 5/0077 Devices for viewing the surface of the body, e.g. camera, magnifying lens
• CPC Class Codes: A61B 5/0205 Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
• CPC Class Codes: A61B 5/02405 Determining heart rate variability
• CPC Class Codes: A61B 5/02416 using photoplethysmograph signals, e.g. generated by infrared radiation A61B5/14552 takes precedence
• CPC Class Codes: A61B 5/02427 Details of sensor
• CPC Class Codes: A61B 5/02438 with portable devices, e.g. worn by the patient
• CPC Class Codes: A61B 5/0245 by using sensing means generating electric signals, i.e. ECG signals
• CPC Class Codes: A61B 5/1102 Ballistocardiography
• CPC Class Codes: A61B 5/14552 Details of sensors specially adapted therefor
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/361 Detecting fibrillation
• CPC Class Codes: A61B 5/363 Detecting tachycardia or bradycardia
• CPC Class Codes: A61B 5/6824 Arm or wrist
• CPC Class Codes: A61B 5/721 using a separate sensor to detect motion or using motion information derived from signals other than the physiological signal to be measured
• CPC Class Codes: A61B 5/7239 using differentiation including higher order derivatives
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 5/7253 : characterised by using transforms
• CPC Class Codes: A61B 5/7275 : Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor

Abstract

A method and corresponding apparatus employ a time-varying spectral analysis approach for reconstructing a heart-related signal that includes motion artifacts. The motion artifacts are produced by motion of a biomedical sensor relative to a sensing location. By comparing time-varying spectra of the heart-related signal and a motion signal, those frequency peaks resulting from the motion artifacts may be suppressed in a time-varying spectrum of the heart-related signal. The time-varying spectral analysis based approach enables the heart-related signal to be reconstructed with accuracy by suppressing the motion artifacts. Example applications for the method and corresponding apparatus include training aids (e.g., runners'"'"' heart-rate monitors) and hospital patient heart-rate monitors.

Claims

1. A method for reconstructing a heart-related signal output by a biomedical sensor, the method comprising:

reconstructing a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes pre-processing the heart-related signal to produce a pre-processed heart-related signal, pre-processing the motion signal to produce a pre-processed motion signal, and computing a first time-frequency spectrum (TFS) of the heart-related signal using the pre-processed heart-related signal and a second TFS of the motion signal using the pre-processed motion signal; and

outputting the reconstructed representation of the heart-related signal.

2. The method of claim 1, wherein the motion artifacts are suppressed in the reconstructed representation of the heart-related signal.

3. The method of claim 1, wherein:

in an event the classification indicates that the movement does not rise to a level causing the motion artifacts, the reconstructed representation is based on an average value of peak-to-peak intervals in the pre-processed heart-related signal.

4. The method of claim 1, wherein the pre-processing includes filtering the heart-related signal to produce a filtered heart-related signal, the filtered heart-related signal being the pre-processed heart-related signal.

5. The method of claim 4, wherein the pre-processing further includes down-sampling the filtered heart-related signal, the down-sampling being at a sampling rate less than an original sampling rate.

6. The method of claim 1, wherein the first TFS computed and the second TFS computed are 3-dimensional spectra, the first TFS computed and the second TFS computed each including a respective time-varying amplitude or power distribution with respect to time and frequency.

7. The method of claim 1, wherein the first TFS computed is a first time-varying power spectral density (PSD) and the second TFS computed is a second time-varying PSD.

8. The method of claim 1, wherein the reconstructed representation is a current reconstructed representation and wherein, based on the classification indicating that the movement does rise to a level causing the motion artifacts and is a pseudo-periodic movement or a periodic movement, the reconstructing includes:

determining whether at least one spectral peak is present having a corresponding frequency within a given frequency range at a point in time in the first TFS computed, the current reconstructed representation being reconstructed for the point in time; and

wherein, based on a determination that the at least one spectral peak is not present, the current reconstructed representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed.

9. The method of claim 8, wherein the given frequency range is 0.5 Hz to 3 Hz.

10. The method of claim 1, wherein the reconstructing includes:

retaining up to a pre-determined number of candidate spectral peaks located at a first point in time in the first TFS computed, the candidate spectral peaks retained based on having corresponding frequencies within a given frequency range and highest amplitude or power values relative to other spectral peaks located at the first point in time, the other spectral peaks having respective frequencies within the given frequency range; and

discarding each of the candidate spectral peaks retained that is associated with a same frequency as a dominant spectral peak located at a second point in time in the second TFS computed, the first point in time and the second point in time having same time values.

11. The method of claim 10, wherein the given frequency range is 0.5 Hz to 3 Hz.

12. The method of claim 10, wherein the representation is a current representation, wherein the reconstructed representation is a current reconstructed representation, and wherein, in an event the discarding discards each of the candidate spectral peaks retained, reconstructing the current representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed, the previous point in time being an earlier point in time with respect to the first point in time.

13. The method of claim 10, wherein the discarding is a first discarding and wherein the representation is a current representation and, wherein, in an event the first discarding results in at least one remaining candidate spectral peak retained, reconstructing the current representation further includes:

second discarding, wherein the second discarding discards remaining candidate spectral peaks, of the at least one remaining candidate spectral peak retained, based on whether a corresponding frequency of the at least one remaining candidate spectral peak retained is distanced by at least a frequency difference threshold from a prior reconstructed representation'"'"'s frequency, the prior reconstructed representation associated with a previous point in time in the first TFS computed, the previous point in time being an earlier point in time with respect to the first point in time.

14. The method of claim 13, wherein:

based on each of the at least one remaining candidate spectral peak retained having been discarded by the second discarding, reconstructing the current representation is based on the prior reconstructed representation; and

based on at least one last candidate spectral peak remaining, the at least one last candidate spectral peak remaining not discarded by the second discarding, reconstructing the current representation is based on a selected candidate spectral peak selected from amongst the at least one last candidate spectral peak remaining having a closest corresponding frequency to the prior reconstructed representation'"'"'s frequency relative to respective frequencies of each of the at least one last candidate spectral peak remaining.

15. The method of claim 10, wherein the dominant spectral peak has a largest power or amplitude value relative to other peaks located at the second point in time in the second TFS.

16. The method of claim 10, wherein the retaining and the discarding are based on:

the classification indicating that the movement is a pseudo-periodic movement or a periodic movement; and

determining that at least one spectral peak with a corresponding frequency in the given frequency range is present at the first point in time.

17. The method of claim 1, wherein the pre-processing includes filtering the heart-related signal and the motion signal to produce a filtered heart-related signal and a filtered motion signal, respectively, the filtered heart-related signal being the pre-processed heart-related signal and the filtered motion signal being the pre-processed motion signal.

18. The method of claim 17, wherein the pre-processing further includes down-sampling both the filtered heart-related signal and the filtered motion signal, the down-sampling being at a sampling rate less than an original sampling rate.

19. The method of claim 1, further including:

classifying the classification of the movement by comparing an amount of amplitude modulation in the second TFS computed to an amplitude modulation threshold, wherein the classification indicates whether the movement rises to a level of causing the motion artifacts based on a result of the comparing.

20. The method of claim 19, wherein the amplitude modulation threshold is dependent on a type of the motion detector sensor.

21. The method of claim 19, wherein, in an event the result indicates that the movement does rise to the level causing the motion artifacts, classifying the classification of the movement further includes determining whether the movement is either a pseudo-periodic movement or a periodic movement, versus a random movement.

22. The method of claim 21, wherein, in determining whether the movement is either the pseudo-periodic movement or the periodic movement, versus the random movement includes, the method further includes:

identifying a first, second, and third frequency associated, respectively, with a first, second, and third spectral peak in the second TFS at a point in time in the second TFS, the first, second, and third peaks having largest power or amplitude values relative to other peaks in the second TFS at the point in time, the first spectral peak being a largest spectral peak amongst the first, second, and third spectral peaks;

computing a first ratio of the second frequency identified to the first frequency identified and determining a first comparison result by comparing the first ratio computed to a first ratio value; and

computing a second ratio of the third frequency identified to the first frequency identified and determining a second comparison result by comparing the second ratio computed to a second ratio value, wherein the classification further indicates that the movement is either the pseudo-periodic movement or the periodic movement, versus the random movement, based on the first comparison result and the second comparison result.

23. The method of claim 22, wherein the first ratio value is 2 and the second ratio value is 3.

24. The method of claim 1, wherein, in an event the classification of the movement indicates random movement, the reconstructing includes:

computing an average value of a number of prior reconstructed representations outputted prior to the reconstructing, the number based on a duration of consecutive motion artifacts, detected prior to the reconstructing;

computing a bandpass filter cutoff frequency based on the average value computed;

filtering the pre-processed heart-related signal by applying the band pass filter computed to produce a filtered, pre-processed heart-related signal; and

wherein the reconstructed representation is based on an average peak-to-peak interval value of the filtered, pre-processed heart-related signal.

25. The method of claim 1, wherein the heart-related signal and the motion signal are produced, concurrently.

26. The method of claim 1, further comprising:

segmenting the heart-related signal into a plurality of windowed heart-related signal data segments and the motion signal into a corresponding plurality of windowed motion signal data segments; and

repeating the reconstructing and the outputting for each windowed heart-related signal data segment and each corresponding motion signal data segment.

27. The method of claim 26, wherein the time window is a value in a range from 2 to 32 seconds.

28. The method of claim 1, wherein the biomedical sensor is at least one of:

a photoplethysmogram (PPG) sensor, piezoelectric sensor, Light Emitting Diode (LED) based sensor, camera sensor, and pulse oximeter sensor, and wherein the motion sensor is an accelerometer.

29. The method of claim 1, wherein the biomedical sensor and the motion sensor are co-located.

30. The method of claim 1, further comprising employing the reconstructed representation to determine a heart rate estimate.

31. The method of claim 1, further comprising employing the reconstructed representation to determine an arterial oxygen saturation (SpO2) estimate.

32. The method of claim 1, further comprising employing the reconstructed representation to detect or predict a heart-related ailment, the heart-related ailment including at least one of a heart rate variability (HRV) condition, atrial fibrillation condition, congestive heart failure condition, and tachycardia condition.

33. The method of claim 1, wherein the reconstructing and the outputting are performed in real-time with respect to production of the heart-related signal and the motion signal.

34. The method of claim 1, wherein the reconstructing and the outputting are performed in non-real-time with respect to production of the heart-related signal and the motion signal.

35. An apparatus for reconstructing a heart-related signal output by a biomedical sensor, the apparatus comprising:

a reconstruction unit configured to reconstruct a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes pre-processing the heart-related signal to produce a pre-processed heart-related signal, pre-processing the motion signal to produce a pre-processed motion signal, and computing a first time-frequency spectrum (TFS) of the heart-related signal using the pre-processed heart-related signal and a second TFS of the motion signal using the pre-processed motion signal; and

an output unit configured to output the reconstructed representation of the heart-related signal.

36. The apparatus of claim 35, wherein the motion artifacts are suppressed in the reconstructed representation of the heart-related signal.

37. The apparatus of claim 35, wherein:

in an event the classification indicates that the movement does not rise to level causing the motion artifacts, the reconstructed representation is based on an average value of peak-to-peak intervals in the pre-processed heart-related signal.

38. The apparatus of claim 35, wherein the pre-processing includes filtering the heart-related signal to produce a filtered heart-related signal, the filtered heart-related signal being the pre-processed heart-related signal.

39. The apparatus of claim 38, wherein the pre-processing further includes down-sampling the filtered heart-related signal, the down-sampling being at a sampling rate less than an original sampling rate.

40. The apparatus of claim 35, wherein the first TFS computed and the second TFS computed are 3-dimensional spectra, the first TFS computed and the second TFS computed each including a respective time-varying amplitude or power distribution with respect to time and frequency.

41. The apparatus of claim 35, wherein the first TFS computed is a first time-varying power spectral density (PSD) and the second TFS computed is a second time-varying PSD.

42. The apparatus of claim 35, wherein the reconstructed representation is a current reconstructed representation and wherein, based on the classification indicating that the movement does rise to a level causing the motion artifacts and is a pseudo-periodic movement or a periodic movement, to reconstruct the representation includes:

determining whether at least one spectral peak is present having a corresponding frequency within a given frequency range at a point in time in the first TFS computed, the current reconstructed representation being reconstructed for the point in time; and

wherein, based on a determination that the at least one spectral peak is not present, the reconstructed current representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed.

43. The apparatus of claim 42, wherein the given frequency range is 0.5 Hz to 3 Hz.

44. The apparatus of claim 35, wherein to reconstruct the representation includes:

retaining up to a pre-determined number of candidate spectral peaks located at a first point in time in the first TFS computed, the candidate spectral peaks retained based on having corresponding frequencies within a given frequency range and highest amplitude or power values relative to other spectral peaks located at the first point in time, the other spectral peaks having respective frequencies within the given frequency range; and

discarding each of the candidate spectral peaks retained that is associated with a same frequency as a dominant spectral peak located at a second point in time in the second TFS computed, the first point in time and the second point in time having same time values.

45. The apparatus of claim 44, wherein the given frequency range is 0.5 Hz to 3 Hz.

46. The apparatus of claim 44, wherein the representation is a current representation, wherein the reconstructed representation is a current reconstructed representation, and wherein, in an event the discarding discards each of the candidate spectral peaks retained, reconstructing the current representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed, the previous point in time being an earlier point in time with respect to the first point in time.

47. The apparatus of claim 44, wherein the discarding is a first discarding and wherein the representation is a current representation and, wherein, in an event the first discarding results in at least one remaining candidate spectral peak retained, reconstructing the current representation further includes:

second discarding, wherein the second discarding discards remaining candidate spectral peaks, of the at least one remaining candidate spectral peak retained, based on whether a corresponding frequency of the at least one remaining candidate spectral peak retained is distanced by at least a frequency difference threshold from a prior reconstructed representation'"'"'s frequency, the prior reconstructed representation associated with a previous point in time in the first TFS computed, the previous point in time being an earlier point in time with respect to the first point in time.

48. The apparatus of claim 47, wherein:

based on each of the at least one remaining candidate spectral peak retained having been discarded by the second discarding, reconstructing the current representation is based on the prior reconstructed representation; and

based on at least one last candidate spectral peak remaining, the at least one last candidate spectral peak remaining not discarded by the second discarding, reconstructing the current representation is based on a selected candidate spectral peak selected from amongst the at least one last candidate spectral peak remaining having a closest corresponding frequency to the prior reconstructed representation'"'"'s frequency relative to respective frequencies of each of the at least one last candidate spectral peak remaining.

49. The apparatus of claim 44, wherein the dominant spectral peak has a largest power or amplitude value relative to other peaks located at the second point in time in the second TFS.

50. The apparatus of claim 44, wherein the retaining and the discarding are based on:

the classification of the movement indicating that the movement is a pseudo-periodic movement or a periodic movement; and

determining that at least one spectral peak with a corresponding frequency in the given frequency range is present at the first point in time.

51. The apparatus of claim 35, wherein the pre-processing includes filtering the heart-related signal and the motion signal to produce a filtered heart-related signal and a filtered motion signal, respectively, the filtered heart-related signal being the pre-processed heart-related signal and the filtered motion signal being the pre-processed motion signal.

52. The apparatus of claim 51, wherein the pre-processing further includes down-sampling, at a sampling rate less than an original sampling rate, both the filtered heart-related signal and the filtered motion signal.

53. The apparatus of claim 35, further including a classification unit, the classification unit configured to:

classify the classification of the movement by comparing an amount of amplitude modulation in the second TFS computed to an amplitude modulation threshold, wherein the classification indicates whether the movement rises to a level causing the motion artifacts based on a result of the comparing.

54. The apparatus of claim 53, wherein the amplitude modulation threshold is dependent on a type of the motion detector sensor.

55. The apparatus of claim 53, wherein, in an event the result indicates that the movement does rise to the level causing the motion artifacts, the classification unit is further configured to classify the classification of the movement by determining whether the movement is either a pseudo-periodic movement or a periodic movement, versus a random movement.

56. The apparatus of claim 55, wherein to determine whether the movement is either the pseudo-periodic movement or the periodic movement versus a random movement includes:

identifying a first, second, and third frequency associated, respectively, with a first, second, and third spectral peak in the second TFS at a point in time in the second TFS, the first, second, and third peaks having largest power or amplitude values relative to other peaks in the second TFS at the point in time, the first spectral peak being a largest spectral peak amongst the first, second, and third spectral peaks;

computing a first ratio of the second frequency identified to the first frequency identified and determining a first comparison result by comparing the first ratio computed to a first ratio value; and

computing a second ratio of the third frequency identified to the first frequency identified and determining a second comparison result by comparing the second ratio computed to a second ratio value, wherein the classification further indicates that the movement is either the pseudo-periodic movement or the periodic movement, versus the random movement, based on the first comparison result and the second comparison result.

57. The apparatus of claim 56, wherein the first ratio value is 2 and the second ratio value is 3.

58. The apparatus of claim 55, wherein, in an event the classification of the movement indicates random movement, the reconstructing includes:

computing an average value of a number of prior reconstructed representations outputted prior to the reconstructing, the number based on a duration of consecutive motion artifacts, detected prior to the reconstructing;

computing a bandpass filter cutoff frequency based on the average value computed;

filtering the pre-processed heart-related signal by applying the band pass filter computed to produce a filtered, pre-processed heart-related signal; and

wherein the reconstructed representation is based on an average peak-to-peak interval value of the filtered, pre-processed heart-related signal.

59. The apparatus of claim 35, wherein the heart-related signal and the motion signal are produced, concurrently.

60. The apparatus of claim 35, further comprising segmentation unit configured to:

segment the heart-related signal into a plurality of windowed heart-related signal data segments and the motion signal into corresponding plurality of windowed motion signal data segments, wherein the reconstruction unit and the output unit reconstruct the representation and output the reconstructed representation, respectively, based on each windowed heart-related signal data segment and each corresponding motion signal data segment.

61. The apparatus of claim 60, wherein the time window is a value in a range from 2 to 32 seconds.

62. The apparatus of claim 35, wherein the biomedical sensor is at least one of:

a photoplethysmogram (PPG) sensor, piezoelectric sensor, Light Emitting Diode (LED) based sensor, camera sensor, and pulse oximeter sensor, and wherein the motion sensor is an accelerometer.

63. The apparatus of claim 35, wherein the apparatus includes the biomedical sensor and the motion sensor.

64. The apparatus of claim 35, further comprising a heart rate unit configured to employ the reconstructed representation to determine a heart rate estimate.

65. The apparatus of claim 35, further comprising an arterial oxygen saturation (SpO2) unit configured to employ the reconstructed representation to determine a SpO2 estimate.

66. The apparatus of claim 35, further comprising an ailment unit configured to employ the reconstructed representation to detect or predict a heart-related ailment, the heart-related ailment including at least one of a heart rate variability (HRV) condition, an atrial fibrillation condition, a congestive heart failure condition, and a tachycardia condition.

67. The apparatus of claim 35, wherein the reconstruction unit and the output unit are configured to reconstruct and to output, respectively, in real-time with respect to production of the heart-related signal and the motion signal.

68. The apparatus of claim 35, wherein the reconstruction unit and the output unit are configured to reconstruct and to output, respectively, in non-real-time with respect to production of the heart-related signal and the motion signal.

69. A non-transitory computer-readable medium having encoded thereon a sequence of instructions which, when loaded and executed by a processor, causes the processor to reconstruct a heart-related signal output by a biomedical sensor, the processor reconstructing the heart-related signal by:

reconstructing a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes pre-processing the heart-related signal to produce a pre-processed heart-related signal, pre-processing the motion signal to produce a pre-processed motion signal, and computing a first time-frequency spectrum (TFS) of the heart-related signal using the pre-processed heart-related signal and a second TFS of the motion signal using the pre-processed motion signal; and

outputting the reconstructed representation of the heart-related signal.

70. A method for reconstructing a heart-related signal output by a biomedical sensor, the method comprising:

reconstructing a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal; and

outputting the reconstructed representation of the heart-related signal, wherein the reconstructed representation is a current reconstructed representation and wherein, based on the classification indicating that the movement does rise to a level causing the motion artifacts and is a pseudo-periodic movement or a periodic movement, the reconstructing includes;

determining whether at least one spectral peak is present having a corresponding frequency within a given frequency range at a point in time in the first TFS computed, the current reconstructed representation being reconstructed for the point in time; and

wherein, based on a determination that the at least one spectral peak is not present, the current reconstructed representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed.

71. A method for reconstructing a heart-related signal output by a biomedical sensor, the method comprising:

reconstructing a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal; and

outputting the reconstructed representation of the heart-related signal, wherein the reconstructing includes;

retaining up to a pre-determined number of candidate spectral peaks located at a first point in time in the first TFS computed, the candidate spectral peaks retained based on having corresponding frequencies within a given frequency range and highest amplitude or power values relative to other spectral peaks located at the first point in time, the other spectral peaks having respective frequencies within the given frequency range; and

discarding each of the candidate spectral peaks retained that is associated with a same frequency as a dominant spectral peak located at a second point in time in the second TFS computed, the first point in time and the second point in time having same time values.

72. A method for reconstructing a heart-related signal output by a biomedical sensor, the method comprising:

reconstructing a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal;

classifying the classification of the movement by comparing an amount of amplitude modulation in the second TFS computed to an amplitude modulation threshold, wherein the classification indicates whether the movement rises to a level of causing the motion artifacts based on a result of the comparing; and

outputting the reconstructed representation of the heart-related signal.

73. An apparatus for reconstructing a heart-related signal output by a biomedical sensor, the apparatus comprising:

a reconstruction unit configured to reconstruct a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal; and

an output unit configured to output the reconstructed representation of the heart-related signal, wherein the reconstructed representation is a current reconstructed representation and wherein, based on the classification indicating that the movement does rise to a level causing the motion artifacts and is a pseudo-periodic movement or a periodic movement, to reconstruct the representation includes;

determining whether at least one spectral peak is present having a corresponding frequency within a given frequency range at a point in time in the first TFS computed, the current reconstructed representation being reconstructed for the point in time; and

wherein, based on a determination that the at least one spectral peak is not present, the reconstructed current representation is based on a prior reconstructed representation, the prior reconstructed representation associated with a previous point in time in the first TFS computed.

74. An apparatus for reconstructing a heart-related signal output by a biomedical sensor, the apparatus comprising:

a reconstruction unit configured to reconstruct a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal; and

an output unit configured to output the reconstructed representation of the heart-related signal, wherein to reconstruct the representation includes;

retaining up to a pre-determined number of candidate spectral peaks located at a first point in time in the first TFS computed, the candidate spectral peaks retained based on having corresponding frequencies within a given frequency range and highest amplitude or power values relative to other spectral peaks located at the first point in time, the other spectral peaks having respective frequencies within the given frequency range; and

discarding each of the candidate spectral peaks retained that is associated with a same frequency as a dominant spectral peak located at a second point in time in the second TFS computed, the first point in time and the second point in time having same time values.

75. An apparatus for reconstructing a heart-related signal output by a biomedical sensor, the apparatus comprising:

a reconstruction unit configured to reconstruct a representation of the heart-related signal to produce a reconstructed representation of the heart-related signal, the reconstructing based on (i) a time-varying spectral analysis of the heart-related signal and a motion signal, the motion signal output by a motion sensor and representative of motion artifacts in the heart-related signal, the motion artifacts being signal artifacts produced by movement of the biomedical sensor relative to a sensing location, and (ii) a classification of the movement, wherein the time-varying spectral analysis includes computing a first time-frequency spectrum (TFS) of the heart-related signal and computing a second TFS of the motion signal;

a classification unit configured to classify the classification of the movement by comparing an amount of amplitude modulation in the second TFS computed to an amplitude modulation threshold, wherein the classification indicates whether the movement rises to a level causing the motion artifacts based on a result of the comparing; and

an output unit configured to output the reconstructed representation of the heart-related signal.