Techniques for monitoring and detecting respiration
Inventors
Lympouridis, Vangelis • Sackman, Josh • Espinoza, Eric • Robin, Aaron • Graves, Jonathan
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11073900-B2
Publication Date
2021-07-27
Expiration Date
Abstract
A device for monitoring respiration by a user determines whether sound represented by audio data corresponds to an exhalation. For example, the device receives audio data representing sound, determines a frequency spectrum for the audio data (e.g., amplitude distributions of one or more frequencies, etc.), and identifies various frequency bands of interest. Here, the device identifies a first frequency band corresponding to respiration, a second frequency band corresponding to noise, and a third frequency band corresponding to voice harmonics. The device applies a gain to the first frequency band to create an amplified first frequency band, filters or nulls the second frequency band, and inverts the third frequency band to create an inverted third frequency band. The device determines the frequency spectrum corresponds to an exhalation based on a slope value over the amplitude distributions of the amplified first frequency band and the inverted third frequency band.
Core Innovation
The invention provides a method for monitoring respiration that receives audio data representing sound detected by an input interface on a device. A processor determines a frequency spectrum of amplitude distributions for one or more frequencies represented by the audio data and identifies a first frequency band, a second frequency band, and a third frequency band from the frequency spectrum of amplitude distributions. The first frequency band includes frequencies associated with respiration, the second frequency band includes frequencies associated with noise, and the third frequency band includes frequencies associated with voice harmonics.
The method applies gain to the first set of frequencies to create an amplified first frequency band, applies a filter to the second set of frequencies to create a nulled second frequency band, and inverts the third set of frequencies of the third frequency band to create an inverted third frequency band. The method determines that the frequency spectrum corresponds to an exhalation based on a slope value between amplitude distributions of the amplified first frequency band and amplitude distributions of at least a portion of the inverted third frequency band.
The method then modifies an experience presented to a user of the device based on the exhalation. In additional implementations, the slope value is determined using a moving average or is used with a threshold value, and respiration can be determined over a time period to derive a respiration rate.
Claims Coverage
The independent claims are clm-00001, clm-00014, and clm-00018. Across these claims, the inventive scheme centers on band-based spectral processing of audio data using gain, filtering, and inversion, followed by exhalation determination via a slope value, and then modifying a user experience based on the exhalation.
Band-based spectral separation of respiration, noise, and voice harmonics
A processor determines a frequency spectrum of amplitude distributions and identifies a first frequency band, a second frequency band, and a third frequency band, where the first band includes frequencies associated with respiration, the second band includes frequencies associated with noise, and the third band includes frequencies associated with voice harmonics.
Gain, nulling filter, and inversion processing of the identified bands
The processor applies gain to the first set of frequencies to create an amplified first frequency band, applies a filter to the second set of frequencies to create a nulled second frequency band, and inverts the third set of frequencies of the third frequency band to create an inverted third frequency band.
Slope-based exhalation determination from processed amplitude distributions
The processor determines that the frequency spectrum corresponds to an exhalation based on a slope value between amplitude distributions of the amplified first frequency band and amplitude distributions of at least a portion of the inverted third frequency band.
Modify a user experience based on exhalation
The processor modifies an experience presented to a user of the device based on the exhalation.
Across the independent claims, respiration monitoring is implemented by receiving audio data, computing a frequency spectrum of amplitude distributions, partitioning into respiration, noise, and voice-harmonics bands, applying gain, filtering, and inversion, determining exhalation using a slope value between amplitude distributions, and modifying an experience presented to the user based on the exhalation.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?