System and method for anticipating the onset of an obstructive sleep apnea event

Abstract

Systems and methods for detecting a general breathing event and for anticipating an onset of an obstructive sleep apnea (OSA) event. The method for detecting a general breathing event includes receiving a plurality of signals from at least one microphone. The method also includes determining a one-sided power spectral density from the received signals. The method further includes distinguishing each received signal as either a breath signal or a background noise signal. The method still further includes calculating a breath signature by processing each breath signal.

1 . An method for detecting a general breathing event, the method comprising:
 receiving a plurality of signals from at least one sensing device;   determining a one-sided power spectral density from the received signals;   distinguishing each received signal as a signal associated with a breath or a signal associated with a background noise; and   calculating a breath signature by processing each signal associated with a breath.   
     
     
         2 . The method as set forth in  claim 1 , wherein receiving the plurality of signals from the at least one sensing device comprises:
 sampling raw sound signals from the at least one sensing device at a first rate;   applying a filter to the raw sound signals to reduce noise; and   resampling the raw sound signals from the at least one sensing device at a second rate, the second rate lower than the first rate.   
     
     
         3 . The method as set forth in  claim 1 , wherein determining the one-sided power spectral density from the received signals comprises:
 forming a data window using resampled sound signals, the data window comprising a plurality of data points;   multiplying the data window by a window function;   performing a Fourier transform on the data window; and   determining the one-side power spectral density from a result of the Fourier transform.   
     
     
         4 . The method as set forth in  claim 1 , wherein distinguishing each received signal comprises:
 determining a pedestal-adjusted auxiliary spectral density;   integrating the pedestal-adjusted auxiliary spectral density to obtain an auxiliary total power;   determining a filtered noise index using the auxiliary total power;   categorizing each received signal as a signal associated with a breath if the filtered noise index is greater than a threshold; and   categorizing each received signal as a signal associated with a background noise if the filtered noise index is less than the threshold.   
     
     
         5 . The method as set forth in  claim 1 , wherein the step of calculating a breath signature further comprises:
 calculating a pedestal-adjusted spectral density using the one-sided power spectral density;   examining the pedestal-adjusted spectral density over a frequency range; and   filtering the pedestal-adjusted spectral density using a low-pass filter.   
     
     
         6 . The method as set forth in  claim 1 , further comprising:
 for each signal associated with a breath, determining if the breath is a short breath; and   determining a peak signature for each breath that is not a short breath.   
     
     
         7 . A system for detecting a general breathing event, the system comprising:
 at least one sensing device; and   a processor in communication with the at least one sensing device, the processor configured to:
 receive a plurality of signals from the at least one sensing device; 
 determine a one-sided power spectral density from the received signals; 
 distinguish each received signal as a signal associated with a breath or a signal associated with a background noise; and 
 calculate a breath signature by processing each signal associated with a breath. 
   
     
     
         8 . The system as set forth in  claim 7 , wherein the processor is further configured to:
 sample raw sound signals from the at least one sensing device at a first rate;   apply a filter to the raw sound signals to reduce noise; and   resample the raw sound signals from the at least one sensing device at a second rate, the second rate lower than the first rate.   
     
     
         9 . The system as set forth in  claim 7 , wherein the processor is further configured to:
 form a data window using resampled sound signals, the data window comprising a plurality of data points;   multiply the data window by a window function;   perform a Fourier transform on the data window; and   determine the one-side power spectral density from a result of the Fourier transform.   
     
     
         10 . The system as set forth in  claim 7 , wherein the processor is further configured to:
 determine a pedestal-adjusted auxiliary spectral density;   integrate the pedestal-adjusted auxiliary spectral density to obtain an auxiliary total power;   determine a filtered noise index using the auxiliary total power;   categorize each received signal as a signal associated with a breath if the filtered noise index is greater than a threshold; and   categorizing each received signal as a signal associated with a background noise if the filtered noise index is less than the threshold.   
     
     
         11 . The system as set forth in  claim 7 , wherein the processor is further configured to:
 calculate a pedestal-adjusted spectral density using the one-sided power spectral density;   examine the pedestal-adjusted spectral density over a frequency range; and   filter the pedestal-adjusted spectral density using a low-pass filter.   
     
     
         12 . The system as set forth in  claim 7 , the processor further configured to:
 for each signal associated with a breath, determine if the breath is a short breath; and   determine a peak signature for each breath that is not a short breath.   
     
     
         13 .- 20 . (canceled) 
     
     
         21 . A non-transitory computer-readable storage medium encoded with executable instructions that when executed cause one or more processors to perform:
 receiving a plurality of signals from at least one sensing device;   determining a one-sided power spectral density from the received signals;   distinguishing each received signal as a signal associated with a breath or a signal associated with a background noise; and   calculating a breath signature by processing each signal associated with a breath.   
     
     
         22 . The non-transitory computer-readable storage medium as set forth in  claim 21 , wherein the executable instructions cause the one or more processors to perform receiving the plurality of signals from the at least one sensing device by:
 sampling raw sound signals from the at least one sensing device at a first rate;   applying a filter to the raw sound signals to reduce noise; and   resampling the raw sound signals from the at least one sensing device at a second rate, the second rate lower than the first rate.   
     
     
         23 . The non-transitory computer-readable storage medium as set forth in  claim 21 , wherein the executable instructions cause the one or more processors to perform determining the one-sided power spectral density from the received signals by:
 forming a data window using resampled sound signals, the data window comprising a plurality of data points;   multiplying the data window by a window function;   performing a Fourier transform on the data window; and   determining the one-side power spectral density from a result of the Fourier transform.   
     
     
         24 . The non-transitory computer-readable storage medium as set forth in  claim 21 , wherein the executable instructions cause the one or more processors to perform distinguishing each received signal by:
 determining a pedestal-adjusted auxiliary spectral density;   integrating the pedestal-adjusted auxiliary spectral density to obtain an auxiliary total power;   determining a filtered noise index using the auxiliary total power;   categorizing each received signal as a signal associated with a breath if the filtered noise index is greater than a threshold; and   categorizing each received signal as a signal associated with a background noise if the filtered noise index is less than the threshold.   
     
     
         25 . The non-transitory computer-readable storage medium as set forth in  claim 21 , wherein the executable instructions cause the one or more processors to perform calculating a breath signature by:
 calculating a pedestal-adjusted spectral density using the one-sided power spectral density;   examining the pedestal-adjusted spectral density over a frequency range; and   filtering the pedestal-adjusted spectral density using a low-pass filter.   
     
     
         26 . The non-transitory computer-readable storage medium as set forth in  claim 21 , wherein the executable instructions further cause the one or more processors to perform:
 for each signal associated with a breath, determining if the breath is a short breath; and   determining a peak signature for each breath that is not a short breath.   
     
     
         27 . The non-transitory computer-readable storage medium as set forth in  claim 26 , wherein the executable instructions further cause the one or more processors to perform:
 discarding each breath that is a short breath.   
     
     
         28 . The method as set forth in  claim 6 , further comprising:
 discarding each breath that is a short breath.