Systems and methods for autocorrelation based assessment of ppg signal quality

Abstract

Systems and methods for assessing PPG signals generated based on transdermal optical data can include a computing device generating a color intensity signal using an acquired sequence of transdermal images of a subject. The computing device can compute a signal quality metric of the color intensity signal. The computing device can provide an indication of a quality of the color intensity signal for display on a display device associated with the computing device. The indication of the quality of the color intensity signal can be determined based on the signal quality metric.

What is claimed is: 
     
         1 . A computing device comprising:
 a processor; and   a memory storing computer code instructions, the computer code instructions when executed by the processor cause the computing device to:
 obtain one or more photoplethysmographic (PPG) signals generated from a sequence of images obtained from a photodetector; 
 extract a plurality of features of the PPG signal; and 
 generate a blood pressure classification using the plurality of features of the PPG signal and a blood pressure classification model. 
   
     
     
         2 . The computing device of  claim 1 , wherein the blood pressure classification model is a machine learning model associated with a corresponding set of parameter variables used to determine the blood pressure classification. 
     
     
         3 . The computing device of  claim 2 , wherein the computer code instructions when executed by the processor further cause the computing device to train the machine learning model using labeled data to determine the corresponding set of parameter variables. 
     
     
         4 . The computing device of  claim 1 , wherein the plurality of features of the PPG signal includes one or more signal features extracted from a logarithmic PPG signal generated from the PPG signal. 
     
     
         5 . The computing device of  claim 1 , wherein the plurality of features of the PPG signal includes one or more pulse related features extracted from pulses of the PPG signal. 
     
     
         6 . The computing device of  claim 1 , wherein the plurality of features of the PPG signal includes at least one of:
 a first systolic blood pressure estimate generated using the PPG signal as an input signal;   a first diastolic blood pressure estimate generated using the PPG signal as an input signal;   a second systolic blood pressure estimate generated using a logarithmic PPG signal as an input signal, the logarithmic PPG signal generated from the PPG signal; or   a second diastolic blood pressure estimate generated using the logarithmic PPG signal as an input signal.   
     
     
         7 . The computing device of  claim 1 , wherein the blood pressure classification is further generated using one or more demographic features of a user of the computing device. 
     
     
         8 . A method comprising:
 obtaining, by a computing device, one or more photoplethysmographic (PPG) signals generated from a sequence of images obtained from a photodetector;   extracting, by the computing device, a plurality of features of the PPG signal; and   generating, by the computing device, a blood pressure classification using the plurality of features of the PPG signal and a blood pressure classification model.   
     
     
         9 . The method of  claim 8 , wherein the blood pressure classification model is a machine learning model associated with a corresponding set of parameter variables used to determine the blood pressure classification. 
     
     
         10 . The method of  claim 9 , further comprising training the machine learning model using labeled data to determine the corresponding set of parameter variables. 
     
     
         11 . The method of  claim 8 , wherein the plurality of features of the PPG signal includes one or more signal features extracted from a logarithmic PPG signal generated from the PPG signal. 
     
     
         12 . The method of  claim 8 , wherein the plurality of features of the PPG signal includes one or more pulse related features extracted from pulses of the PPG signal. 
     
     
         13 . The method of  claim 8 , wherein the plurality of features of the PPG signal includes at least one of:
 a first systolic blood pressure estimate generated using the PPG signal as an input signal;   a first diastolic blood pressure estimate generated using the PPG signal as an input signal;   a second systolic blood pressure estimate generated using a logarithmic PPG signal as an input signal, the logarithmic PPG signal generated from the PPG signal; or   a second diastolic blood pressure estimate generated using the logarithmic PPG signal as an input signal.   
     
     
         14 . The method of  claim 8 , wherein the blood pressure classification is further generated using one or more demographic features of a user of the computing device. 
     
     
         15 . A computer readable medium including computer code instructions stored thereon, the computer code instructions when executed cause one or more processors to:
 obtain one or more photoplethysmographic (PPG) signals generated from a sequence of images obtained from a photodetector;   extract a plurality of features of the PPG signal; and   generate a blood pressure classification using the plurality of features of the PPG signal and a blood pressure classification model.   
     
     
         16 . The computer readable medium of  claim 15 , wherein the blood pressure classification model is a machine learning model associated with a corresponding set of parameter variables used to determine the blood pressure classification. 
     
     
         17 . The computer readable medium of  claim 16 , wherein the computer code instructions when executed further cause the one or more processors to train the machine learning model using labeled data to determine the corresponding set of parameter variables. 
     
     
         18 . The computer readable medium of  claim 15 , wherein the plurality of features of the PPG signal includes one or more signal features extracted from a logarithmic PPG signal generated from the PPG signal. 
     
     
         19 . The computer readable medium of  claim 15 , wherein the plurality of features of the PPG signal includes one or more pulse related features extracted from pulses of the PPG signal. 
     
     
         20 . The computer readable medium of  claim 15 , wherein the plurality of features of the PPG signal includes at least one of:
 a first systolic blood pressure estimate generated using the PPG signal as an input signal;   a first diastolic blood pressure estimate generated using the PPG signal as an input signal;   a second systolic blood pressure estimate generated using a logarithmic PPG signal as an input signal, the logarithmic PPG signal generated from the PPG signal; or   a second diastolic blood pressure estimate generated using the logarithmic PPG signal as an input signal.