System and method for determining data quality for cardiovascular parameter determination

Abstract

The system for cardiovascular parameter data quality determination can include a user device and a computing system, wherein the user device can include one or more sensors, the computing system, and/or any suitable components. The computing system can optionally include a data quality module, a cardiovascular parameter module, a storage module, and/or any suitable modules. The method for cardiovascular parameter data quality determination can include acquiring data and determining a quality of the data. The method can optionally include processing the data, and/or determining a cardiovascular parameter, training a data quality module, any suitable steps.

We claim: 
     
         1 . A system, comprising:
 a processor configured to:
 receive a set of images of a body region of a user, the set of images sampled by an image sensor; 
 using a first model, determine a first data quality classification based on a first set of attributes extracted from the set of images, wherein the first data quality classification comprises a binary classification for detected contact or lack thereof between the body region and the image sensor; 
 using a second model, determine a second data quality classification based on a second set of attributes extracted from the set of images, wherein the second data quality classification comprises a multiclass classification for placement of the body region relative to the image sensor; 
 generate a high-quality plethysmogram (PG) dataset from the set of images based on the first data quality classification and the second data quality classification; and 
 determine a cardiovascular parameter based on the high-quality PG data. 
   
     
     
         2 . The system of  claim 1 , wherein the first set of attributes comprises a total luminance of each image in the set of images, and wherein the second set of attributes comprises summed luminance values across each row and across each column of each image in the set of images. 
     
     
         3 . The system of  claim 2 , wherein the first set of attributes further comprises a total red chroma and a total blue chroma of each image in the set of images. 
     
     
         4 . The system of  claim 1 , wherein the multiclass classification comprises: a first set of classes associated with location of the body region relative to the image sensor; and a second set of classes associated with contact pressure between the body region and the image sensor. 
     
     
         5 . The system of  claim 4 , wherein the first set of classes comprises at least: proper placement, improper placement associated with a first direction, and improper placement associated with a second direction. 
     
     
         6 . The system of  claim 1 , wherein the second model comprises a machine learning model trained using sets of training images, wherein each set of training images corresponds to a time window, wherein at least a portion of the time windows comprise overlapping time windows. 
     
     
         7 . The system of  claim 1 , wherein the first data quality classification and the second data quality classification are determined contemporaneously. 
     
     
         8 . The system of  claim 1 , wherein generating the high-quality PG dataset comprises: combining the first data quality classification and the second data quality classification to determine an overall data quality; and generating the high-quality PG dataset from the set of images based on the overall data quality. 
     
     
         9 . The system of  claim 1 , wherein a user device is configured to guide the user based on the second data quality classification to adjust at least one of: a motion of the body region, a placement of the body region relative to the image sensor, or a contact pressure between the body region and the image sensor. 
     
     
         10 . The system of  claim 1 , wherein the cardiovascular parameter is displayed at a user device. 
     
     
         11 . A method, comprising:
 sampling a set of images of a body region of a user;   using a trained machine learning model, determining a data quality based on attributes extracted from the set of images;   generating a high-quality plethysmogram (PG) dataset from the set of images based on the data quality;   determining a fit parameter of a fiducial model by fitting at least one of a first derivative, second derivative, or third derivative of the fiducial model to a corresponding derivative of the high-quality PG dataset; and   determining a cardiovascular parameter based on the fit parameter.   
     
     
         12 . The method of  claim 11 , wherein the fit parameter comprises a timing fit parameter, wherein the timing fit parameter is determined by fitting at least one of the first derivative or the second derivative of the fiducial model to the high-quality PG dataset. 
     
     
         13 . The method of  claim 12 , wherein the timing fit parameter is associated with a location of the fiducial model. 
     
     
         14 . The method of  claim 11 , wherein the fit parameter comprises an amplitude fit parameter, wherein the amplitude fit parameter is determined by fitting the third derivative of the fiducial model to the high-quality PG dataset. 
     
     
         15 . The method of  claim 11 , wherein the fiducial model is one of a set of fiducial models, wherein determining the cardiovascular parameter comprises:
 determining a set of fiducials based on a fit parameter for each of the set of fiducial models; and   calculating a synthetic fiducial based on the set of fiducials, wherein the cardiovascular parameter is determined based on a linear relationship between the synthetic fiducial and the cardiovascular parameter.   
     
     
         16 . The method of  claim 15 , wherein the synthetic fiducial comprises a linear combination of the set of fiducials. 
     
     
         17 . The method of  claim 11 , wherein the fiducial model comprises a combination of radial basis functions. 
     
     
         18 . The method of  claim 11 , wherein determining the data quality comprises:
 determining a contact data quality based on a first set of attributes extracted from the set of images;   determining a body region placement data quality based on a second set of attributes extracted from the set of images; and   determining the data quality based on the contact data quality and the body region placement data quality.   
     
     
         19 . The method of  claim 18 , wherein the body region placement data quality comprises a multiclass classification for a location of the body region in at least two dimensions. 
     
     
         20 . The method of  claim 11 , wherein the cardiovascular parameter comprises at least one of a blood pressure or a heart rate.