This document describes the data used to obtain the plots and results shown in 
the publication:

A. R. Guazzi, M. Villarroel, J. Jorge, J. Daly, M. C. Frise, P. A. Robbins, 
L. Tarassenko, ``Non-contact measurement of oxygen saturation with an RGB
camera,'' Biomedical Optics Express (submitted in 2015).


This directory contains the following files:

sophia_subject1.csv
sophia_subject2.csv
sophia_subject3.csv
sophia_subject4.csv
sophia_subject5.csv

sophia_badata_fit.csv
sophia_badata_loo.csv

sophia_roi_subject3.csv

r2_comparisons.csv

************************** SOPHIA DATA *****************************************
Used in Figures 3,4
Used in Tables  2,3 

The camera estimates using the Sophia algorithm described in the paper, and the 
references used for these estimates may be found, per subject, in the files:

sophia_subjectN.csv

Where N represents the subject number.

The rows in the csv files represent the data points, and the columns have the 
following headings:

Resampled Time [s], Camera Estimated HR [bpm], Camera Estimated R (proportional 
to SpO2) [a.u.], Left Reference HR [bpm], Left Reference Spo2 [%], Right 
Reference HR [bpm], Right Reference Spo2 [%]

************************ FIGURES 2 & 6 *****************************************

The Bland-Altman plots may instead be taken from:

sophia_badata_fit.csv
  Corresponding to the Bland-Altman of the camera estimates using Sophia 
  linearly regressed onto the oxygen saturation reference
sophia_badata_loo.csv
  Corresponding to the Bland-Altman of the camera estimates using Sophia as 
  inferred from the gradients of the other fits (leave-one-out method).
  
In both cases the data is arranged such that the rows represent the data points 
and the columns have the following headings:

Mean(Estimate,Reference) Spo2 [%], Estimate SpO2 - Reference SpO2 [%]

*************************** FIGURE 5 *******************************************

The data with which the comparisons were made may be generated using:

r2_comparison.csv

The comparison was made against ROI selection methods described in:

    L. Kong, Y. Zhao, Y. Dong, Y. Jian, X. Jin, B. Li, Y. Feng, M. Liu, X. Liu, 
    H. Wu, ``Non-contact detection of oxygen saturation based on visible light 
    imaging device using ambient light,'' Opt. Express 21(15), 1764--1771 (2013)
    
    U. Bal, ``Non-contact estimation of heart rate and oxygen saturation using 
    ambient light,'', Biomed. Opt. Express 6(10), 86-97 (2015)
    
    L. Tarassenko, M. Villarroel, A. Guazzi, J. Jorge, D. A. Clifton, C. Pugh, 
    ``Non-contact video-based vital sign monitoring using ambient light and 
    auto-regressive models,'' Physiol. Meas. 35(5), 805--831 (2014)

and against three separate steps in Sophia, as described in Equation 13 of the 
associated publication.

The data in r2_comparison.csv is arranged as follows:

The rows correspond to the subject number, from 1 to 5.
The columns correspond to the coefficient of determination (r^2, [0,1]) found
when applying each of the following methods:
Kong (2013), Tarassenko (2014), Bal (2015), Sophia (no threshold), Sophia (HR 
threshold only), Sophia (HR and BR thresholds), Sophia (all inclusion criteria)

**************************** FIGURE 7 ******************************************

The graph showing the effect of increasing the minimum number of windows used in
the estimation of oxygen saturation is instead generated by:

sophia_roi_subject3.csv

The columns correspond to data points (from 0 through to 30), and the rows 
represent:

r^2 [0 to 1] (coefficient of determination)
SpO2 MAE [%] (mean absolute error between camera estimate of SpO2 and mean 
reference Spo2) Windows Rejected [%] (percentage of windows rejected as a 
result of this condition)
