| Summary: | <p>This document describes the data used to obtain the plots and results shown in the publication:</p>
<p>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).</p>
<br/>
<p>This directory contains the following files:</p>
<br/>
sophia_subject1.csv<br/>
sophia_subject2.csv<br/>
sophia_subject3.csv<br/>
sophia_subject4.csv<br/>
sophia_subject5.csv<br/>
<br/>
sophia_badata_fit.csv<br/>
sophia_badata_loo.csv<br/>
<br/>
sophia_roi_subject3.csv<br/>
<br/>
r2_comparisons.csv<br/><br/>
************************** SOPHIA DATA *****************************************<br/><br/>
Used in Figures 3,4<br/>
Used in Tables 2,3 <br/><br/>
<p>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:</p>
sophia_subjectN.csv<br/>
<br/>
Where N represents the subject number.<br/>
<br/>
The rows in the csv files represent the data points, and the columns have the
following headings:<br/>
<br/>
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 [%]
<br/><br/>
************************ FIGURES 2 & 6 *****************************************<br/><br/>
The Bland-Altman plots may instead be taken from:<br/>
<br/>
sophia_badata_fit.csv<br/>
Corresponding to the Bland-Altman of the camera estimates using Sophia
linearly regressed onto the oxygen saturation reference<br/>
sophia_badata_loo.csv<br/>
Corresponding to the Bland-Altman of the camera estimates using Sophia as
inferred from the gradients of the other fits (leave-one-out method).<br/>
<br/>
In both cases the data is arranged such that the rows represent the data points and the columns have the following headings:<br/><br/>
Mean(Estimate,Reference) Spo2 [%], Estimate SpO2 - Reference SpO2 [%]<br/><br/>
*************************** FIGURE 5 *******************************************<br/><br/>
The data with which the comparisons were made may be generated using:<br/>
<br/>
r2_comparison.csv<br/>
<br/>
The comparison was made against ROI selection methods described in:<br/>
<br/>
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)
<br/><br/>
U. Bal, ``Non-contact estimation of heart rate and oxygen saturation using
ambient light,'', Biomed. Opt. Express 6(10), 86-97 (2015)
<br/><br/>
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)
<br/><br/>
and against three separate steps in Sophia, as described in Equation 13 of the associated publication.
<br/><br/>
The data in r2_comparison.csv is arranged as follows:
<br/><br/>
The rows correspond to the subject number, from 1 to 5.<br/>
The columns correspond to the coefficient of determination (r^2, [0,1]) found
when applying each of the following methods:<br/>
Kong (2013), Tarassenko (2014), Bal (2015), Sophia (no threshold), Sophia (HR threshold only), Sophia (HR and BR thresholds), Sophia (all inclusion criteria)<br/><br/>
**************************** FIGURE 7 ******************************************
<br/><br/>
The graph showing the effect of increasing the minimum number of windows used in the estimation of oxygen saturation is instead generated by:
<br/><br/>
sophia_roi_subject3.csv<br/><br/>
The columns correspond to data points (from 0 through to 30), and the rows
represent:
<br/><br/>
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).
|
|---|