Online Supplement A

ONLINE SUPPLEMENT A

1. The Mel-scale spectrum is basically the Discrete Cosine Transform (DCT) of the logarithmic energy spectrum of a signal, where the spectral energies are calculated using logarithmically spaced filters with increasing bandwidth (mel-filters). In more details, the mel- frequency ceptral coefficients, MFCCs, were derived as follows: First, the spectral representation of each 2-sec segment was passed through a bank of band-pass filters. The filters had a triangular-shape frequency response. Their central frequencies (CF) were linearly spaced in the frequency axis for low frequencies but logarithmically spaced for high frequencies, forming the so-called mel-scale filter bank. The logarithm of the resulting mel-spectrum is calculated and transformed via the DCT, yielding an MFCC sequence for each corresponding CF 1.
2. The adjusted R-square (coefficient of determination) is defined as, where n is the number of responses and d the degrees of the polynomial. measures the total sum of squares: , with the mean of the observed data. measures the residual sum of squares:. Note that SSTotal and SSResid are indicators of the sample variances of the dependent variable and the estimated residuals respectively. Adjusted R-square is a refinement of the R-square statistic accounting for the degrees of freedom, and thus making model fits comparable.
3. Point-wise (or non-simultaneous) prediction boundsmeasure the confidence that a new observation will lie within the bound interval, given the predictor values. These bounds are given by where s2 is the mean squared error, t is the inverse of Student's-t cumulative distribution with respect to the corresponding confidence level, and S is the covariance matrix,(, of the estimates .

REFERENCES

1. Davis S, Mermelstein P. Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences [Internet]. IEEE; 1980. Available from:

ONLINE SUPPLEMENTB

E-Table 1.Linear Regression Line fits of extracted features (rows) with respect to individual subject‘s characteristic Age, Height and Weight (columns). The adjusted R-square is also shown.

AGE / HEIGHT / WEIGHT
Fittedregression line / R2a / Fittedregression line / R2a / Fittedregression line / R2a
RR / 30.449
− 1.953 ×age / 0.222 / 44.511
− 0.216 ×height / 0.263 / 35.462
− 0.770 ×weight / 0.243
HR / 123.525
− 5.076 ×age / 0.184 / 155.267
− 0.504 ×height / 0.175 / 136.919
− 2.031 ×weight / 0.208
MFCC1 / 4.637
+ 0.115 ×age / 0.140 / 3.881
+ 0.012 ×height / 0.143 / 4.459
+ 0.036 ×weight / 0.092
MFCC2 / −0.068
+ 0.088 ×age / 0.140 / −0.607
+ 0.009 ×height / 0.129 / −0.205
+ 0.027 ×weight / 0.093
MFCC3 / 0.991
+ 0.041 ×age / 0.081 / 0.719
+ 0.004 ×height / 0.085 / 0.936
+ 0.012 ×weight / 0.046
PW / 171.631
− 6.142 ×age / 0.024 / 213.992
− 0.657 ×height / 0.028 / 182.102
− 1.980 ×weight / 0.016
SL / −9.505
− 0.293 ×age / 0.175 / −7.468
− 0.032 ×height / 0.196 / −8.981
− 0.097 ×weight / 0.132
PR / 0.990
+ 0.002×age / 0.098 / 0.981
+ 0.000×height / 0.082 / 0.987
+ 0.000×weight / 0.064
PLN / 8586.491
+ 528.046 ×age / 0.164 / 4912.753
+ 56.866 ×height / 0.185 / 7756.135
+ 164.376 ×weight / 0.110

E-Table 2.Linear Regression Line fits of extracted features (rows) with respect to the combined subject‘s characteristic Age, Height and Weight (columns). The adjusted R−square is also shown.

AGE / HEIGHT / WEIGHT
Fitted regression line / R2a
RR / 44.974
+ 0.456 ×age− 0.196 ×height− 0.260 ×weight / 0.259
HR / 118.632
− 3.453 ×age + 0.277 ×height− 1.842 ×weight / 0.208
MFCC1 / 3.973
+ 0.050 ×age + 0.013 ×height− 0.024 ×weight / 0.142
MFCC2 / −0.244
+ 0.074 ×age + 0.004 ×height− 0.009 ×weight / 0.131
MFCC3 / 0.694
+ 0.015 ×age + 0.006 ×height− 0.013 ×weight / 0.085
PW / 220.637
− 0.670 ×age− 0.892 ×height + 1.226 ×weight / 0.016
SL / −7.109
− 0.013 ×age− 0.043 ×height + 0.050 ×weight / 0.192
PR / 0.991
+ 0.002 ×age− 0.000023×height− 0.000041×weight / 0.086
PLN / 3844.605
+ 30.715 ×age + 89.111 ×height− 144.062 ×weight / 0.190

ONLINE SUPPLEMENTC

E-Figure 1: Linear fit (solid line) for each feature (rows, y axis) with respect to patient characteristics (columns, × axis). Point-wise prediction bounds with 95% confidence level are also shown with dashed lines. Inset: R2a, the adjusted coefficient of determination of the quadratic fit; r, the linear correlation coefficient only if significant correlation was achieved.

E-Table 3Linear Regression Line fits of extracted features (rows) with respect to individual subject’s z-scores Weight-for-Height WHZ, Weight-for-Age WAZ, Height-for-Age HAZ and Body mass index-for-Age BAZ. The adjusted R-square is also shown.

WHZ / HAZ / WAZ / BAZ
Fittedregression line / R2a / Fittedregression line / R2a / Fittedregression line / R2a / Fittedregression line / R2a
RR / 26.362
0.571 ×WHZ / 0.015 / 26.266
+ 0.123 ×HAZ / −0.006 / 26.203
− 0.592 ×WAZ / 0.009 / 26.387
− 0.617 ×BAZ / 0.020
HR / 113.080
− 2.213 ×WHZ / 0.033 / 112.794
+ 0.717 ×HAZ / −0.004 / 112.472
− 2.051 ×WAZ / 0.016 / 113.164
− 2.349 ×BAZ / 0.041
MFCC1 / 4.894
− 0.033 ×WHZ / 0.007 / 4.884
− 0.006 ×HAZ / −0.006 / 4.884
− 0.042 ×WAZ / 0.007 / 4.894
− 0.032 ×BAZ / 0.006
MFCC2 / 0.127
− 0.017 ×WHZ / −0.001 / 0.117
− 0.016 ×HAZ / −0.003 / 0.122
− 0.029 ×WAZ / 0.005 / 0.127
− 0.016 ×BAZ / −0.001
MFCC3 / 1.083
− 0.017 ×WHZ / 0.011 / 1.078
− 0.002 ×HAZ / −0.006 / 1.078
− 0.021 ×WAZ / 0.010 / 1.083
− 0.017 ×BAZ / 0.011
PW / 157.639
+ 2.831 ×WHZ / 0.001 / 158.105
− 0.641 ×HAZ / −0.006 / 158.423
+ 2.818 ×WAZ / −0.002 / 157.529
+ 3.015 ×BAZ / 0.002
SL / −10.156
+ 0.066 ×WHZ / 0.004 / −10.138
+ 0.003 ×HAZ / −0.007 / −10.137
+ 0.078 ×WAZ / 0.003 / −10.155
+ 0.060 ×BAZ / 0.002
PR / 0.993
− 0.000 ×WHZ / −0.006 / 0.993
− 0.001 ×HAZ / 0.008 / 0.993
− 0.000 ×WAZ / 0.001 / 0.993
− 0.000 ×BAZ / −0.007
PLN / 9769.056
− 160.492 ×WHZ / 0.011 / 9732.245
+ 7.307 ×HAZ / −0.007 / 9724.069
− 177.698 ×WAZ / 0.008 / 9769.243
− 148.156 ×BAZ / 0.009

E-Table 4.Linear Regression Line fits of extracted features (rows) with respect to the combined subject’s z-scores Weight-for-Height (WHZ), Weight-for-Age (WAZ), Height-for-Age (HAZ) and Body mass index-for-Age (BAZ). The adjusted R-square is also shown.

Fittedregression line / R2a
RR / 26.236
+ 9.971 ×WHZ− 0.188 ×HAZ− 1.125 ×WAZ / 0.039
HR / 112.863
+ 27.695 ×WHZ− 5.253 ×HAZ + 5.220 ×WAZ / 0.061
MFCC1 / 4.888
− 0.003 ×WHZ + 0.353 ×HAZ− 0.609 ×WAZ / 0.002
MFCC2 / 0.118
+ 0.073 ×WHZ + 0.134 ×HAZ− 0.266 ×WAZ / −0.011
MFCC3 / 1.080
+ 0.076 ×WHZ + 0.036 ×HAZ− 0.081 ×WAZ / −0.005
PW / 158.082
− 40.475 ×WHZ + 7.350 ×HAZ− 6.515 ×WAZ / −0.009
SL / −10.151
+ 0.590 ×WHZ− 1.333 ×HAZ + 2.163 ×WAZ / 0.024
PR / 0.993
− 0.001 ×WHZ + 0.001 ×HAZ− 0.003 ×WAZ / −0.010
PLN / 9762.467
− 1027.470 ×WHZ + 2581.475 ×HAZ− 4195.872 ×WAZ / 0.034