Table 1.

Methods for the study of heart rate variability^(7,8,9)

1. Linear methods - time domain

a. Statistical measures

i. SDNN - Standard deviation of all normal NN intervals

ii. SDANN - Standard deviation of the average normal NN interval calculated over 5-minute intervals

iii. SDNNi - Mean of the standard deviations of all normal NN calculated over 5-minute intervals

iv. rMSSD - Square root of the mean squared differences of successive normal NN intervals

v. SDSD - Standard deviation of differences between adjacent normal NN intervals

vi. NN50 - Number of pairs of adjacent normal NN intervals differing by more than 50 milliseconds

vii. pNN50 - Percentage of normal NN intervals differing by more than 50 milliseconds from the adjacent interval

b. Geometric measures

i. Triangular index

ii. TINN - Triangular interpolation of normal NN intervals histogram

iii. Differential index

iv. Logarithmic index

2. Linear methods - frequency domain

a. Long-term analysis (5 minutes)

i. Total power

ii. VLF - Very low frequency

iii. LF - Low frequency

iv. LFn - Low frequency in normalized units

v. HF - High frequency

vi. HFn - High frequency in normalized units

vii. LF/HF - Low frequency/high frequency ratio

b. Long-term analysis (24 hours)

i. Total power

ii. ULF - Ultra low frequency

iii. VLF - Very low frequency

iv. LF - Low frequency

v. HF - High frequency

vi. α - Slope of the linear interpolation of the spectrum in a logarithmic scale

3. Time-frequency analysis methods

a. Time-varying parametric models

i. Autoregression models

b. Non-parametric methods

i. Short-time Fourier transform (STFT)

ii. Wavelet transform (WT)

iii. Hilbert-Huang transform

iv. Wigner-Ville transform

4. Non-linear methods

a. Detrended fluctuation analysis (total DTA, α1, α2 and α1/α2)

b. Correlation function

c. Hurst exponent

d. Fractal dimension

e. Lyapunov exponent

f. Sample entropy

g. Multiscale entropy

h. Approximate entropy (ApEn)

i. Shannon entropy