Dynamical complexity detection in short-term physiological series using base-scale entropy

Abstract

Physiological systems generate complex fluctuations in their output signals that reflect the underlying dynamics. The base-scale entropy method was proposed as a complexity measure to investigate the complexity of time series. The advantages of this method are simplicity and extremely fast calculation for very short data sets. This method enables analyzing very short, nonstationary, and noisy data sets. We employed this method for short-term physiological time series for analysis of heart-rate variability signals. The results show that the simple and easily calculated measure can effectively detect the complexity dissimilarity of physiological time series in different physiological or pathological states, which is convenient for clinical applications.

Multiscale Symbolic Transfer Entropy on Closing Eyes and Being Idle in rhythm of Electroencephalogram

“…Therefore, it is of great importance to learn more about EEG [1][2][3][4][5]. In this work, we analyze the value of multiscale symbolic transfer entropy (MSTE) on closing eyes and being idle in all four rhythms of EEG, and finally, we found out that it in the δ rhythm of EEG are definitely different.…”

Multiscale Symbolic Transfer Entropy (MSTE) Analysis in Closing Eyes and Being Idle

“…As the formula [16][17][18] above, we can change the original data into a symbolic sequence, the result is like the Fig. 1.…”

Analysis of Sleep Staging Based on Multivariate Symbolic Transfer Entropy

“…When the time delay L is selected as 1, the number of m-dimensional vector is N-m+1. For any m-dimensional vector, the BS (basic scale) [10] is calculated by the root mean square of the difference between the adjacent points of the m-dimensional vector: mark all the regions, so it is meaningless to divide this value. α is a special constant parameter, so it is very important to choose the right α for the interval and the division of the symbol.…”

“…after the data collected on the whole partition, then solve the problem on the templates of each small interval. Dynamic adaptive partitioning method [10,16] as follows: Set a length of N points for the time series:…”

“…The traditional symbolic dynamics uses static division method. In this paper, we use the dynamic adaptive [10] method to divide the time series then use multivariate symbolic transfer entropy algorithm to obtain the consistent conclusion…”