Entropy Analysis In Biological Information Processing

In this paper, we have analytical studied two biological signal, the one is the gait timing sequence, and the other is ECG signal. We have used two algorithms, the approximate entropy and the power spectrum entropy, and used them to analytical study the biological signal.The study has two parts: the first part, using approximate entropy to analytical study the children's gait timing sequence, and observing the changing discipline of the value of approximate entropy, and find that children in three years old to fourteen years old, the value of approximate entropy is decreasing as time goes on, and gait is more and more stable. The study relative use linear analysis method to study these time sequences usually, however, this paper use the algorithm approximate entropy to study, and find that the approximate entropy can reflect the difference of the timing sequence's complexity, it needs little data, and it is sensitive and stable. The second part, we put forward power spectrum entropy on sub-frequency and reducing frequency band, and use them to study the ECG signal of a healthy group and a myocardial infarction group, and we want to investigate the difference of the power spectrum entropy of the healthy group and the myocardial infarction group. Then we find that the power spectrum entropy on sub-frequency band can differentiate the old myocardial infarction group and the old healthy well. Analytical studying using reducing frequency band, and as time goes on, the power spectrum entropy of the healthy is decreasing, but the myocardial infarction's is almost the same. In the middle years old, the entropy of the healthy is great than the entropy of the myocardial infarction (<0.05), but in the elderly, the entropy of the healthy is less than the myocardial infarction's (<0.001). However, using the ordinary whole-frequency band, because it cannot reflect the information in part band, so it could hardly get the answer like this paper. Power spectrum entropy on sub-frequency to differentiate the ECG signal of the healthy and of the myocardial infarction is simple and feasible, and to be a easy calculable and sensitive specificity index.