Castiglioni Paolo, Parati Gianfranco, Faini Andrea
IRCCS Fondazione Don Carlo Gnocchi, 20148 Milan, Italy.
Istituto Auxologico Italiano, IRCCS, Department of Cardiovascular, Neural and Metabolic Sciences, S.Luca Hospital, 20149 Milan, Italy.
Entropy (Basel). 2019 May 31;21(6):550. doi: 10.3390/e21060550.
Multiscale entropy (MSE) provides information-domain measures of the systems' complexity. The increasing interest in MSE of the cardiovascular system lies in the possibility of detecting interactions with other regulatory systems, as higher neural networks. However, most of the MSE studies considered the heart-rate (HR) series only and a limited number of scales: actually, an integrated approach investigating HR and blood-pressure (BP) entropies and cross-entropy over the range of scales of traditional spectral analyses is missing. Therefore, we aim to highlight influences of higher brain centers and of the autonomic control on multiscale entropy and cross-entropy of HR and BP over a broad range of scales, by comparing different behavioral states over 24 h and by evaluating the influence of hypertension, which reduces the autonomic control of BP. From 24-h BP recordings in eight normotensive and eight hypertensive participants, we selected subperiods during daytime activities and nighttime sleep. In each subperiod, we derived a series of 16,384 consecutive beats for systolic BP (SBP), diastolic BP (DBP), and pulse interval (PI). We applied a modified MSE method to obtain robust estimates up to time scales of 334 s, covering the traditional frequency bands of spectral analysis, for three embedding dimensions and compared groups (rank-sum test) and conditions (signed-rank test) at each scale. Results demonstrated night-and-day differences at scales associable with modulations in vagal activity, in respiratory mechanics, and in local vascular regulation, and reduced SBP-PI cross-entropy in hypertension, possibly representing a loss of complexity due to an impaired baroreflex sensitivity.
多尺度熵（MSE）提供了系统复杂性的信息域度量。对心血管系统MSE的兴趣日益增加，在于其有可能检测与其他调节系统（如高级神经网络）的相互作用。然而，大多数MSE研究仅考虑心率（HR）序列且尺度数量有限：实际上，缺少一种综合方法来研究传统频谱分析尺度范围内的HR和血压（BP）熵以及交叉熵。因此，我们旨在通过比较24小时内不同行为状态，并评估高血压对BP自主控制的影响（高血压会降低BP的自主控制），来突出高级脑中枢和自主控制对广泛尺度范围内HR和BP的多尺度熵及交叉熵的影响。从8名血压正常参与者和8名高血压参与者的24小时BP记录中，我们选择了白天活动和夜间睡眠期间的子时段。在每个子时段，我们获取了一系列连续16384次搏动的收缩压（SBP）、舒张压（DBP）和脉搏间期（PI）。我们应用一种改进的MSE方法，以获得高达334秒时间尺度的稳健估计，涵盖频谱分析的传统频段，针对三个嵌入维度，并在每个尺度上比较组间（秩和检验）和不同条件（符号秩检验）。结果表明，在与迷走神经活动、呼吸力学和局部血管调节调制相关的尺度上存在昼夜差异，且高血压患者的SBP - PI交叉熵降低，这可能代表由于压力反射敏感性受损导致的复杂性丧失。
