THE SEARCH FOR HIDDEN PERIODICITIES IN BIOLOGICAL TIME-SERIES REVISITED
THE SEARCH FOR HIDDEN PERIODICITIES IN BIOLOGICAL TIME-SERIES REVISITED
复制标题
DOI：
10.1016/s0022-5193(89)80067-0
复制
发表时间：
1989-08-22
影响因子：
2
通讯作者：
RINGO, JM
中科院分区：
文献类型：
--
作者：
DOWSE, HB;RINGO, JM
在研究生物系统的节律性时，必须解决两个关键问题：所研究的过程是否具有显著的周期性，如果是，振荡周期的最佳估计是什么。惠特克-罗宾逊“周期图”已被广泛用于回答这两个问题，傅立叶分析在较小程度上。数字信号处理的进步已经产生了优于两者的上级技术，我们已经将其中之一（最大熵谱分析或梅萨）应用于生物数据。我们还开发了一种新的方法，用于分析信号噪声比的生物节律数据使用的自回归模型的基本梅萨。我们在这里回顾了目前的生物时间序列分析方法，并描述了我们的应用这些技术。这种技术组合的上级性能使用以前发表的数据证明。此外，采用经验的方法，我们已经证明，循环，但aperiocompatibility（即混沌）系统可以区分从嘈杂的周期性或随机使用这些分析的组合。这对超昼夜节律和昼夜节律的工作的影响进行了讨论。
In the study of rhythmicity in biological systems, two critical questions must be addressed: whether the process under investigation is significantly cyclic, and if so, what is the best estimate of the period of the oscillation. The Whittaker-Robinson "periodogram" has been used extensively to answer both of these questions, Fourier Analysis to a lesser extent. Advances in digital signal processing have produced techniques superior to both, and we have applied one of these (Maximum Entropy Spectral Analysis or MESA) to biological data. We have additionally developed a novel method for analyzing signal-to-noise ratios in biological rhythm data using the autoregressive model underlying MESA. We review here the current methodology for the analysis of biological time series and describe our application of these techniques. The superior performance of this combination of techniques is demonstrated using previously published data. In addition, employing an empirical approach, we have demonstrated that cyclic but aperioidic (i.e. chaotic) systems may be distinguished from noisy periodic or stochastic ones using a combination of these analyses. The implications of this for work on ultradian and circadian rhythms are discussed.