Home Page of the Group Home Page of the Author



This page provides a copy of the supplementary material to [1], where the method described in [2] was used:
[1] D. Iatsenko et al (2013), "Evolution of cardio-respiratory interactions with age", Phil. Trans. R. Soc. A 371, 20110622.
[2] T. Stankovski et al (2012), "Inference of time-evolving coupled dynamical systems in the presence of noise", Phys. Rev. Lett. 109, 024101.


The videos below show time-evolution of cardiorespiratory coupling function, as well as instantaneous (IHF, IRF) and natural (\omega_{h,f} in Eq.4.1 of [1]) cardiac and respiratory frequencies. Time-averaged coupling functions for the same subjects are shown in Figure 8 of [1].

Technical details are shown below the video. 1 second of video = 1 minute of real life.



Time-evolution of cardiorespiratory coupling for young subject (21 years old male):






Time-evolution of cardiorespiratory coupling for old subject (71 years old male):







Coupling parameters were obtained by the use of Bayesian method [2]. For approximation of coupling functions, 3rd order Fourier series were used plus the same "synchronization" terms as in [1]. Consecutive 50 s time windows overlapping at 49 s were used for calculation of accurate and continuous time-evolution of coupling. Propagation constant was chosen as 0.02 (on the basis of overlap, 50/49-1).

Only central 20 min out of 30 min recording are shown in the video. First and last 5 min were discarded for convenience of presentation (e.g. during few first minutes young male had more intense respiratory coupling function, exceeding the borders of the figure in video, but changing limits to incorporate this makes time-evolution at later times less visible). If you have any questions you can contact me at dmytro.iatsenko@gmail.com.