Fig 1.
Phase–amplitude coupling between hippocampal DG low frequency rhythm (2–16 Hz, theta and alpha) and high frequency rhythm (30–80 Hz, gamma) measured by Modulation Index.
(a) PAC between low frequency rhythm (filter bandwidth = 1 Hz) and high frequency rhythm (filter bandwidth = 1 Hz). (b) PAC between low frequency rhythm (filter bandwidth = 2 Hz) and high frequency rhythm (filter bandwidth = 1 Hz). (c) Difference (b-a). (d) PAC between low frequency rhythm (filter bandwidth = 1 Hz) and high frequency rhythm (filter bandwidth = 2 Hz). (e) Difference (d-a).
Fig 2.
Simulation data were generated by Von-Mises Coupling.
(a) From top panel to bottom panel: a 5Hz slow oscillation; a 40Hz fast oscillation; a 40 Hz fast oscillation, whose amplitude was modulated by the phase of 5Hz slow oscillation; Simulation data without noise; simulation data with noise (σ = 0.1). (b) Simulation data with different absolute amplitude levels of fast rhythm. The parameter c was increased from 0.5 to 1.5. (c) Simulation data with different modulation phases. The parameter θ0 was changed from 0π to .
Fig 3.
Performance of three PAC approaches in simulation type I.
The interferential oscillations were sinusoidal oscillations. (a) Simulation data were constructed without noise. The PAC intensity k varied from 0 to 1. Each curve was averaged over 20 trails. PAC values derived from three approaches were normalized by the average of PAC values when k = 1. This simulation procedure was utilized throughout the present study. (b) Noise level σ = 0.1. (c) Noise level σ = 0.2.
Fig 4.
Performance of three PAC approaches in simulation type I.
The interferential oscillations were filtered from experimental LFP. (a) The interferential oscillations were filtered from the CA1 LFP of the puberty rat. (b) The interferential oscillations were filtered from the CA3 LFP of the adult rat.
Fig 5.
Performance of three PAC approaches in simulation type II.
The interferential oscillations were sinusoidal oscillations. (a) Simulation data were constructed without noise. (b) Noise level σ = 0.1. (c) Noise level σ = 0.2.
Fig 6.
Performance of three PAC approaches in simulation type II.
The interferential oscillations were filtered from experimental LFP. (a) The interferential oscillations were filtered from the CA1 LFP of the puberty rat. (b) The interferential oscillations were filtered from the CA3 LFP of the adult rat.
Fig 7.
Performance of three PAC approaches in simulation type III.
The interferential oscillations were sinusoidal oscillations. (a) Simulation data were constructed without noise. (b) Noise level σ = 0.1. (c) Noise level σ = 0.2.
Fig 8.
Performance of three PAC approaches in simulation type III.
The interferential oscillations were filtered from experimental LFP. (a) The interferential oscillations were filtered from the CA1 LFP of the puberty rat. (b) The interferential oscillations were filtered from the CA3 LFP of the adult rat.
Fig 9.
The phase-amplitude plots for the three simulation types and different coupling intensity k.