Figures and data
Sequential activation of two discrete brain areas appears as waves in simulated EEG phase latency maps.
(a) Schematic illustration of simulated EEG. Two brain areas (green and red ellipses) are defined on the cortex by a set of dipoles that are activated uniformly with a time delay (ΔT) between areas. A finite element forward model in Brainstorm 50 is used to simulate the resulting potentials measured with EEG electrodes (purple dots). (b) Latency to peak measured across electrodes following the sequential activation of two sets of dipoles (green and red circles in the background) with a distance of 11cm (Δx/σx=16.9) and time delay of 20ms (ΔT/σT=0.2). A continuous delay patterns propagating from posterior (P) to anterior (A) is observed. (c) Phase-latency distance correlations (PLDC) for different distances and different delays between two activated areas. PLDC remains close to 1 (indicating apparent wave propagation) for a wide range of values.
A simple analytic model of sequentially activated Gaussians exhibits travelling peaks with a diversity of patterns.
(a) The spatiotemporal potential of two sequentially activated one dimensional spatial Gaussians (centers; x1=-0.5 and x2=0.5; standard deviation σx=1), each with a temporal Gaussian activation profile (peaking at t1=-1 and t2=1; standard deviation σt=1), as in Equation 1. Notice wave-like propagation. (b) Spatial profile of Equation 1 at time 
Sequential activation of discrete spiking populations underlies LFP waves in turtle cortex.
(a) Schematic of the experimental setup. The brain with eye attached is extracted and the visual cortex is flattened on a MEA and its activity is measured in response to visual stimulation of the retina. (b) A cortical response following visual stimulation at t=0. Traces show the following data: raw electrode (grey), low-pass filtered (<2Hz, blue with dotted line marking the first oscillation cycle), high-pass filtered (>200Hz, black), ALSA (red) and the instantaneous phase of the low-pass filtered trace (green). (c) Power spectral density (averaged over 1000 randomly selected trials from one experiment). (d) Dynamics of low-pass filtered (<2Hz) and averaged (n=4000) cortical responses to visual stimulation. Amplitude in each time window (200-750ms post stimulus in 50ms intervals) is color coded and presented on the physical space of the electrode. Filled color-coded circles mark the electrodes recorded in (e). M=Medial, A=Anterior, L=Lateral, P=Posterior. (e) Averaged (n=4000) low-pass filtered (<2Hz) traces for channels along the delay gradient. The maximum is marked by a circle and its location on the array is color coded in (d). (f) Phase Latency-distance correlation (PLDC) distribution for all trials with strong waves (n=2415, blue) from this experiment and for the same trials with shuffled latencies (red). (g) Spiking responses (black dots) in a single visual stimulation trial. Empty circles mark the π/2 phase crossing. Red circles denote the ALSA onset for each electrode. Electrodes were reordered according to the LFP phase crossings. (h) The distribution of LFP phase crossing times and first spike times for the response in (g) with corresponding DIP values (0.85 and 0.0002, respectively). (i-j) Latency maps on the electrode array’s physical space extracted from the Hilbert phase (i) and ALSA (j). Black line denotes the wave center path extracted from the Hilbert phase (see Methods). (k) DIP-test p-value distribution of the 1st spikes (blue), Hilbert phase crossings (red) and differences between both per trial (gray) for strong waves with enough spikes (n=902). Distribution medians (solid black line) are connected by a dashed black line (blue line connects the distributions’ means). (l) Same as (k) but for ALSA. (m,n) Medians of the DIP-test p-values (black lines in k,l, respectively) from 6 recordings in 4 animals. Dashed black line marks the median over all experiments. All results except (m,n) are from the same recording.
