Fig. 12.

A graphical depiction of the 3rd step of the SigMax detection method. Synchronization was measured from the empirical cumulative distribution of recurrence times between stationary MUAPTs with dependent firing instances. Recurrence times were divided into different orders based on the recurrence interval, i, and the mean IPI of the alternate MUAPT, $\overline{IP{I}_{Alt}}$, as detailed by Eq. 1 in the text. A window of width W_j was selected and used to calculate the number of recurrence times at different latencies, L. The latency that provided the maximum number of recurrence times, k_max, was selected for significance testing. Statistical significance was computed as the probability, P_j, of finding k_max occurrences within a window of width W_j, using the equations in step 3 of appendix. This procedure was repeated for windows of different widths defined by Eq. 7 in the text. The most statistically significant density of recurrence times was identified as the j^th window width that produced the minimum probability, P_min, of all P_j. If P_min were a <0.05 significance threshold, then the pair of MUAPTs manifested statistically significant synchronization.