Fig. 4: Same state probability and perceptual performance.

a (Top) Normalized pre-stimulus ongoing power ratio (PR) for 120 detection task trials on one example electrode. The green dotted line represents the median. Trials below the median are in the low PR group (blue), and those above are in the high PR group (red). (Middle) The mean normalized PR of the remaining simultaneously recorded electrode population (excluding the example electrode at the top) for the same example session, divided into low PR (blue) and high PR (red) groups using the median value (green line). (Bottom) Black lines indicate the Low/High states for the example electrode (top) and the remaining population (bottom) for the 120 trials of the example session. High PR trials are above the session median, and low PR trials are below it. Orange shaded trials correspond to instances where the example electrode’s state matches that of the rest of the population. The example electrode had the same state as the remaining population in 65% of the trials. b Population distribution of same-state probabilities between individual temporal electrodes and the average of the rest of the population of the temporal electrodes. On average, individual electrodes shared the same state as the rest of the population for the majority of trials (0.61 ± 0.0054%, mean ± SEM, across all subjects). Inset: same-state probabilities between individual temporal electrodes and the average of all the electrodes from all the lobes (global state) of the brain. Arrows represent the mean. c Perceptual performance for detection (blue) and discrimination (red) tasks for low and high ongoing PR state trials. The high ongoing PR trials were associated with significantly better performance than the low ongoing PR trials for the detection task (low PR: 57.85 ± 2.81% vs. high PR: 70.92 ± 2.57%, mean ± SEM, p = 0.003, n = 97 electrodes, bootstrapping with 1,000,000 iterations). Conversely, for the discrimination task, the low ongoing PR trials were associated with significantly higher performance values than the high ongoing PR trials (low PR: 65.1 ± 2.48 vs. high PR: 46.5 ± 2.58, mean ± SEM, p < 1e-6, n = 97 electrodes, bootstrapping with 1,000,000 iterations). **p < 0.01.