Figure 4.

Stimulus- and percept-specific spectrotemporal modulations of cortical activity during restored rhythm. (a) Wavelet power correlograms, in a 1–25 Hz frequency range, reveal qualitative differences in steady neural responses post probe onset, across participants (N = 32). Color arrows indicate spectrogram pairs submitted to difference contrasts as follows. (b) Differences between spectrograms reveal differential processing under alternative percepts, whether based on different physical sounds (top left), or on endogenous restorative processes (bottom left), in both cases specific to the target 5 Hz frequency band. The latter case of filling-in generates enhanced sustained power in the first harmonic band (~10 Hz) as well. Synchronization maps are shown masked by regions of group-level significance, as determined by permutations within contrast pairs, performed independently across subjects (‘driven’, p = 3.3 × 10⁻⁴; ‘filling-in’near 5 Hz, p = 6.7 × 10⁻⁴, ‘filling-in’ near 10 Hz, p = 0.01). The lower-rate rhythmic enhancements (~5 Hz) coincide spectrotemporally even though the sensory bases for each are different (right). White vertical lines indicate noise probe temporal edges.