SUPPLEMENTARY INFORMATION – Captions
Figure SI 1: Spontaneous vs. Evoked Amplitudes
The mean amplitude of the 100 frames that were maximally correlated with the horizontal map. Blue – evoked session, red - spontaneous session. Result from 5 different experiments (hemispheres). For each experiment, the values were normalized to the amplitude of the evoked session. When only 3 maximally correlated frames were used, the difference in amplitude varied between 10% and 50%, with a 30% mean difference, as stated in the text.
Movie SI 2: Ongoing activity and the emergence of intrinsic states.
The left panel displays a sequence of single frames from a movie (recording session) showing the instantaneous cortical activity in the absence of stimulation (raw data; not processed). The time relative to the first frame is displayed in the center. Whenever a state that is significantly correlated with one of the orientation maps emerges, that map appears on the right panel, its corresponding orientation is displayed in the center, and the correlation between the two frames is displayed over the left panel. When the maximal correlation for a state is reached, the movie is briefly paused. To facilitate the perception of the similarity between the structures, outlines of the major patches have been added to both panels.
Figure SI 3: Similarity of Spontaneous vs. Evoked States.
A: Top trace: for each orientation, the average of the 30 best correlated single evoked frames with the corresponding orientation map. Bottom trace: Same with spontaneous frames. B: Same as A, from another experiment. All frames are unsmoothed. To maximize pattern visibility, all frames are displayed with maximal gain.
Movie SI 4: The dynamics of ongoing activity revealed by the Kohonen templates; its relation to the orientation maps.
Right panel - dynamics of correlations with Kohonen templates; Left panel - the same for orientation maps. By comparing the dynamics in two panels one can verify that the Kohonen templates correspond to the real maps. The right panel displays the sequence of vectors pointing to the template that is maximally correlated with the instantaneous pattern of ongoing activity. The length of the vector is proportional to the value of correlation coefficient. The orientations were assigned to the templates by selecting the template corresponding to horizontal orientation and adding an equal angle of 4.5 deg (180 deg divided by 40, the number of templates), to every subsequent template. The left panel illustrates the same analysis of the data, but using the orientation of the map that is maximally correlated with instantaneous pattern of ongoing activity. The radius of the circle shows the threshold for significant correlation (P < 0.01; |0.22|). The time relative to the first frame is displayed in the center.
Comments SI 5: On the modeling of the dynamics of cortical states.
Recent theoretical models introduced the notion of marginal regime of cortical activity1, for which stereotyped patterns (states) that reflect intrinsic connections emerge spontaneously. The fact that we observed the corresponding states appearing for a 20% of the time only, may indicate that the visual cortex under our experimental conditions operated close to the marginal regime and undergoes occasional transitions into it. Alternatively states other than orientation appeared but were not identified here. The overall, patchy shape of the orientation states is compatible with the network models characterized by combination of intra-cortical excitation and inhibition2 (Mexican-hat type of interactions). However, we found that further theoretical work, taking into account more intricate geometrical properties of the states, need to be made for precise predictions of the intrinsic dynamics and the connectivity profile that can explain the present results.
Reference
1. Ben-Yishai, R., Bar-Or, R. L. & Sompolinsky, H. Theory Of Orientation Tuning In Visual-Cortex. Proceedings of the National Academy of Sciences of the United States of America 92, 3844-3848 (1995).
2. Ernst, U., Pawelzik, K., Sahar-Pikelny, C & Tsodyks, M. Intracortical origin of visual maps. Nature Neuroscience 4:431-436 (2001).