The Journal of Neuroscience, June 6, 2007, ():
Enhanced Category Tuning Revealed by Intracranial Electroencephalograms in High-Order Human Visual Areas
J. Neurosci. Privman et al.
27: 6234
Supplemental Data
Files in this Data Supplement:
- supplemental material -
Supplementary Table 1 – Summary of all the electrodes having highest co-localization with fMRI.
Columns (left to right) mark: (i),(ii) Electrode and patient identification. (iii) Modality and Cortical Region – the estimated electrode location base on sulcal topography & Visual regional definition taken from (Hasson et al., 2003). (iv) Reproducibility – the correlation between the Gamma BLP of the same electrode in two runs of the movie experiment. (v) iEEG to fMRI correlation for this electrode (see Methods) (vi) The estimated anatomical location of the electrode in Talairach space (based on CT images of the electrode sites).
- supplemental material -
Supplementary Table 2 –Correlation of iEEG signal across patients according to cortical regions.
Electrodes from all patients that were found correlated (r>0.2) with any electrode from another patient, are grouped according to functional regions (low order visual; high order visual; auditory). The number of electrodes per region is indicated in parenthesis. Each cell indicates the percentage of electrode pairs that were found correlated across subjects. As can be seen, the level of correlation across patients is much higher for similar cortical regions.
- supplemental material -
Supplementary Table 3 – Ictal epileptic foci.
Ictal epileptic foci per patient as identified through ECoG.
Columns (left to right) mark: (i) Patient identification,(ii) Hemisphere of seizure foci. (iii) Anatomical region of seizure foci (iv) Estimated coordinates of seizure foci in Talairach space (based on CT images of the electrodes sites).
- supplemental material
-
Supplementary Figure 1. Correspondence between iEEG and fMRI.
Correlation maps between gamma BLP of the 10 iEEG electrodes seen in Figure 3, and fMRI data. The correlation to each electrode’s activity is presented separately on a flat cortical surface. Blue dots denote the estimated anatomical location of the electrodes on the patient’s cortex, as determined by co-registration of post-operative CT to the pre-operative MRI. E1 to E20 refer to the different electrodes, whose details are provided in Supplementary Table 1, and the color dot below the label is reference to the color code of Figure 3.
- supplemental material
-
Supplementary Figure 2. Category selectivity of fMRI responses in a healthy subject.
(a) The red patch depicts the fMRI voxels of a healthy subject that best correlated to the iEEG signal of face selective electrode in the movie experiment. The white contour shows the face-selective region-of-interest (ROI) found to be face-selective in the fMRI localizer experiment (see methods) in the same subject. (b) Face selectivity of the correlated voxels (red patch in panel a). Plots of the average fMRI response (percent signal change) to faces (red), tools (blue) and houses (green). (c) Face selectivity of the most selective voxels in the entire hemisphere (see Methods for details). Plots are the same as in b.
- supplemental material
-
Supplementary Figure 3. ‘Gallery’ of pictures that evoked the strongest VEP
The galleries are presented in descending order (strongest response, top left corner). conventions as in Figure 5. (a) and (b) show galleries of face-selective electrodes anatomically located in the Fusiform Face Area (FFA). (c) and (d) show galleries of face-selective electrodes, anatomically located in the Lateral Occipital (LO) region. (e) A house-selective electrode. (f) A tool-selective electrode.
- supplemental material
-
Supplementary Figure 4. Power spectrum of the iEEG signal
Mean power spectrum of the activity recorded from all electrodes exhibiting reproducible activity across repeated movie presentations (log-log scale). The drop at 50Hz is due to our preprocessing procedures. As can be seen, power decreases with increasing frequency.
- supplemental material
-
Supplementary Figure 5. Spatial selectivity of the iEEG signal
(a) to (d) Mean visual evoked potentials (VEP) plot of four adjacent electrodes (8 mm inter-electrode separation) from a ventral strip of electrodes (see panel e). In each category odd trials are plotted separately from the even trials. As can be seen, the category selectivity of each electrode was different from its neighbors. (e) A ventral view of the brain, the estimated anatomical locations of the electrodes are marked with red letters according to the panel of each plot.
- supplemental material
-
Supplementary Figure 6. Distribution of the reproducibility level across all electrodes
Each panel depicts the distribution of the reproducibility (correlation of the signal between the two movie runs) in all iEEG electrodes. Panels (a) to (f) correspond to the different patients (P1 to P6). Red bars mark electrodes that crossed the threshold of r = 0.2 and were defined as responsive to the movie (see Methods). (g) A ‘pseudo random’ distribution of reproducibility values obtained by correlating the signal of the first run to the reverse signal of the second run, where no real correlations are expected. Note that in this distribution no electrodes reached a reproducibility level above r = 0.1, implying that our threshold of r = 0.2 was quite conservative.
