Figure illustrates typical examples of the OIS response in SI of three different subjects in the absence of stimulation (control), and during low (50 μm) versus high (400 μm) amplitude stimulation. Each image shown in Figure represents the sum of frames taken from the time of stimulus onset to 5 seconds after stimulus offset (frames 1–16). Areas of high absorbance are indicated by dark patches within each image; regions of high absorbance in each case correspond to the SI locus that represents the stimulated site on the skin. SI in each experiment underwent a larger increase in absorbance within the region of interest (ROI) in response to the 400 μm stimulus than evoked in the same region by the 50 μm stimulus. Moreover, in each subject the increase in absorbance appears more evenly distributed and less diffuse throughout the ROI under the 400 μm condition.
Figure 1 OIS response to low vs. high amplitude stimulation. OIS images taken from three subjects (A, B, C). All images are anatomically oriented as shown in the top left image. Images were obtained by averaging across 10 experimental trials and then summing frames (more ...)
To identify the boundaries of the regions of increased absorbance spatial histograms were constructed. Figure illustrates the average results obtained from all experiments (n = 5) as well as the methodology used to evaluate the spatial extent of the stimulus-evoked activation. In each experiment, the image was segmented along a line 6 mm long and roughly centered on the area of increased absorbance, as shown in the top panel. Pixels along the line were binned (bin size 40 × 200 μm) and absorbance values averaged and plotted as a function of distance along the line. The plots demonstrate that at all amplitudes of stimulation, the spatial extent of the region of above-background absorbance (ie. absorbance values larger than control) is similar and at every stimulus amplitude occupies a circular-shaped territory in SI between 1.8 – 2.24 mm in diameter. The ROI (to be used for further analysis) was therefore defined as the region displaying above background levels of absorbance within the activated region of SI.
Figure 2 Spatial histograms of activity at different amplitudes. Absorbances were measured at each amplitude along the red line shown in the OIS image at top left. Each plotted value represents an average of pixels spanning 100 μm above and below the line (more ...)
Figure demonstrates (for one exemplary subject) the method used to evaluate the time course of stimulus-evoked SI absorbance. Panel A shows a green filter image of the cortical surface, which highlights the vasculature. Panels B&C are the OIS responses (dark regions) evoked by the low (50 μm) and high (400 μm) stimulus amplitudes respectively. The ROI is the circular territory enclosed by the dashed white lines. Absorbance values within the ROI were averaged for each amplitude of stimulation and plotted as a function of time. The time course of the absorbance values measured between 1 and 22 sec after stimulus onset is plotted for each of the stimulus conditions, and in the absence of stimulation ("control"). Arrows along x-axis of plots at bottom left of Figure indicate stimulus onset (1 sec) and stimulus offset (11 sec), and reveal how absorbance increased with increasing amplitude of stimulation. For each stimulus amplitude absorbance is maximal near to stimulus offset.
Figure 3 Absorbance time course and anatomical registration in SI. A) Green filter image of SI cortex in vivo, used for anatomical registration of OIS images. B&C) Resulting OIS image obtained from averaging across 10 experimental trials and then summing (more ...)
The analysis approach illustrated in Figure was performed on all experiments (n = 5) and the resulting absorbance plots were averaged (Figure ). Similar to Figure , the plots in Figure demonstrate that absorbance increases with increasing stimulus amplitude. To quantify this relationship a measure of ΔAbsorbanceevoked was used. ΔAbsorbanceevoked was defined as the difference between the absorbance measured at 1 sec (prior to stimulus onset) and 11 sec (point of stimulus offset), and is shown in the plot at the bottom of Figure . The plot of ΔAbsorbanceevoked vs. amplitude is well described (coefficient of determination R2 = 0.9921) by the linear function (solid line) ΔAbsorbanceevoked = (4 × 10-6)*d + 0.0005. This type of analysis, however, gives little or no information about the spatial properties of the response.
Figure 4 A) Plots of absorbance and standard deviation averaged across all experiments (n = 5). All data was normalized prior to being averaged. Plotted absorbances were measured within the ROI which was defined as all pixels within a 1 mm radius of the center (more ...)
Radial histograms were constructed to better visualize the spatiotemporal relationship of the OIS response at different amplitudes. The radial histograms shown in Figure are representative time-space plots at each amplitude. From the ROI center (as determined by spatial histogram analysis), average absorbance values were determined for the pixels within concentric rings located at progressively larger distances from the center at each frame acquired. Absorbance values are color coded (red indicating areas of high absorbance, blue indicating areas of low absorbance) and plotted as a function of time and radial distance from the center of the ROI. Figure demonstrates that the major differences that exist in the SI global responses to different amplitudes of stimulation are differences in the magnitude of absorbance and not the spatial properties of the absorbance pattern (this also is apparent in the spatial histogram analysis of Figure ). As would be expected based on the absorbance curves shown in Figures &, higher stimulus amplitudes evoked a more intense and discrete region of increased absorbance than did the lower amplitudes. Interestingly, one of the more robust differences between low- and high-amplitude stimulation, is the magnitude of decreasing absorbance detected in the territory that surrounds the region in which absorbance increases. This response is most pronounced under the 400 μm condition where it can be seen to occur much sooner after stimulus onset at radial distances as small as 1.5 mm from the ROI center. Spatially, the regions of above- and below-background absorbance are nearly the same at each stimulus amplitude, with the above-background portion extending nearly 1 mm away in all directions from the center of activity, whereas the below-background portion of the response comprises a ring beginning at a radial distance of 1.5 mm from the ROI center and extending out beyond the area that was analyzed.
Figure 5 Radial time space plots. Radial histograms were performed on OIS data at all stimulus amplitudes. Radii were measured from the center of activation as demonstrated in the image at top-left. The dashed blue line is the maximum radial distance used in the (more ...)
Using similar techniques to those we used to analyze above-background activity in the ROI (as in Figures &), regions outside the designated ROI were examined to determine whether a similar amplitude-dependency could be established for the time courses of the below-background absorbance observed in the surround. Figure shows plots constructed from averaging the absorbance values in pixels lying 1.5 – 2 mm away from the center of the ROI. Data were normalized and then averaged across experiments (n = 5). It is apparent that the time courses at different amplitudes of stimulation are different with respect to the stimulus timing, (compared with above background levels of activity, which all show maximum absorbance at the point of stimulus offset). Accordingly, a different measure was adopted to quantify this relationship: In this case ΔAbsorbancemax was defined as the difference between the minimum absorbance and the maximum absorbance value obtained at any point during the recording. Interestingly, the relationship between stimulus amplitude and ΔAbsorbancemax in the surround is not linear (Figure ). Instead, each of the higher stimulus amplitudes employed (100–400 μm) evoked a very similar level of below-background absorbance. The sole difference between the different curves (Figure ) is the time required to reach the peak of the decrease in absorbance. That is, as amplitude is increased from 100 to 400 μm the point of minimum absorbance was attained earlier in time. This is also apparent in Figure .
Figure 6 Absorbance trends surrounding the ROI across five experiments. A) Plots of absorbance and standard deviation averaged across all experiments (n = 5). Plotted absorbances were measured at radial distances between 1 and 1.5 mm away from the center of the (more ...)
Correlation maps were constructed to further characterize the spatial properties of the SI response to 25 Hz flutter. A correlation map compares every pixel in the image with the signal referenced from the ROI, and assigns a correlation coefficient (r) to the location of the pixel being compared. This gives a fairly good approximation of the signal at all locations in the image. Since there is no significant difference in the spatial properties between stimuli at intermediate amplitudes (as demonstrated by radial histograms) only the 50 and 400 μm amplitudes will be compared with this technique. Figure shows correlation maps of the OIS responses to stimulus amplitudes of 50 and 400 μm (Top panels). The bottom panels of the figure show the input signal (solid dark red line) used for correlation of each pixel in the map, and the negative (opposite) of the input signal (dotted blue line). A coefficient of +1 (although it never appears in the map) indicates that a pixel's time course perfectly matches the input signal while a coefficient of -1 indicates that pixel's time course perfectly matches the opposite of the input signal (dotted blue line). At 50 μm the correlation map (color-coded image) shows that the correlation is weak and more dispersed within the ROI in area 3b. At the highest amplitude, however, there is a pronounced and well-defined positive correlation within the ROI that is more evenly distributed throughout the ROI. A large region of negatively correlated activity (corresponding to strong below background activity) surrounds the ROI in the high-amplitude map.
Figure 7 Correlation maps for stimulus amplitudes of 50 (left) and 400 (right) microns. The schematic at top-right indicates the anatomical orientation of the OIS image as well as all maps: A-anterior, M-medial, P-posterior, L-lateral. Color bars show coefficient (more ...)
To examine the spatial dynamics of the SI response in more detail we examined the patterns of activity generated by low- and high-amplitude stimulation in a 16 mm2 (4 × 4) area centered around the ROI. Figure demonstrates the patterns of activity evoked at three time intervals during the delivery of the stimulus. The 3D surface plots show activity measured within the boxel indicated by the dashed box in the image at the top. In each 3D plot absorbance is plotted in two-dimensional space and is indicated by two measures: height of the peak along the z-axis (as shown in the schematic at the top right), and the color (indicated by the color bar to the right of each row of 3D plots). These data make it apparent that after a short period of stimulation (1 sec) the activity pattern is very similar for the different amplitudes. That is, at this early time interval both patterns are diffuse and occupy much of the ROI. However as stimulus duration increases, the pattern of increased absorbance evoked by high-amplitude stimulation tends to become restricted to the center of the ROI and within this region becomes homogeneous. Standard deviation was used to measure the variability within the ROI at low and high amplitudes. At 10s after stimulus onset, standard deviations for low- and high-amplitude surface plots were 0.1415 and 0.1166 respectively. Average standard deviation across all sets of maps (n = 5) differs very little from these values (0.1448 at low amplitude vs. 0.1201 at high amplitude). These differences suggest that within the ROI evoked absorbance levels are more homogeneous in response to high- vs low-amplitude stimulation. In addition, at 5 and 10s after onset of high-amplitude (but not low-amplitude) stimulation the territory surrounding the ROI becomes dominated by below-background changes in absorbance.
Figure 8 Spatial plots of activity evoked by low- (50 μm) and high-amplitude (400 μm) stimulation. The schematic at the top-left indicates the anatomical orientation of the cortex: A-anterior, M-medial, P-posterior, L-lateral. Absorbances were (more ...)