Vergence angle significantly modulated planning period activity in 74% (101/137) of the population for coupled reach targets with constant disparity (ANOVA; P < 0.05). 58% (44/76) of neurons that were sensitive to coupled reach target disparity (n = 76) had a main effect of vergence angle (ANOVA; P < 0.05). A large proportion (52%, 32/61) of disparity insensitive neurons (61/137) was also significantly modulated by vergence angle (ANOVA; P < 0.05); these neurons directly represent fixation depth during planning. (See Figure 4-6A.) It remains possible that this disparity insensitive
population of neurons is sensitive to disparity as well as vergence in other frontoparallel locations (Figure 4-6B). In addition, it has been shown that neurons in area 7a are sensitive to fixation position in 3D (Sakata, Shibutani et al. 1980); the use of a single of frontoparallel location for fixation targets in depth likely underestimates the degree of vergence angle sensitivity in the PRR population.
A Vergence Tuning Index (VTI; see 3-3), similar to DTI, was based on response modulation by vergence angle for reach targets at a constant disparity and computed for PRR neurons. The mean of the maximum planning period VTI from each neuron in the population (n = 137) is 0.4326 (±0.2348) with a median VTI of 0.3612, and is nearly identical to and correlated with the DTI (see Figure 4-7). VTI is likely underestimated due to the fact that only 3 samples (13°, 9.7°, 6.5° vergence angle) were obtained at each disparity.
Figure 4-7 VTI from Experiment 1. A: Histograms of planning period VTI for all neurons (n = 137). The average VTI was similar across disparity, however there was a significant difference with a lower VTI for targets at 0° disparity (Kruskal-Wallis; P = 0.0277), which was found to be only between targets at 0° and -1.0° disparity (Kruskal-Wallis with Bonferroni correction for multiple comparisons, P < 0.05). B: Histograms for cue period VTI for all neurons (n = 137). There is no difference in VTI due to target disparity (Kruskal-Wallis; P > 0.17). C: VTI is paired for planning and cue periods for all neurons. The VTI averaged across disparity is similar for cue and planning periods (µcue =
0.3103±0.1896, µplanning = 0.2678±0.1606; Kruskal-Wallis; P > 0.07), and no significant
differences exist between cue and planning VTI at each level of disparity (Kruskal-Wallis, all P > 0.06). The average difference between cue and planning VTI across neurons and disparity is -0.0425, which is a 13.69% reduction of the cue VTI during planning, and the difference does not vary with disparity (Kruskal-Wallis; P > 0.90). The correlation of the average VTI across disparity between the cue period and planning period is r = 0.84 (P < 1e-5) and correlations between VTI during the cue period and movement planning at each level of disparity are shown (P < 1e-5 for all r). D: The planning period DTI and VTI, averaged across vergence angle and disparity respectively, is paired and shown for all neurons (n = 137), with a correlation of r = 0.92 (P < 1e-5). E: Proportion of vergence sensitive cells by epoch.
E D
Fixation period activity was least modulated by vergence angle in this study. Figure 4-7E shows the percentage of cells in the population significantly modulated by vergence angle during coupled reaches. In this case, activity during an epoch was averaged across
disparities within a vergence angle, and differences in firing rates between vergence levels were examined. The visual stimulation of the cue had a greater effect due to vergence; however the modulation by vergence during the memory period effected over half the population. An even larger proportion was modulated during the reaching movement. This illustrates the visuomotor nature of the neurons, where the neural firing rate during the visual guidance of the hand critically depends on the position of the eyes that determines the viewing distance. The firing during the reach could be a result of pure feedback from the motor areas that are coordinating the muscle movements required to make the reach, however the modulation by vergence angle shows this is not the case. The modulation by vergence during this motor action instead suggests that the firing is related to the visual guidance of the hand, however since vergence angle and reach depth covary in the coupled design (Figure 3-2C), different regions of space or sampled with each vergence angle as seen in Figure 3-7A. Comparisons on vergence angle modulation when reach depth is constant (Figure 3-2A) are performed with decoupled targets, and can be found in 4.10 as Index A.
The fact that vergence angle modulates the planning period neural activity of many disparity tuned neurons in PRR suggests that target depth is directly encoded by these cells. A population of neurons sensitive to only vergence angle (pure vergence) supports the functionality of an eye centered encoding scheme by explicitly representing the depth
of fixation (Figure 4-6A; yellow). Pure vergence encoding cells may be the source of the vergence modulation observed in disparity sensitive cells though lateral connections. In addition, vergence encoding cells serve as a relay for the vergence signal to downstream cortical areas. These downstream cortical areas could in turn use a pure disparity signal (not modulated by vergence) and combine the information from a pure vergence encoder to infer the egocentric depth of a target. Again, it remains possible that pure vergence encoding cells for one frontoparallel target configuration may additionally encode the disparity of a reach target in another frontoparallel configuration. Figure 4-6B shows a hypothetical 3D receptive field for a neuron. The reach target does not fall in the frontoparallel receptive field, and thus the neuron does not show sensitivity to the disparity of the reach target, however the neuron may still encode the vergence angle when fixating the fixation stimulus. Several examples of neuronal tuning during
movement planning for pure vergence encoders are shown for all experimental conditions of disparity and vergence in Figure 4-10 to Figure 4-9. A robust, statistically significant change in planning period firing rates is observed for vergence only.
Figure 4-8 - Vergence Encoding Cell. Preference for “near” fixation Vergence ANOVA P = 1.4688e-013.
Figure 4-9 - Vergence Enocding Cell. This cell encodes vergence for only positive disparities. Vergence ANOVA P = 0.00071504. Preference for “far” fixation.
Figure 4-10 - Vergence Encoding cell with near fixation preference. No significant response to disparity; Vergence ANOVA P = 6.2172e-015