Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 11 (2), 216-23

Integrating Motion and Depth via Parallel Pathways

Affiliations

Integrating Motion and Depth via Parallel Pathways

Carlos R Ponce et al. Nat Neurosci.

Abstract

Processing of visual information is both parallel and hierarchical, with each visual area richly interconnected with other visual areas. An example of the parallel architecture of the primate visual system is the existence of two principal pathways providing input to the middle temporal visual area (MT): namely, a direct projection from striate cortex (V1), and a set of indirect projections that also originate in V1 but then relay through V2 and V3. Here we have reversibly inactivated the indirect pathways while recording from MT neurons and measuring eye movements in alert monkeys, a procedure that has enabled us to assess whether the two different input pathways are redundant or whether they carry different kinds of information. We find that this inactivation causes a disproportionate degradation of binocular disparity tuning relative to direction tuning in MT neurons, suggesting that the indirect pathways are important in the recovery of depth in three-dimensional scenes.

Figures

Figure 1
Figure 1
Effects of cooling on visually evoked activity in the anterior lunate bank. (a) Principal feedforward projections to MT. M, magnocellular pathways; P, parvocellular pathways; 4B, layer 4B; SS, spiny stellate cells; PYR, pyramidal cells. (b) Intra-operative photograph of cryoloop implants in monkey K. Image shows posterior bank of the lunate sulcus. Broken line shows the midline. Arrows point to the border between V1 and V2, which is evident from the difference in vascularity. Each cryoloop is numbered. Scale bar, 5 mm. D, dorsal; L, lateral. (c) Cut-away view of cryoloop placement along the posterior bank of the lunate sulcus. The surface of the brain is at the top; the fundus of the lunate sulcus is represented by the broken line at the bottom. Cryoloops were implanted in the lunate sulcus of the right hemisphere in three different monkeys and lowered to a temperature of 0–6 °C to inactivate all cortical layers. All distances are in millimeters. (d) Raster plot showing spontaneous and visually evoked multi-unit responses from layers 5/6 in the anterior lunate bank. Each row shows a single trial of the neural response to a light bar at the preferred orientation moving into the receptive field. Arrows indicate the time points when cooling pumps were turned on and off. (e) Mean multi- (MU) and single-unit (SU) responses from neurons in layers 5/6 of the anterior bank of the lunate sulcus. Curves show the firing frequency aligned at the onset and termination of the cooling pump cycle. Each point is the running average of three stimulus sweeps.
Figure 2
Figure 2
Effects of V2/V3 inactivation on the firing rates of MT neurons. (a) Responses of an individual MT neuron to its preferred stimulus at cooling onset (blue curve) and during rewarming (green line). (b) Population change in mean firing rate. Each point shows the response of a single recording site before and during cooling (mean ± s.e.m., n = 80). Inset, distribution of the blocking index, BI (median blocking index = 0.20). (c) Distribution of blocking index as a function of receptive field position in visual space. Each circle represents the location and approximate size of a hand-mapped response field; its color represents the absolute value of its corresponding blocking index.
Figure 3
Figure 3
Changes in direction tuning of MT neurons during V2/V3 inactivation. (a) Direction-tuning curves measured for a single unit in MT before (red), during (blue) and after (magenta) cooling of the indirect pathways. Mean ± s.e.m. responses are presented in polar coordinates; arrows indicate the mean vector of each tuning curve (vector lengths have been multiplied by maximum firing rate). (b) Preferred directions recorded before and during cooling (error bars indicate 95% CI, n = 74). Filled points represent neurons that showed statistically significant changes in preferred direction (P < 0.05, two-sample Watson-Williams test). (c) Changes in tuning bandwidth during cooling (mean ± s.e.m.). Points below and above the equality line represent neurons that became, respectively, more broadly and more narrowly tuned. Filled symbols indicate statistically significant changes in mean vector length (P < 0.05, Wilcoxon rank sum test, n = 74).
Figure 4
Figure 4
Comparison of changes in disparity and direction tuning of MT neurons during V2/V3 inactivation. (a) Paired sets of direction and disparity (mean ± s.e.m.) tuning curves recorded from the same sites before (red) and during (blue) cooling. ΔDI values show the changes in DI for direction and disparity (ΔDI = DIpre-cool − DIcool). (b) DI values before and during cooling for direction (left) or disparity (right). Open symbols, multi-unit sites; filled symbols, single-unit recordings (mean ± s.e.m., n = 41). Broken diagonal lines indicate the line of equality.
Figure 5
Figure 5
Quantitative and qualitative changes in tuning for direction and disparity during V2/V3 inactivation. (a,b) Direction- and disparity-tuning curves (mean ± s.e.m.) recorded from the same neuron before (red) and during (blue) cooling. The correlation coefficients (r2) of determination for direction and disparity were 0.99 and 0.54, respectively. (c) Correlation between pre-cooling and cooling tuning curves for direction and disparity. Each circle represents a recording site. Abscissa shows the r2 value for direction; ordinate shows the r2 value for disparity. Filled symbols indicate that both r2 values are significantly different from each other (P < 0.05, ref. 19).
Figure 6
Figure 6
Correlation of direction and disparity modulation-amplitude ratio distributions. Each point represents a recording site. Abscissa shows the modulation amplitude ratios (mean ± s.e.m.) computed for direction, ordinate shows the ratios for disparity (where modulation amplitude ratio = (RmaxRmin)cool/(RmaxRmin)pre-cool). Standard error bars were derived through bootstrapping (500 iterations). Gray circles indicate units that showed no significant change in amplitude ratio for either visual modality.
Figure 7
Figure 7
Bootstrap test of the global gain hypothesis. (a) Direction-tuning curves recorded before (red) and during (blue) cooling (mean ± s.e.m.). (b,c) Gain and offset distributions obtained by bootstrapping direction-tuning data (200 iterations). (d) Disparity-tuning curves recorded before (red) and during (blue) cooling (mean ± s.e.m.). Light blue traces are example bootstrap disparity-tuning curves calculated by resampling pre-cooling disparity data and applying direction gain and offset values sampled randomly from the distributions in b and c. (e) Predicted disparity DI distribution from bootstrap (1,000 iterations). Blue arrow indicates the actual DI value obtained during cooling. (f) Plot of predicted disparity DI against measured cooling disparity DI values. Each point represents a recorded site. Abscissa shows the mean predicted disparity DI values; ordinate shows the measured values (n = 41).
Figure 8
Figure 8
Effects of V2/V3 inactivation on disparity-dependent eye movements. Shown are vergence-tuning curves as a function of temperature condition. Normalized vergence velocity responses (mean ± s.e.m.) were plotted against stimulus binocular disparity before (red) and during (blue) cooling for monkeys J, K and M. Asterisks indicate responses that were significantly different, as determined by a multiple comparison procedure (see text).

Similar articles

See all similar articles

Cited by 45 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback