Limits...
Role of visual and non-visual cues in constructing a rotation-invariant representation of heading in parietal cortex.

Sunkara A, DeAngelis GC, Angelaki DE - Elife (2015)

Bottom Line: Here we demonstrate that the brain implements an alternative solution in which retinal velocity patterns are themselves used to dissociate translations from rotations.These results reveal a novel role for visual cues in achieving a rotation-invariant representation of heading in the macaque ventral intraparietal area.These findings further suggest that the brain is capable of performing complex computations to infer eye movements and discount their sensory consequences based solely on visual cues.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, United States.

ABSTRACT
As we navigate through the world, eye and head movements add rotational velocity patterns to the retinal image. When such rotations accompany observer translation, the rotational velocity patterns must be discounted to accurately perceive heading. The conventional view holds that this computation requires efference copies of self-generated eye/head movements. Here we demonstrate that the brain implements an alternative solution in which retinal velocity patterns are themselves used to dissociate translations from rotations. These results reveal a novel role for visual cues in achieving a rotation-invariant representation of heading in the macaque ventral intraparietal area. Specifically, we show that the visual system utilizes both local motion parallax cues and global perspective distortions to estimate heading in the presence of rotations. These findings further suggest that the brain is capable of performing complex computations to infer eye movements and discount their sensory consequences based solely on visual cues.

Show MeSH

Related in: MedlinePlus

Problems with previous approaches to measuring shifts in the absence of full tuning curve measurements.Previous studies (Bradley et al., 1996; Shenoy et al., 1999; Shenoy et al., 2002; Bremmer et al., 2010; Kaminiarz et al., 2014) evaluated heading tuning curves in a narrow range around straight ahead (white region around 90° in panel A). (A) Two hypothetical tuning curves with different bandwidths and amplitudes. (B) Correcting for the difference in response amplitudes reveals a clear difference in bandwidths, which may reflect a lack of compensation for rotation. The rank order of the responses (1–1, 2–2, 3–3) would be identical for the two curves, which previous methods would erroneously interpret as evidence for rotation-invariance (Bremmer et al., 2010; Kaminiarz et al., 2014). (C) Simulated tuning curves with peaks roughly near straight ahead. The underlying von Mises functions (to which Poisson noise was added) have peaks at headings of 80°, 100° and 120°, resulting in a simulated shift of 20°. (D) The cross-correlation function between the translation only (black) and rotation-added (red—leftward rotation, blue—rightward rotation) tuning curves. Reflecting the true shift of 20° that was introduced into the tuning curves (before noise was added), the cross-correlation functions peak near a lag of +20°. (E) Simulated tuning curves with peaks at 180° and different bandwidths. The difference in the width of the full tuning curves corresponded to a 20° shift (at half-height). (F) Cross-correlation functions for the simulated neurons with lateral heading preferences are quite flat and show no evidence of a peak near a lag of +20°. (G) Shifts from 10 sets of simulated tuning curves (with different noise samples) were measured using cross-correlation for heading preferences ranging from 0° to 180°. The gain and offsets of the tuning curves were randomized for each set. The mean shift (black markers) approaches the true shift (20°, dashed line) for tuning curves with heading preferences near 90°, but the mean shifts are grossly inaccurate for simulated neurons with lateral heading preferences. In contrast, Figure 4–figure supplement 1E shows that our analysis method correctly estimates tuning curve shifts regardless of heading preference.DOI:http://dx.doi.org/10.7554/eLife.04693.012
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4337725&req=5

fig4s2: Problems with previous approaches to measuring shifts in the absence of full tuning curve measurements.Previous studies (Bradley et al., 1996; Shenoy et al., 1999; Shenoy et al., 2002; Bremmer et al., 2010; Kaminiarz et al., 2014) evaluated heading tuning curves in a narrow range around straight ahead (white region around 90° in panel A). (A) Two hypothetical tuning curves with different bandwidths and amplitudes. (B) Correcting for the difference in response amplitudes reveals a clear difference in bandwidths, which may reflect a lack of compensation for rotation. The rank order of the responses (1–1, 2–2, 3–3) would be identical for the two curves, which previous methods would erroneously interpret as evidence for rotation-invariance (Bremmer et al., 2010; Kaminiarz et al., 2014). (C) Simulated tuning curves with peaks roughly near straight ahead. The underlying von Mises functions (to which Poisson noise was added) have peaks at headings of 80°, 100° and 120°, resulting in a simulated shift of 20°. (D) The cross-correlation function between the translation only (black) and rotation-added (red—leftward rotation, blue—rightward rotation) tuning curves. Reflecting the true shift of 20° that was introduced into the tuning curves (before noise was added), the cross-correlation functions peak near a lag of +20°. (E) Simulated tuning curves with peaks at 180° and different bandwidths. The difference in the width of the full tuning curves corresponded to a 20° shift (at half-height). (F) Cross-correlation functions for the simulated neurons with lateral heading preferences are quite flat and show no evidence of a peak near a lag of +20°. (G) Shifts from 10 sets of simulated tuning curves (with different noise samples) were measured using cross-correlation for heading preferences ranging from 0° to 180°. The gain and offsets of the tuning curves were randomized for each set. The mean shift (black markers) approaches the true shift (20°, dashed line) for tuning curves with heading preferences near 90°, but the mean shifts are grossly inaccurate for simulated neurons with lateral heading preferences. In contrast, Figure 4–figure supplement 1E shows that our analysis method correctly estimates tuning curve shifts regardless of heading preference.DOI:http://dx.doi.org/10.7554/eLife.04693.012

Mentions: Because of these changes in tuning curve bandwidth or shape, analysis of the effects of rotation on heading tuning requires more complex and rigorous approaches (Figure 4—figure supplement 1) than the cross-correlation or rank-order methods used in previous studies (Bradley et al., 1996; Shenoy et al., 1999, 2002; Bremmer et al., 2010; Kaminiarz et al., 2014). It is also critical to distinguish between changes in response gain and changes in the shape (Figure 4—figure supplement 2; see Discussion) of tuning curves, which our analysis allows because we sample the entire heading tuning curve (Mullette-Gillman et al., 2009; Chang and Snyder, 2010; Rosenberg and Angelaki, 2014). As shown in Figure 4—figure supplement 1, the first step in the analysis involves normalizing each RP and SP tuning curve to match the dynamic range of the pure translation tuning curve. Following this transformation, the change in the shape of the RP and SP tuning curves can be measured without ambiguity. To account for the expected changes in bandwidth and skew, partial shifts of the tuning curve were measured separately for forward (0°:180°) and backward (180°:360°) headings. Thus, four shift values were obtained from each neuron for both real and simulated pursuit, corresponding to forward/backward headings and left/right rotation directions. These four values were averaged for each neuron to quantify the transformation in shape and obtain one shift metric for RP tuning curves and one for SP tuning curves (see ‘Materials and methods’, Figure 4—figure supplement 1).


Role of visual and non-visual cues in constructing a rotation-invariant representation of heading in parietal cortex.

Sunkara A, DeAngelis GC, Angelaki DE - Elife (2015)

Problems with previous approaches to measuring shifts in the absence of full tuning curve measurements.Previous studies (Bradley et al., 1996; Shenoy et al., 1999; Shenoy et al., 2002; Bremmer et al., 2010; Kaminiarz et al., 2014) evaluated heading tuning curves in a narrow range around straight ahead (white region around 90° in panel A). (A) Two hypothetical tuning curves with different bandwidths and amplitudes. (B) Correcting for the difference in response amplitudes reveals a clear difference in bandwidths, which may reflect a lack of compensation for rotation. The rank order of the responses (1–1, 2–2, 3–3) would be identical for the two curves, which previous methods would erroneously interpret as evidence for rotation-invariance (Bremmer et al., 2010; Kaminiarz et al., 2014). (C) Simulated tuning curves with peaks roughly near straight ahead. The underlying von Mises functions (to which Poisson noise was added) have peaks at headings of 80°, 100° and 120°, resulting in a simulated shift of 20°. (D) The cross-correlation function between the translation only (black) and rotation-added (red—leftward rotation, blue—rightward rotation) tuning curves. Reflecting the true shift of 20° that was introduced into the tuning curves (before noise was added), the cross-correlation functions peak near a lag of +20°. (E) Simulated tuning curves with peaks at 180° and different bandwidths. The difference in the width of the full tuning curves corresponded to a 20° shift (at half-height). (F) Cross-correlation functions for the simulated neurons with lateral heading preferences are quite flat and show no evidence of a peak near a lag of +20°. (G) Shifts from 10 sets of simulated tuning curves (with different noise samples) were measured using cross-correlation for heading preferences ranging from 0° to 180°. The gain and offsets of the tuning curves were randomized for each set. The mean shift (black markers) approaches the true shift (20°, dashed line) for tuning curves with heading preferences near 90°, but the mean shifts are grossly inaccurate for simulated neurons with lateral heading preferences. In contrast, Figure 4–figure supplement 1E shows that our analysis method correctly estimates tuning curve shifts regardless of heading preference.DOI:http://dx.doi.org/10.7554/eLife.04693.012
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4337725&req=5

fig4s2: Problems with previous approaches to measuring shifts in the absence of full tuning curve measurements.Previous studies (Bradley et al., 1996; Shenoy et al., 1999; Shenoy et al., 2002; Bremmer et al., 2010; Kaminiarz et al., 2014) evaluated heading tuning curves in a narrow range around straight ahead (white region around 90° in panel A). (A) Two hypothetical tuning curves with different bandwidths and amplitudes. (B) Correcting for the difference in response amplitudes reveals a clear difference in bandwidths, which may reflect a lack of compensation for rotation. The rank order of the responses (1–1, 2–2, 3–3) would be identical for the two curves, which previous methods would erroneously interpret as evidence for rotation-invariance (Bremmer et al., 2010; Kaminiarz et al., 2014). (C) Simulated tuning curves with peaks roughly near straight ahead. The underlying von Mises functions (to which Poisson noise was added) have peaks at headings of 80°, 100° and 120°, resulting in a simulated shift of 20°. (D) The cross-correlation function between the translation only (black) and rotation-added (red—leftward rotation, blue—rightward rotation) tuning curves. Reflecting the true shift of 20° that was introduced into the tuning curves (before noise was added), the cross-correlation functions peak near a lag of +20°. (E) Simulated tuning curves with peaks at 180° and different bandwidths. The difference in the width of the full tuning curves corresponded to a 20° shift (at half-height). (F) Cross-correlation functions for the simulated neurons with lateral heading preferences are quite flat and show no evidence of a peak near a lag of +20°. (G) Shifts from 10 sets of simulated tuning curves (with different noise samples) were measured using cross-correlation for heading preferences ranging from 0° to 180°. The gain and offsets of the tuning curves were randomized for each set. The mean shift (black markers) approaches the true shift (20°, dashed line) for tuning curves with heading preferences near 90°, but the mean shifts are grossly inaccurate for simulated neurons with lateral heading preferences. In contrast, Figure 4–figure supplement 1E shows that our analysis method correctly estimates tuning curve shifts regardless of heading preference.DOI:http://dx.doi.org/10.7554/eLife.04693.012
Mentions: Because of these changes in tuning curve bandwidth or shape, analysis of the effects of rotation on heading tuning requires more complex and rigorous approaches (Figure 4—figure supplement 1) than the cross-correlation or rank-order methods used in previous studies (Bradley et al., 1996; Shenoy et al., 1999, 2002; Bremmer et al., 2010; Kaminiarz et al., 2014). It is also critical to distinguish between changes in response gain and changes in the shape (Figure 4—figure supplement 2; see Discussion) of tuning curves, which our analysis allows because we sample the entire heading tuning curve (Mullette-Gillman et al., 2009; Chang and Snyder, 2010; Rosenberg and Angelaki, 2014). As shown in Figure 4—figure supplement 1, the first step in the analysis involves normalizing each RP and SP tuning curve to match the dynamic range of the pure translation tuning curve. Following this transformation, the change in the shape of the RP and SP tuning curves can be measured without ambiguity. To account for the expected changes in bandwidth and skew, partial shifts of the tuning curve were measured separately for forward (0°:180°) and backward (180°:360°) headings. Thus, four shift values were obtained from each neuron for both real and simulated pursuit, corresponding to forward/backward headings and left/right rotation directions. These four values were averaged for each neuron to quantify the transformation in shape and obtain one shift metric for RP tuning curves and one for SP tuning curves (see ‘Materials and methods’, Figure 4—figure supplement 1).

Bottom Line: Here we demonstrate that the brain implements an alternative solution in which retinal velocity patterns are themselves used to dissociate translations from rotations.These results reveal a novel role for visual cues in achieving a rotation-invariant representation of heading in the macaque ventral intraparietal area.These findings further suggest that the brain is capable of performing complex computations to infer eye movements and discount their sensory consequences based solely on visual cues.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, United States.

ABSTRACT
As we navigate through the world, eye and head movements add rotational velocity patterns to the retinal image. When such rotations accompany observer translation, the rotational velocity patterns must be discounted to accurately perceive heading. The conventional view holds that this computation requires efference copies of self-generated eye/head movements. Here we demonstrate that the brain implements an alternative solution in which retinal velocity patterns are themselves used to dissociate translations from rotations. These results reveal a novel role for visual cues in achieving a rotation-invariant representation of heading in the macaque ventral intraparietal area. Specifically, we show that the visual system utilizes both local motion parallax cues and global perspective distortions to estimate heading in the presence of rotations. These findings further suggest that the brain is capable of performing complex computations to infer eye movements and discount their sensory consequences based solely on visual cues.

Show MeSH
Related in: MedlinePlus