| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Report
Departments of 1Neurobiology and 2Biomedical Engineering, Washington University School of Medicine, St. Louis, Missouri 63110
Submitted 14 February 2003; accepted in final form 19 March 2003
 |
ABSTRACT |
|---|
|
|
|
INTRODUCTION |
|---|
|
;
Clement et al. 1986;
Himi et al. 1990;
Matsuo and Cohen 1984;
Snyder and King 1988). If
these asymmetries all derive from the same source, then they should be
dependent on the same variables. The low-frequency (long duration) components
of the VOR and OKN show asymmetries that cease or change direction when the
subjects are reoriented with respect to gravity or when the gravitational
field is altered (Angelaki and Hess
1994; Clarke et al.
2000; Clement et al.
1986,
1993;
Mittelstaedt and Mittelstaedt
1996; Raphan and Sturm
1991). By convention, we call these gravity-dependent asymmetries
allocentric. In contrast, we call any asymmetry that is independent of gravity
and maintains a constant relationship with the body an egocentric
asymmetry. The aim of this investigation was to test whether the vertical asymmetry in memory saccades is egocentric or allocentric. Because all previous studies of memory saccades were performed with subjects oriented upright, this upward asymmetry could be either egocentric or allocentric, or reflect an interaction between the two. Knowing the reference frame of the vertical memory saccade asymmetry will provide clues to its origin. If it is allocentric, then it is more likely to arise from a vestibular-related source; if it is egocentric, then a vestibular origin can be ruled out. We recorded memory saccades while animals were in one of four different orientations in space: upright, left-side-down (LSD), right-side-down (RSD), and supine. Evaluation of the magnitude and direction of the errors in these orientations demonstrated that the up/down asymmetry is egocentric, but the magnitude of the error does decrease in tilted orientations.
|
|
METHODS |
|---|
|
During experiments, animals were seated in a head-fixed primate chair inside a three-dimensional turntable (Acutronics) equipped with a three-field magnetic system (CNC Engineering). The animals' bodies were secured with shoulder and lap belts, whereas the extremities were loosely fixed to the chair. The chair and magnetic coils could be tilted in a variety of orientations relative to gravity. Saccades were made from four distinct orientations: upright, LSD, RSD, and supine. This configuration allowed dissociation between gravity and head coordinates, although prohibited dissociation between a head- and a body-centered frame of reference. Because the magnetic coils moved with the animal, the recorded eye movements were measured relative to the animals' head and body axes.
The memory saccade task is outlined in Fig. 1A. A laser projected a target on a screen 22 cm from the monkey at the center of his visual field in a completely dark room. After the monkey fixated on the central light for ≥1 s, a peripheral target light flashed for 200 ms. The flash appeared randomly at 1 of 16 possible locations (45° increments around a full circle with 15 or 20° eccentricity). The animal was required to maintain fixation on the central target and ignore the peripheral flash. The central target was turned off between 1.75 and 2.25 s after the peripheral flash, signaling the animal to make a saccade to the remembered location of the flash. The animal was required to hold fixation at the extinguished peripheral target location for 400 ms before the peripheral target was turned back on. The animal received a juice reward only after he satisfactorily fixated on the re-illuminated target for 400 ms. Behavioral windows were small for visual fixation (typically ±2°) but large (typically ±10°) for the memory fixation.
|
Only successful trials (i.e., trials in which a reward was delivered) were saved for off-line analyses. A Cambridge Electronics peripheral interface device (CED Power 1401) using Spike2 software-controlled stimulus presentation, behavioral control, and data acquisition (833.33 Hz, 16-bit resolution). Eye-position signals were anti-alias filtered (200 Hz, 6-pole Butterworth) and calibrated based on a daily horizontal/vertical fixation task. Eye velocity was calculated as the time derivative of eye position.
Saccade onset and offset were defined as the time when the magnitude of eye velocity first exceeded and fell below 25°/s, respectively. Horizontal and vertical pre- and postsaccadic positions were calculated by averaging eye position over a 20-ms period beginning 50 ms before and after saccade onset and offset, respectively. Horizontal and vertical saccade amplitudes were then calculated as the difference between pre- and postsaccadic eye positions.
Systematic error was defined on a trial-by-trial basis as the vector
difference between the target location and the eye position immediately after
the saccade. Variable error was computed as in White et al.
(1994) to indicate variability
within each group, where groups are defined as trials that share orientation,
direction, target eccentricity, and animal. Positive numbers were defined as
upward and rightward errors (relative to the animal) for the vertical and
horizontal components, respectively. Systematic errors were analyzed by
assuming that the asymmetry was either ego- or allocentric. We therefore
defined two distinct coordinate reference frames: allocentric with its
ordinate perpendicular to the ground, and egocentric with its ordinate
perpendicular to the horizontal (or transverse) body plane. Because these two
coordinate systems are identical with the animal upright, non-upright
orientations were used to dissociate the two. Specifically, for each side-down
orientation, the egocentric reference frame was rotated by 90° with
respect to the allocentric frame. We calculated the mean saccadic errors for
each animal in every orientation by averaging the appropriate systematic
errors. We then expressed the vertical component of the error in side-down
trials as a fraction of the vertical error in upright trials. As a final step,
we averaged these ratios across the three animals. Statistical comparisons
used ANOVA to ascertain that the upshift was influenced by position, and
Student's t-test to test the significance of individual
comparisons.
|
|
RESULTS |
|---|
|
;
White et al. 1994), the upward
memory-guided saccade overshot its target while the downward saccade undershot
its target. When the peripheral target was turned on later in each trial,
downward corrective visually guided saccades brought the eyes closer to the
target (Fig. 1B). To
determine whether this up/down asymmetry is egocentric (related to body or
head position) or allocentric (related to the direction of gravity), the
errors were evaluated in three different body orientations—LSD, RSD, and
supine—as well as upright. The systematic (lines) and variable (circles)
errors for two monkeys in the eight different 20° target location
eccentricities are plotted in Fig.
2, separately for each orientation. For both monkeys and at every
orientation, nearly every saccade landed a few degrees above the target. The
data are consistent for 15° targets as well, which are not shown for
simplicity. Because eye position is expressed relative to the head and body,
that is, in egocentric coordinates, it appears that the asymmetry is
egocentric. Had it been allocentric, the LSD lines would point to the right,
the RSD lines would point to the left, and supine lines would average to no
asymmetry at all.
|
The results for all three animals are summarized in Fig. 3. The up/down asymmetry persisted and remained positive (upward) in all four orientations. However, the upward systematic error decreased in tilted orientations in all three animals (although the difference was statistically significant only for animal A, Fig. 3A, top; white fills). In contrast to the vertical error, which was always upward, the horizontal error was small and idiosyncratic: to the left for animal A, to the right for animal B, and variable for animal C (Fig. 3A, bottom).
|
A quantitative summary is provided in Table 1 and Fig. 3B. The mean up/down asymmetry calculated in an egocentric frame of reference is two-thirds as large as the asymmetry in the upright position (egocentric ratio = 0.68). In contrast, the mean asymmetry calculated in the allocentric frame of reference is close to zero (allocentric ratio = 0.05). Figure 3B plots the mean systematic errors separately for each animal and orientation in both ego- and allocentric coordinates. The figure illustrates that the errors in side-down orientation were nearly identical to the errors for the upright orientation, when plotted using egocentric coordinates. However, when the same data are plotted using allocentric coordinates, the error in each side-down orientation clearly diverges from the error in the upright orientation.
|
|
|
DISCUSSION |
|---|
|
Our results from the side-down orientations are consistent with the asymmetry being egocentric (Fig. 3 and Table 1). In addition, the asymmetry persisted for animals in the supine position, which is also consistent with an egocentric frame of reference. However, the egocentric ratio was reduced to approximately two-thirds in the side-down positions, consistent with a second-order modulatory effect of position with respect to gravity on the up/down asymmetry.
Our results indicate that the vestibular system is minimally involved in
the vertical asymmetries seen in memory-guided saccades. Asymmetries that are
closely linked to the vestibular system, such as those seen in the VOR and
OKN, differ from the memory-guided saccade asymmetry in two important
respects. First, the low-frequency VOR and OKN effects are allocentric
(Angelaki and Hess 1994;
Clement et al. 1986;
Mittelstaedt and Mittelstaedt
1996; Raphan and Sturm
1991), whereas the memory saccadic effect is egocentric. Second,
the up/down asymmetry in the time constant and gain of the VOR and OKN,
although present in upright orientations, increases in tilted body positions
(Angelaki and Hess 1994;
Clarke et al. 2000;
Clement and Lathan 1991;
Matsuo and Cohen 1984;
Pettorossi et al. 1993;
van den Berg and Collewijn
1988), whereas the saccadic asymmetry decreases in tilted body
positions.
Findings from White et al.
(1994) indicate that the
up/down asymmetry in memory-guided saccades does not derive from the memory
system itself. In one of their tasks, monkeys initiated saccades as soon as
the target light appeared, but the target was extinguished 200 ms later. As a
result, the target was absent at the time that the saccade ended. The up/down
asymmetry in the saccade persisted despite the fact that the memory period was
extremely brief, leading these authors to suggest that memory per se was not
responsible for the asymmetry.
If asymmetries in memory saccades derive neither from the vestibular nor
the memory system, what is the cause of these errors? One possibility is that
the asymmetry reflects a strategic bias rather than an error. Consider three
other up/down asymmetries in the visual system. First, the highest density of
rods in the retina is found around the superior vertical meridian
(corresponding to lower visual field)
(Curcio and Allen 1990;
Packer et al. 1989;
Wikler and Rakic 1990;
Wikler et al. 1990). Second,
because rods are optimized for scotopic vision, dim targets viewed against a
dark background are appropriately placed in the lower visual field
(Barash et al. 1998). Third,
He, Cavanagh, and Intriligator
(1996) have shown that
attentional resolution is greater in the lower than the upper visual
field.
All three of these asymmetries— higher rod density in the superior
retina, a tendency to place dim targets in the lower visual field and better
attentional resolution in the lower visual field—suggest that a target
that has disappeared from sight against a dim or dark background would be more
easily found if it lay below rather than above the current fixation point.
This might explain why saccades to a target that has disappeared from view
would land above the best estimate of that target's location. Because the eyes
generally rotate with the head (ocular counter-roll is minimal at steady
state) (Haslwanter et al.
1992), these asymmetries are head-fixed, that is, egocentric.
Therefore the finding that the upward asymmetry for memory-guided saccades is
egocentric rather than allocentric is consistent with our hypothesis of a
strategic bias. When a target has disappeared from view, placing the fovea
slightly above the best estimate of its location may optimize the speed with
which the target can be reacquired.
|
|
ACKNOWLEDGMENTS |
|---|
|
|
|
FOOTNOTES |
|---|
Address for reprint requests: D. Angelaki, Dept. of Anatomy and Neurobiology, Box 8108, Washington University School of Medicine, 660 S. Euclid Ave., St. Louis, MO 63110 (E-mail: angelaki{at}pcg.wustl.edu).
|
|
REFERENCES |
|---|
|
Barash S,
Melikyan A, Sivakov A, and Tauber M. Shift of visual fixation dependent on
background illumination. J Neurophysiol
79: 2766–2781,
1998.
Chaudhuri A. Pursuit afternystagmus asymmetry in humans. Exp Brain Res 83: 471–476, 1991.[Medline]
Clarke AH, Grigull J, Mueller R, and Scherer H. The three-dimensional vestibule-ocular reflex during prolonged microgravity. Exp Brain Res 134: 322–334, 2000.[Medline]
Clement G and Lathan CE. Effects of static tilt about the roll axis on horizontal and vertical optokinetic nystagmus and optokinetic after-nystagmus in humans. Exp Brain Res 84: 335–341, 1991.[Web of Science][Medline]
Clement G, Popov KE, and Berthoz A. Effects of prolonged weightlessness on horizontal and vertical optokinetic nystagmus and optokinetic after-nystagmus in humans. Exp Brain Res 94: 456–462, 1993.[Web of Science][Medline]
Clement G, Vieville T, Lestienne F, and Berthoz A. Modifications of gain asymmetry and beating field of vertical optokinetic nystagmus in microgravity. Neurosci Lett 63: 271–274, 1986.[Web of Science][Medline]
Curcio CA and Allen KA. Topography of ganglion cells in human retina. J Comp Neurol 300: 5–25, 1990.[Web of Science][Medline]
Gnadt JW, Bracewell RM, and Andersen RA. Sensorimotor transformation during eye movements to remembered visual targets. Vision Res 31: 693–715, 1991.[Web of Science][Medline]
Haslwanter T, Straumann D, Hess BJ, and Henn V. Static roll and pitch in the monkey: shift and rotation of Listings plane. Vision Res 32: 1341–1348, 1992.[Web of Science][Medline]
He S, Cavanagh P, and Intriligator J. Attentional resolution and the locus of visual awareness. Nature 383: 334–337, 1996.[Medline]
Himi T, Igarashi M, and Takeda N. Effect of vestibule-cerebellar lesions on asymmetry of vertical optokinetic functions in the squirrel monkey. Acta Otolaryngol 109: 188–194, 1990.[Medline]
Matsuo V and Cohen B. Vertical optokinetic nystagmus and vestibular nystagmus in the monkey: up-down asymmetry and effects of gravity. Exp Brain Res 53: 197–216, 1984.[Web of Science][Medline]
Mittelstaedt M-L and Mittelstaedt H. The influence of otoliths and somatic graviceptors on angular velocity estimation. J Vestib Res 6: 355–366, 1996.[Medline]
Packer O, Hendrickson AE, and Curcio CA. Photoreceptor topography of the retina in the adult pigtail macaque (Macaca nemestrina). J Comp Neurol 288: 165–183, 1989.[Web of Science][Medline]
Pettorossi VE, Bruni R, Draicchio F, Errico P, Ferraresi A, and Santarelli R. Asymmetries in ocular and cervical responses to sagittal vestibular stimulations. Arch Ital Biol 131: 105–115, 1993.[Medline]
Raphan T and
Sturm D. Modeling the spatiotemporal organization of velocity storage in
the vestibuloocular reflex by optokinetic studies. J
Neurophysiol 66:
1410–1421, 1991.
Snyder LH, King
WM. Vertical vestibuloocular reflex in cat: asymmetry and adaptation.
J Neurophysiol 59:
279–298, 1988.
van den Berg AV and Collewijn H. Directional asymmetries of human optokinetic nystagmus. Exp Brain Res 70: 597–604, 1988.[Web of Science][Medline]
White JM, Sparks DL, and Stanford TR. Saccades to remembered target locations: an analysis of systematic and variable errors. Vision Res 34: 79–92, 1994.[Web of Science][Medline]
Wikler KC and Rakic P. Distribution of photoreceptor subtypes in the retina of diurnal and nocturnal primates. J Neurosci 10: 3390–3401, 1990.[Abstract]
Wikler KC, Williams RW, and Rakic P. Photoreceptor mosaic: number and distribution of rods and cones in the rhesus monkey. J Comp Neurol 297: 499–504, 1990.[Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  N. Li, M. Wei, and D. E. Angelaki Primate Memory Saccade Amplitude After Intervened Motion Depends on Target Distance J Neurophysiol, July 1, 2005; 94(1): 722 - 733. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
