239 research outputs found
Sort by
Perceptual decision influences V1 neuronal responses to ambiguous threedimensional objects
No full textWe studied spike responses of V1 superficial layer neurons in a perceptual decision task.
A rhesus monkey was trained to hold fixation during presentations of a three-dimensional
(structure-from-motion) object and to make a perceptual decisions in an alternative
choice paradigm while extracellular responses were obtained by single electrode
penetrations. The disparity of constituent dots was varied from trial-to-trial to render
perceptually ambiguous or unambiguous objects. Neurons with modulated disparity
responses were selected. We estimated the certainty at which the firing rate of a given V1
neuron would allow an ideal observer to predict the monkey's perceptual choice in the
task. Neuronal responses to zero-disparity (ambiguous) objects were sorted according to
the perceptual decision and the choice probability was determined for each neuron
(Britten et al., 1996). Based on the sample of n>100 neurons the firing rate ROC curves
showed significant bias from chance starting at 500 msec after the stimulus onset. The
choice probability was different from 0.5 for the significant majority of cells. The long
latency of the perceptual bias in the V1 responses suggests a feed-back from higher visual
cortical areas including MT/MST that further raises the question of V1's involvement in
perceptual awareness. (Supported by NEI and J.G. Boswell Professorship)
Acknowledgments: We thank to Christof Koch and Melissa Saenz for useful discussion
Recognition of non face objects, designed to require the same stimulus processing as that for faces, show only minimal effects of differences in contrast polarity or orientation direction
No full textPlease see attached pd
Motion deficits in dyslexia are restricted to high external noise displays
No full textStudies of motion perception in dyslexia have usually used random dot kinetograms with high external noise. Is the reported motion deficit in dyslexia due to deficiencies in
motion perception per se, or due to deficiencies in excluding noise in the displays? In this study, we compared the motion perception thresholds of both dyslexic and nondyslexic children, and dyslexic and non-dyslexic adults using first-order coherent motion displays that varied in noise level and signal salience. Both dyslexic children and adults
had higher motion thresholds than non-dyslexic children and adults when the task
involved first-order motion processing in high noise. Dyslexics performed as well as
non-dyslexics, however, when the signal was clearly separated from the noise or noise
was reduced. Thus dyslexics appear to have normal motion perception, but have
difficulty processing motion in high external noise. The ability to exclude noise or ignore
distractors while focusing on the what is relevant may play a role the creation of
appropriately flexible yet solid phonological and orthographic categories, a fundamental
process in learning to read
Characterizing and modeling temporal dynamics of perceptual decision making
No full textWe combined the external noise method (1) with the cue-to-respond speed accuracy trade-off
(SAT) paradigm (2) to characterize the temporal dynamics of perceptual decision making.
Observers were required to identify the orientation of one of eight briefly presented peripheral
Gabor targets (+/- 12 deg) in both zero and high noise. An arrow, occurring in the center of the
display cued the observer to the target location 234 ms before the onset of a brief target display;
an auditory beep, occurring at one of eight delays (SOA=25 to 800 ms) after the target onset,
cued the observers to respond. Five Gabor contrasts, spanning a wide range of performance
levels, were tested in each external noise condition. Increasing accuracy of discrimination (d')
was measured over processing times from 210 to 940 ms (as a function of SOA to the cue) in
each external noise and Gabor contrast condition. All ten SAT functions were well fit by
exponential functions with identical time constant and intercept but different asymptotic levels.
This suggests that, despite enormous variation in the external noise and contrast energy in the
stimulus, and in the ultimate accuracy of performance, information accumulated with the same
rate and starting time across all the external noise and contrast conditions. In addition, we
conducted a standard response time version of the experiment both before and halfway through
the SAT procedure. Data from the response time version of the experiment were all consistent
with the speed-accuracy trade-off data, but primarily differed in response accuracy. A simple
elaboration of the perceptual template model (3) with a dynamic decision process in which
information accumulates with the same rate but with step sizes proportional to the signal to noise
ratio in the perceptual representation of the visual input fully accounts for the results.
(1) Pelli, Dissertation; (2) Dosher, Cognitive Psychology'76; (3) Lu & Dosher, JOSA'99
Sine-wave grating adaptation selectively reduces the gain of the adapted stimulus
No full textAdapting to a sinusoidal grating selectively reduces contrast sensitivity to subsequent
stimuli of the adapted stimulus orientation and spatial frequency (Blakemore & Campbell,
1969). Both pre-synaptic and post-synaptic suppression have been proposed as the
cellular mechanism underlying reduced neuronal sensitivity in early visual cortex
following adaptation (Finlayson & Cynader, 1995; Sanchez-Vives, et al., 2000). At the
behavioral level, these two cellular mechanisms can be distinguished using an external
noise method and a noisy observer model (Dao, Lu, & Dosher, 2003). We studied
adaptation processes in two experiments; in the first, we measured adaptation effects at a
full range of white external noise levels; in the second, we measured adaptation effects in
orientation-filtered external noise. In both experiments, observers adapted to a 45 deg, 1
Hz counter-flickering sine grating of 0.8 contrast, then performed a two-interval forced
choice detection of a sine-wave grating oriented 45 deg in the presence of external noise.
In Experiment 1, we measured the impact of adaptation in six contrast levels of white
external noise. Based on an analysis of full psychometric functions in multiple white
external noise levels and at multiple stimulus contrasts, we concluded that adaptation
reduces the gain to the adapted stimulus and we predicted that the perceptual template is
changed about the adapted stimulus orientation. In Experiment 2, we tested this
prediction by measuring adaptation effects using orientation-filtered noise. This allowed
us to infer the impact of adaptation on the shape of the perceptual template. Nine
different external noise bandwidths were tested: +/- 0, 1, 5, 10, 15, 30, 45, 60, and 90 deg
about the center orientation of 45 deg. Compared to performance in unadapted control
conditions, adaptation led to a general elevation in threshold in all noise conditions.
Results from both experiments can be accounted for by the contrast gain control
Perceptual Template Model (cgcPTM), which is mathematically equivalent to the
Perceptual Template Model (PTM). The cgcPTM consists of perceptual templates in a
signal path and in a contrast gain control path, non-linearity transducer functions, contrast
gain control, pre and post gain control internal noise, and a decision structure. In this
framework, adaptation selectively reduces the gain of the perceptual templates around the
adapted orientation in both the signal and contrast gain control paths; adaptation does not
alter pre- or post gain control internal noise; and adaptation does not change transducer
non-linearity. These findings are consistent with post-synaptic changes following
adaptation at the neural level
A model of the mammalian muscle spindle
No full textProprioceptors such as muscle spindles and Golgi tendon organs provide the
central nervous system with sensory feedback for motor control and kinesthesia. It is
difficult to record afferent activity from such receptors during motor behavior, so theories
of motor control usually depend on implicit or explicit assumptions about such activity.
The muscle spindle is the most important proprioceptor, playing a dominant role
in kinesthesia and in reflexive adjustments to perturbations. Each muscle spindle
accurately senses and encodes length and velocity information of the extrafusal muscle
fibers over a wide range of movements despite the relatively restricted dynamic range of
firing rates for action potentials. It does this by shifting the relative importance and
sensitivity to length and velocity in response to specialized fusimotor efferents (gamma
motoneurons), at the cost of complicating the interpretation of their signals by the
nervous system. We have constructed a physiologically realistic model of the spindle that
is composed of mathematical elements closely related to the anatomical components
found in the biological spindle. The spindle model consists of three nonlinear intrafusal
fiber models: bag1, bag2 and chain. The bag1 fiber model is the only one that receives
dynamic fusimotor control and is primarily responsible for velocity sensitivity of the
spindle. The bag2 and chain receive ?static fusimotor control and contribute mainly to
length sensitivity. All three fiber types give rise to primary afferent activity, while only
bag2 and chain to secondary afferent activity. In the case of the primary afferent, the
model incorporates the experimentally observed effect of partial occlusion, where
primary afferent activity results from a competition between two impulse generator sites,
one located on the bag1 and other on bag2 and chain fibers. When both sites are active,
the dominant generator wins and suppresses all activity in the weaker generator by resetting
its spike generator. While that results in total occlusion, the mechanism
responsible for partial occlusion observed in the case of primary afferent is believed to
include electrotonic current spread between the suppressed and dominant generator,
resulting in increased impulse generation at the dominant site. The model also
incorporates the appropriate temporal properties of three types of intrafusal fibers during
static or dynamic fusimotor stimulation. The advantage of including these properties is
demonstrated by comparing model simulations with and without these properties to data
from recently published experiments in which both fusimotor efferent and spindle
afferent activity were recorded simultaneously during decerebrate locomotion in the cat
(Taylor et al., J Physiol 529.3: 825-836, 2000). We have inverted the spindle model in
order to use it as a tool to better understand fusimotor control in natural tasks. By
supplying the inverted model with records of afferent activity and kinematics during
natural tasks, the inverted model can be used to infer the underlying fusimotor drive.
Once the principles of fusimotor control are understood, it should be possible to apply the
spindle model to predict more accurately the activity of spindle afferents and their role in
control of motor tasks
Geometric Analysis of the Axo-Dendritic Interface in Neocortical Pyramidal Neurons
No full textSince the time of Hebb, the physical substrate for learning and memory in the brain has been most often discussed in relation to activity-dependent synaptic "weight" changes
mediated by LTP or LTD. However, two recent theoretical studies suggest that long term information storage in neural tissue could also depend (heavily) on structural plasticity at
the interface between axons and dendrites (Poirazi & Mel, 2001; Stepanyants et al. 2002).
According to both theories, the capacity for structure-based information storage depends
on the interaccessibility of afferent axons and their dendritic targets within the neuropil.
For example, how many different axons are likely to be accessible to any given postsynaptic
dendrite with only minor structural modification? How much overlap exists in
the set of axons accessible to two different dendritic branches? Using axonal and
dendritic arborizations of a reconstructed pyramidal neuron from cat visual cortex as a
reference point (courtesy J. Hirsch), we quantified the tradeoffs among several
morphological variables that parameterize the axo-dendritic interface in neocortex,
including spine length, spine density, dendritic branch length, branches per neuron, etc.
We then used an extended version of the formula derived in Poirazi & Mel (2001) to
understand how each of these variables separately and together contribute to the tissue's
capacity to learn
Factors contributing to Post-Traumatic Dentate Hyperexcitability: A Network Model Incorporating Topographic Connectivity Patterns
No full textHead injury is a major risk factor in the etiology of temporal lobe epilepsy (TLE). Studies using a rodent model of concussive head trauma have identified specific patterns of cell loss and synaptic reorganization in the dentate gyrus after brain injury, which are similar to the changes in human TLE. However, the contribution of each of these cellular and synaptic alterations to increased excitability in the dentate neuronal circuits is not known.
In order to independently examine the factors critical to post-traumatic dentate bhyperexcitability, we developed a reduced network model of the dentate gyrus with 500
granule cells, 15 mossy cells 6 basket cells and 6 hilar interneurons. The topographic
networks were constructed with connectivity patterns constrained by the spatial distribution
of the axonal arbors of the cell types. Sprouting was simulated by addition of mossy fiber
to granule cell connections with the maximum sprouting (100%) estimated from the
distribution of sprouted axons in a rodent model of spontaneous recurrent seizures
(Buckmaster and Dudek 1999). Simulations were performed using NEURON (Hines
1993). Our results show that perforant path stimulation evoked greater granule cell firing in
the dentate excitatory network with as low as 10% sprouting compared to the control
topographic network. Additionally, the topographic network was more hyperexcitable than
the non-topographic network with the same degree of sprouting. Mossy cell loss decreased
the spread of hyperexcitability in the network 10% sprouting. With increasing sprouting,
even the complete loss of mossy cells was unable to prevent the spread of hyperexcitability.
Simulations of both purely excitatory network and the full network showed that mossy
fiber sprouting was sufficient to elicit hyperexcitable perforant path evoked responses in all
cell types examined. Mossy cell loss was neither necessary nor sufficient to cause granule
cell hyperexcitability in the dentate network with inhibition.
The network simulations show that mossy fiber sprouting can contribute to
increased excitability in the dentate gyrus even in the absence of cell loss or changes in the
intrinsic properties of the cells. The data from the topographically constrained simulations
indicate that the lamellar topology of the sprouted mossy fibers is important for the spread
of granule cell excitability. The results suggest that the moderate sprouting observed after
concussive head trauma is likely to be a major factor in post-traumatic dentate
hyperexcitability.
Acknowledgment: Supported by the NIH (NS35915) to I.S
Ruby: A Robotic Platform for Real-time Social Interaction
No full textThe majority of our waking hours are spent engaging in social interactions. Some of these interactions occur at the level of long-term strategic planning while others
take place at faster time scales, such as in conversations or card games. The abilityto perceive subtle gestural, postural, and facial cues, in addition to verbal language,
in real-time is a critical component. An understanding of the underlying perceptual
primitives that support this kind of real-time social cognition is key to understanding
social development.
Robots present an ideal opportunity to study the development of social interaction
in infants [Fasel,Deak,Triesch,Movellan 2002]. It is possible to create robots that
exhibit precisely controlled contingency structures. By observing how infants interact
with these robots we gain an opportunity to understand how infants identify the
operating characteristics of the social agents with whom they interact.
We have recently developed a social interaction robot, "Ruby", designed to communicate
with children. Ruby is endowed with the following real-time perceptual
primitives to facilitate social interaction: face tracking, motor control and speech
detection. It communicates via head and eye movements and we have recently run
pilot studies indicating that Ruby is fun and non-threatening to children.
Ruby's face tracking system consist of 3 cues taken from 3 inputs. The first 2 inputs
are high-resolution pan-tilt-zoom color cameras which are the "eyes". The third
input is an omni-directional camera acting as Ruby's peripheral vision. Each eye
uses the MPLab's contrast-feature based frontal face finder [Fasel et al CVIU2004]
and adaptive color-based tracker [Ishiguro et al 2003] [Hershey et al CVPR2004].
Ruby combines both of these to find both frontal and rotated faces at more than
30 frames per second. Ruby's motor control system currently has 3 components;
neck control, eye control, and control of external objects for experiments. Ruby
also features speech detection [Pellom 2004] and response with variable delay parameters.
We are now adding eye and eye-blink detection[Fasel et al CVIU2004],
expression recognition[Littlewort-Ford, Bartlett et al 2004], recognition of common
communicative words in English, arm movements, finger pointing, and touch sensors.
We hope to use Ruby to collect and analyze data on social interaction and contingency
and on the development of social interaction in infants
