|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Seminar
Title: "Fast and Flexible Action Planning in Sensorimotor Systems"
Presenter: Timothy Verstynen, Ph.D.
Location: A53 Baker Hall
Abstract:
Within a hundred milliseconds, perceptual and motor systems can reliably evaluate sensory cues and make appropriate action decisions (e.g., “go” vs. “no-go”). Once an action is selected, these systems take advantage of statistical regularities in the recent past in order to maximize the precision of the current motor command. Using a combination of psychophysics, neuroimaging, and computational modeling, I will highlight the dynamics and neural underpinnings of two dynamic planning processes that bridge perception and action: (1) response selection and (2) movement planning. First, behavioral assessments of inhibitory control reveal that skilled motor performance can be predicted by the precision of perceptual decision-making abilities in an unrelated task. These rapid decision making abilities are associated with the integrity of limbic and executive segments of the cortico-basal ganglia system. Breakdowns in the structure and function of these networks, revealed using multiple neuroimaging measures, lead to impairments in executive control, as evidenced in obese individuals. Second, after a response has been selected, sensorimotor planning systems use the statistics of recent actions to guide the current motor plan. This adaptive process follows the Bayesian principle of a variance-bias tradeoff and can be explained by the presence of associative learning in motor planning networks. In fact, this associative learning model not only generates novel psychophysical predictions but can also predict the dynamics of activity sensorimotor networks. Taken together, this work illustrates that fast and adaptive processes at the cusp of perception and action can underlie the efficiency of subsequent motor performance.
Seminar
Title: Heterogeneity among hair cell synapses: Differences in Threshold Sensitivity and Vulnerability to Noise
Presenter: M. Charles Liberman, PhD.
Location: Room 1495 BST-1 Building
Seminar
Title: Visual Prostheses for the Blind
Presenter: Shawn K. Kelly, PhD.
Location: Scaide Auditorium at CMU
Refreshments at 4pm.
Abstract:
Several groups worldwide are working to develop visual
prostheses to restore useful functional vision to the blind. These
devices deliver electrical stimulation to nerves in the visual pathway
(or sometimes other areas, like the tongue) to create the
representation of pixelated images.
The Boston Retinal Implant Project is developing and testing a
chronically-implantable subretinal visual prosthesis that will
stimulate retinal nerves of patients blinded by retinitis pigmentosa
and macular degeneration. We showed the viability of the concept in a
series of six acute epiretinal stimulation trials with blind
volunteers between 1998 and 2000. We determined that a subretinal
approach provides the safest and most stable placement for the
stimulating electrode array and electronics, and have since pursued a
chronically-implantable subretinal prosthesis that receives power and
data wirelessly.
We have developed several generations of prostheses, progressively
adding functionality, increasing the number of channels, and
assembling the implants into titanium hermetic cases. We have
surgically implanted three generations of prostheses into Yucatan
minipigs, for up to one year, and are actively developing a device
suitable for clinical trials, which we expect to begin within 2-3
years. Progress made and hurdles faced by the Boston Retinal Implant
and the visual prosthesis field will be discussed.