Sparse modeling of auditory receptive fields
We have recently shown that Spectro-Temporal
Receptive Fields (STRFs) can be estimated from both multi-neuron clusters and
high-gamma power responses in human auditory cortex. Traditional methods for
estimating STRFs from single-unit recordings, such as spike-triggered-averages,
tend to be noisy and are less robust to other response signals such as local
field potentials. We use generalized linear models (GLM) and regularization
that penalizes non-zero coefficients, which results in a sparse solution to the
GLM. We have shown that the frequency-time structure of the STRF tends toward
grouping in different regions of frequency-time and we have demonstrated that
group sparsity-inducing penalties applied to GLM
estimates of STRFs reduces the background noise while preserving the complex
internal structure. We used this approach to identify an abrupt change in the
best frequency of estimated STRFs along posteromedial-to-anterolateral
recording locations along the long axis of Heschl’s gyrus. This change correlates well with a proposed
transition from core to non-core auditory fields in humans.
Neuroeconomics
Humans are faced every day with
economic decisions that involve choosing between different alternatives. Should I
invest in stocks versus bonds today? A critical question is where in the
brain is value computed? This is a fundamental question posed in
Neuroeconomics. Neuroeconomics is a relatively new field whose goal is to bridge brain
neuroscience with behavioral economics and decision-making. To answer questions related to simple choice
we have been recording
intracranially from patients undergoing study for medically intractable epilepsy. These patients
are actively engaged in cognitively demanding tasks, and provide the
opportunity to study the neural correlates of decision-making. A number of
areas of the brain compute different forms of value, and our particular focus
has been on the amygdala and orbitofrontal cortex (OFC). Point-process models
are developed to track the dynamics of the evolving value signal in single-unit and multi-unit activity.
Information flow
between the human amygdala and OFC
There is a consensus that value signals computed in the OFC
and amygdala play a critical role in the choice process. What is the nature and
direction of the flow of information between OFC and amygdala at the time of
decision-making? In order to study this question, we simultaneous record local field potentials from OFC and amygdala in human patients while they perform a simple food choice task. Although the interaction of these circuits has been
studied in animals, these studies examine the effective connectivity directly in
the human brain on a moment-by-moment basis. We apply spectral conditional Granger causality analysis to test the
direction of information flow. We have observed that influence from amygdala-to-OFC
is dominant prior to the revealed choice, with a small but significant OFC
influence on the amygdala earlier in the trial. Alpha band oscillation amplitudes
analyzed with the Hilbert-Huang transform reveal differences in choice
valence coincident with temporally specific amygdala influence on the OFC.
Effects of
methylphenidate (MPH) on decision-making
Impulsivity, a cardinal symptom of ADHD, has been linked to alterations within the dopamine system, and treated predominantly with psychostimulants that alter dopamine neurotransmission. Decision-making is often based on selecting what is thought
to be the most rewarding option. However, when faced with a choice between a
small but immediate reward and a larger but more delayed reward, humans and
animals tend to prefer the smaller reward if the difference in magnitude is
sufficiently small or if the delay to receive the larger reward is too long.
This seemingly irrational choice is thought to be the result of the subjective
value of the larger reward decaying as a function of delay, a process known as
temporal discounting. We have modeled the effects of psychostimulants on
inter-temporal choice using logistic probability models and hyperbolic
discounting functions. Under
control conditions the probability of choosing the smaller-sooner (SS) reward becomes
larger when the magnitude of the SS reward increases and the delay of the larger-longer
(LL) reward becomes longer. Generally, as the dose of MPH increases, the probability of choosing the SS reward decreases reflecting a decrease in
impulsivity. Increasing the dose further may also result in an increase in choosing
randomly. We hypothesize that MPH may reduce temporal discounting and
impulsivity by increasing dopamine levels in the striatum and prefrontal
cortex.
Effects of
methylphenidate (MPH) on task switching
Low doses of psychostimulants such as methylphenidate (MPH), which increase extracellular dopamine and norepinephrine by inhibiting their reuptake, are the most commonly used treatment for attention deficit hyperactivity disorder (ADHD). Therapeutic doses of these drugs may improve focused attention at the expense of hindering other cognitive functions, including the ability to adapt behavior in response to changing circumstances—cognitive flexibility. Cognitive flexibility is thought to depend on proper operation of the prefrontal cortex (PFC) and is also linked to reward processing, which is dopamine-dependent. Additionally, reward outcome signals have been recorded from the PFC. Methylphenidate impaired task-switching performance,which can be used as a measure of cognitive flexibility. This detriment may result from degraded outcome signaling within the PFC, and has implications for the use of MPH in the treatment of ADHD.
Auditory spatial attention in the superior colliculus
Previous behavioral work has demonstrated that rhesus monkeys are able to allocate attention about the surrounding space based on
brief, broadband auditory cues. We obtained single-unit recordings taken
from the intermediate layers of the superior colliculus (iSC) while the
subjects oriented to visual and auditory targets in the context of a
cuing task with their heads unrestrained. Our results show a correlation
between behavioral manifestations of attention allocation, attention
capture and inhibition of return, and modulation of target-evoked
responses in single iSC neurons. A linear combination of weighted Laguerre basis functions was used to estimate an impulse response relative to the target event that optimally predicts each trial’s neural response.
