Quizlet Questions Flashcards
(89 cards)
1
Q
What are types of brain lesions?
A
Aspiration lesion - literally take out/suck out parts of the brain (vacuum cleaner).
Excitotoxic lesion - using drugs to either inhibit or activate parts of the brain.
Directed lesions
Naturally occurring lesions
- UPSIDES: can strongly implicate a region as being essential for a task, occur naturally
- DOWNSIDES: Need double dissociation to strongly confirm selectivity, no temporal resolution, and relatively few subjects
Intracranial stimulation
- Can provide specific neural perturbation, BUT limited to animals
2
Q
What is an antagonist, or agonist drug?
A
An antagonist drug: bind to and block receptors
An agonist drug: bind to and activate receptors, mimicking neurotransmitter (more reuptake)
3
Q
What is temporal and spatial resolution?
A
Temporal resolution refers to how closely the measured activity corresponds to the timing of the actual neuronal activity. The temporal resolution with PET is poor compared to both fMRI, EEG and MEG,
Spatial resolution refers to how accurately the measured activity is localised within the brain.
4
Q
What are types of single-neuron electrical recordings?
A
Extracellular: information about small groups of nerve cells
Intracellular - how single cells behave during cognitive functions.
5
Q
What is fMRI?
A
Functional magnetic resonance imaging. It is an indirect measure of brain activity. It measures BOLD signals (Blood Oxygen Level Dependent). It is low temporal (2-6 seconds) and spatial resolution, but most human friendly. It divides the brain into voxels (1-5 mm), each voxel has millions of neurons and tens of billions of synapses.
fMRI is good for analyzing activation patterns within a brain area. It can also look at coactivation, in which two or more brain regions change similarly in response to an experimental condition.
6
Q
What is a PSTH?
A
Peri-stimulus time histogram. It counts the spikes after aligning the spikes, temporally, relative to an event of interest. This measures the change in activity over time, at a marked moment of external stimulus/interest.
7
Q
Where is the perception of motion? How are those neurons specialized?
A
In the middle temporal (MT), V5 - middle temporal visual area specifically.
It is an area known to be best driven by visual motion stimulus
MT neurons are specialized for receptive field (which is anchored to the retina - each neuron is mapped onto the retina!) and direction of motion (cells have preferred motion direction)
8
Q
What is the ventral stream, what is the dorsal stream, and what do those relate to?
A
This is how our brain processes visual information. Ventral stream (to temporal lobes - asks WHAT); dorsal stream (parietal lobes - asks WHERE + HOW (action oriented))
9
Q
Random dot motion correlogram?
A
Discrimination with correlation/coherence - dots moving all in the same direction, or some random. This is a good task for testing visual information processing.
10
Q
What is a psychometric curve?
A
A psychometric curve measures behavior.
11
Q
What is a neurometric function? What are the key assumptions with a neurometric function? (include essential vocab)
A
A neurometric function measures the firing rates of a single neuron - it reflects what the neuron itself is “perceiving.”
The firing rate is essential:
- assume this neuron has all the information in its firing rate, that this neuron is “the ideal observer”
- Assume the neuron votes ‘yes’ when monkeys perceived motion stimuli in the preferred direction of this neuron, it votes ‘yes’ by increasing its firing rate
12
Q
How do you make a neurometric function?
A
We measure Hit rate vs. False Alarm rate.
The criterion is the specific firing rate - at which this neuron votes ‘yes’
Question is - how clear/unambiguous is the information that the neuron is passing along?
If the criterion is set too low, this neuron does poorly, since the two distributions (the firing rate distribution for the preferred motion and the firing rate distribution for the anti-preferred motion direction) are largely overlapping.
As the criterion is set higher, this neuron does better since the two distributions start to separate more.
If the criterion is very high, this neuron does very well since the two distributions don’t overlap much.
The method: change the criterion (spike/s) to get Hit rate and False Alarm rate. Plot Hit vs. False Alarm rate, to make the ROC (receiver operating characteristic curve)
Then, from there, you make a neurometric function, mapping it onto the x = correlation/y = proportion correct graph space. Here, each dot represents the area under the curve from the ROC.
The goal is to directly compare the psychometric (behavior) and neurometric (neuron’s view) functions directly.
13
Q
What is the relationship between a psychometric curve and a neurometric function? What is the purpose of mapping those onto the same space?
A
This is to compare whether a single neuron can do a sufficient job at describing the perceived motion (or worse? Or better?) The result - some neurons are more, or less, or about the same sensitivity as the overall perception of the monkey.
Puts the behavior and neuronal activity in the same space, to figure out whether the neuron is the ideal observer.
can assess how closely neural activity correlates with behavioral outcomes. A strong correlation suggests that the neural activity is relevant to the behavioral task.
14
Q
How would you design a study to ask the central question - does MT cells reflect perceived motion, beyond just visually-driven motion? Can neurons tell us what subjects perceived in motion direction? What do you expect to find?
A
Need to create a graphing space upon which you can map the subject’s/animal’s behavior onto the same space as a neuron’s firing rate. So, create a psychometric function and a neurometric function.
Experiment: use random dot motion correlogram.
Result: some neurons are more, or less, or about the same sensitivity as the overall perception of the monkey.
15
Q
How would you test if we can causally induce motion perception?
A
Microstimulate to MT neurons, to test whether these microstimulations can cause changes/biases in the subjective perception
- For instance, weaker motion becomes more sufficient to perceive the right direction
16
Q
How/where does the brain perceive faces? Why is that so important?
A
Face cells (which are primarily excited when they perceive a combination of face feature dimensions) are in the inferior temporal cortex (IT).
Face perception is so important because evolutionarily, faces carry the most important social and emotional cues for guiding survival and reproduction. So, it makes sense that we have a specialized neural system!
17
Q
What are face patches? What excites them?
A
Face patches are concentrated clusters of mostly face cells in the IT (inferior temporal). They are excited by a combination of critical features - a combination of two black circles and a horizontal line.
18
Q
What is the fusiform face area? What happens if you stimulate the fusiform face area?
A
It is a small region found on the inferior surface of the temporal lobe, which is specialized for facial recognition. Stimulating the fusiform face area impairs ability to perform match-to-sample tasks with faces.
19
Q
The first electrophysiology studies of face patches in the temporal cortex showed that neurons specifically fire when monkeys view faces. How would one test whether these neurons are causally involved in face recognition? Describe a simple experiment to test this idea (task, method, and results).
A
One way to test the causal role of these face patches is to microstimulate the face patches when animals are performing a face identify match-to-sample task. If a given face patch is causally involved in face perception, I would expect that the microstimulation applied either during the sample stimulus period or the test stimulus period would disrupt the processing of face. This disruption would be behaviorally reflected in a lower performance level of the animals in matching the identity.
19
Q
What are Jennifer Aniston cells? Where are they found? Would a Jennifer Aniston cell always encode Jennifer Aniston (forever?) What is going on in these cells?
A
Jennifer Aniston cells are found in the human medial temporal lobe (MTL), found in the hippocampus. They record high spiking activity whenever presented with Jennifer Aniston (sketching, the name, a picture, a group of Friends). This is evidence of invariant (viewing angle and size) representation of person identity - it proves the existence of concept cells, with specific identities! Though it does not imply that the Jennifer Aniston cells will encode Jennifer Aniston forever - perhaps next week will encode something different.
Still a live question - how do these concept cells get formed?
20
Q
What do concept cells tell us about the way the MTL encodes things?
A
The MTL has high-level, “concept”-like visual recognition. It has been associated with recognition of faces, objects, scenes, as well as various aspects of memory.
It shows us that the MTL may transform complex visual percepts into long-term and abstract concept-like representations.
21
Q
What is numerical cognition? Where does numerical cognition take place? Why is it inherently challenging to study?
A
Numerical cognition is non-symbolic. It’s challenging because it is inherently confounded with sensory and spatial information - inherently intertwined, which means something for the brain’s way of doing this/processing information. (same size stimuli results in different cumulative area…same cumulative area results in different overall area, same overall area results in different density).
It takes place in the prefrontal and parietal cortices - specifically, intraparietal sulcus.
22
Q
Studying the neural substrate of numerical cognition has been partially impeded by the presence of multiple confounds inherent to using numerical stimuli. Could you name one example of such confounds? Describe one stimulus design that could be used to eliminate at least one of these confounds.
A
Imagine using number of dots to indicate numerosity - one stimulus set with three dots, and another stimulus set with five dots. If you are using the same-sized dots, then the two stimulus sets would have different total surface areas of the dots, which can lead to confounding results when investigating neural representations in brain regions that are sensitive to visual input. To solve this issue, one can match the total surface area by making the dots in the stimulus set with five dots to be smaller.
23
Q
Is numerical cognition sense-modality specific? (would it care if it’s auditory or visual?)
A
Sensory modality doesn’t matter! As numerosity is non-sensory dependent.
24
What is a delayed match-to-sample experiment?
a procedure in which the participant is shown a sample stimulus and then, after a variable time, a pair of test stimuli and is asked to select the test stimulus that matches the earlier sample stimulus. Correct selection of the matching stimulus is reinforced.
25
What is the attention vs. intention debate? What area was studied in this debate?
The question is whether motor-processing systems are distinct from circuits for controlling covert and overt attention. Do motor-planning (action) and attention processes overlap (or are the same) in the brain? This would redefine attention as motor planning. Attention = sheer motor processing. Intention = action-oriented (as affected by movement goal).
The posterior parietal cortex (implicated in spatial attention and eye movements) was studied to determine whether the signals in the posterior parietal cortex reflected attention or intention.
26
How would you test if LIP (lateral intraparietal area) neurons really track attention (vs. movement goal)?
Design a saccade-to-goal experiment. A target cue always tells where to go on Go trials. But, there is a task-irrelevant distractor (which only appears on 50% of trials), BUT captures the attention at a different location than the potential motor goal. Why use the distractor? It captures attention at a different location than the potential motor goal (testing motorindependent attention) (i.e., distractor captures attention when it's not a target -> so the distractor-driven activity argues for nonmotor-related attention signal)
The result: attention is involuntarily drawn to a distractor early when planning a saccade elsewhere, but the attentional effect lasts a short time. LIP responses reflect that the attention is captured, although it shows greater response to target WHEN there is an attentional benefit. Activity of LIP neurons show similar response to target and distractor when there is no attentional benefit. So, LIP activity is nicely tracking where attention is captured with respect to behavioral aspects of attention-related task performance. LIP activity is tracking where the attention is beneficial for the task.
27
How would you design a study to test the difference between "space" and "motor effect," in attention/intention debate?
Two tasks. First, a cue-delay-target task. In this task, the monkey knows from the beginning whether it'll saccade or reach with its arm (motor effect). But, the information for where the intended motion will be cued is delayed (spatial target comes up afterwards)
Second, a target-delay-cue task. In this task, the monkey knows from the beginning where its spatial target will be. It finds out how it'll gesture (saccade or reach) after a delay period.
By dividing it into two tasks with a delay period, you can measure the monkey's brain activity, to understand how the monkeys response to intention (effector instruction) or spatial information (attention). Even without the spatial information, these cells respond to effector instruction. However, they still are spatially tuned. So, they're responding to BOTH attention and intention.
28
How do you use visual awareness to study consciousness? How does the binocular rivalry paradigm relate to this?
Conscious perception, by testing how the same visual stimuli are "gated by" consciousness or "reach" consciousness.
The binocular rivalry paradigm - to study what's reaching the brain/consciousness stream. Red/Green filter glasses (only house on Left eye and only face on Right eye), leads to alterations in subjective perceptions (don't perceive both). However, the actual brain activity matches the non-rivalry condition.
This indicates the importance of "perception" rather than "just seeing."
29
What is the premotor cortex? How would you test if the PM is the system that integrates information (e.g. - which arm for which target?)
PMd (dorsal) and PMv (ventral) represent movement direction at a higher level than the primary motor cortex. Neurons are tuned to a preferred direction (PD - when moving an arm to a specific location in space), movement field (activity only evoked with motion).
Setup: cue for which arm (left or right), then cue for direction of target (which target, left or right), then GO. The key here - sequential information allows to tease apart the processing.
Result: They have effector instruction selective cells, and target instructions elective cells (pure right arm instruction selective (before knowing which target), and pure right target instruction selective, (before knowing which arm)). And, there is action selective neuron (moving right arm to left target - integration order doesn't matter as long as that's the final action).
30
When measuring the effects of the premotor cortex in selecting motor decisions (attention vs. intention): What would you expect to see if a task has two possible targets, with a memory moment, and then the correct target later revealed? (versus the one target task, the target known the entire time)?
I would expect to see the build-up cell show activity for both directions, but activity for the undesired direction shuts down once target is given. The selected-response cell will have no activty until target is known, where activity shoots up for desired direction.
The immediate drop in cell activity (when revealed that that stored direction is no longer the target) shows that premotor cortex "selects" motor decisions.
31
What is mirroring in the premotor cortex? How would you design an experiment to record these neurons?
Mirror neurons are those neurons which light up for both the direct action, and seeing the action (monkey see, monkey do).
Method: Three tasks. First, experimenter and monkey do the same action (monkey seeing and monkey doing). Second, the experimenter and monkey do different actions (monkey doing, but not seeing). Finally, as a control, have the monkey grab the food in the darkness (ruling out simple visual feedback as driving these neurons).
The result - in all three cases, the same neuron lights up for the correct action (either perceived or done themselves). The neuron does not light up for the seeing the experimenter do a different action.
Significance: confirming that these mirror neurons in the premotor cortex really are responding to that action, independent of whether it's perceived or done by the animal itself.
32
How would you design an experiment to see whether PMv mirror neurons take into account peri- vs. extra-personal space? What is the significance of this question?
What would you expect to happen with this same experiment, if an experimenter put up a panel in the peripersonal space - so that the object touched was now outside the workspace, but still inside peripersonal space?
The purpose of this is to test if the mirror neuron cares about intention; if it’s further away, you know that you cannot reach it and vice versa.
Experiment design: a monkey sitting on a table. An object is either placed inside or outside their peripersonal space. Three trials - an object inside the peripersonal space is grabbed by an experimenter, or it is grabbed by the monkey. Or, an object in the extra-personal space is grabbed by the experimenter.
Result: some neurons were turned to extrapersonal space (higher when farther) and some were turned to peripersonal space (higher when closer). Evidence that some mirror neurons change how they fire based on whether the monkey might interact with what’s happening.
33
How do human mirror neurons relate to action / intention?
Method: Manipulated video clips - picking up cup (context: before tea), versus picking up cup (context: after tea). The intention is inferred differently - before tea, the action is linked with the intention of drinking. With after tea, the action is linked with the intention of cleaning up.
Result: PMv related areas are active for intention (look at the overall activation patterns, subtract intention from action), processing intention and goal.
34
What are hypotheses about the function of mirror neurons?
Motor rehearsal: even if you're not doing the action, you're still practicing the circuite loop.
Simulating others? (figuring out the intention of others?)
35
What are the differences you'd expect between mirror neuron system in autism, and in control patients?
The activity for movement observation and movement execution in controls and individuals with autism are the same.
However, there are more variabilities in signals noted for autism as compared to controls - there's more error, more noise, variable signals.
36
What is the cingulate motor area (CMA)? If you wanted to study internal motivation signal in CMA, how would you design an experiment, and what sort of activity response would you expect to get?
Cingulate cortex has many functional roles for integrating information about sensory, affective, and movement. Important for processing behaviorally relevant movement.
Train two different arm movements (push or turn a handle) in response to visual signal. Animals voluntarily selected which movement to perform based on reward size (juice drop). If a reward is reduced, they change to a different movement, whereas maintain the same movement if the reward is constant.
Result: CMA process rewards for internally-generated action. (they don't respond when constant motion, and when they didn't switch the motion). Also, instructed change does not engage the rCMA in the same way). The implication here is that it's a different process for internally versus externally cued behavioral response.
37
What is working memory?
The ability to temporarily hold and manipulate information for daily cognitive tasks (for a few seconds, can only hold five/seven items at a time).
It depends on control of attention and mental effort, and is tied to cognitive/executive control
38
Question: how do you test to see if LIP (lateral intraparietal) neurons have a role in spatial working memory?
Task: visually guided movement versus memory guided movement. Record a LIP (lateral intraparietal) neuron, which has a preferred direction of motion. Result: shows that the LIP neuron is holding a spatial location for the upcoming eye movement - high activity up until saccade (holds activity without stimulus).
39
How would you compare the LIP and dlPFC's causal influence on short term memory? What would you expect to see as a result?
Method: Use cooling to determine causal relationship. Record neuronal activity from LIP while cooling the prefrontal cortex, AND record from dlPFC while cooling the LIP.
Measure the saccade error.
Result: saccade error increased more strongly following PFC cooling, which suggests that spatial working memory required PFC
40
How would you test to see if PFC memory is modality dependent? (would the same neuron react differently to different tones, for example?) What sort of result would you expect?
Task: match to sample - press button to indicate if the second sound frequency is higher or lower than the original.
Parametric modulation (less activity for 10hz, gradated up to 34hz) - monotonically increasing or decreasing activity.
This shows that the neuron scales activity for the type of information found in the brain. There is a linear relationship between the stimulus frequency and the activity. It is carrying specific information - easier to read out for downstream neurons
41
Let's say that you found a brain region that is activated for holding sensory information (tactile stimulation frequency) in working memory. Being a good scientist, you decided to test how neural activity depends on the sensory information that is being held in working memory (high, medium, low frequency). Describe what activity patterns you might observe during which period in a single neuron that is holding the three frequency information in working memory and draw a simple PSTH plot illustrating this pattern of activity.
Because this neuron is holding information in working memory, I expect that during the delay period with no sensory information, this neuron exhibit frequency-specific modulations in firing rates. In particular, I would expect this neuron to show monotonically increasing or decreasing activity as a function of high, medium, and low frequency stimulation.
[Plus a drawing of PSTH (activity over time, with time zero being the time of the time of tactile frequency delivery) showing that there are three traces with different activity levels that persist during the delay period when there is no sensory information]
42
what is an n-back memory task?
look for target letter.
0-Back condition: look for X.
1-Back condition: any letter identical to immediately preceding one
2-Back condition: any letter identical to 2 trials back.
43
At the whole brain level, which parts are responsible for maintenance of information? How would you test this?
Task: n-back memory task
Scan the whole brain with a fMRI -
Result: parametric modulation (increased activity with harder task) in dlPFC and Broca's area - verbal rehearsal provoked
44
What is the role of D1 and D2 receptors in spatial working memory? How would you test this? What results would you expect to see?
Focal infusion sites of D1 antagonist
Method: task w/ a spatial working memory (fixation, cue, delay, saccade), and a control with the same task, but no delay (vision-guided).
Result: Blocking dopamine receptors abolishes working memory. These effects were measured by the end point discrepancy, and the saccade onset time
45
What does it mean to sharpen the tuning curve of a neuron?
Sharpening of the turning curve (higher signal-to-noise ratio). The neuron passes along more unambiguous information - it responds more sharply to the preferred direction, and less with non-preferred direction (less noise).
46
What is the relationship between schizophrenia and working memory in the PFC? How would you measure this, and what sorts of results would you expect?
Working memory in PFC is a known deficit in schizophrenia. This is more generally related to cognitive disorganization. With schizophrenia, there is already much higher activation with small working memory loads (curve is shifted tot he left).
Method: n-back working memory task.
I would expect to see hypo-activity in the dlPFC, and hyper-activity in the anterior cingulate cortex (perhaps linked to efforts).
47
How would you test perceptual discrimination in the primary somatosensory cortex? The real question here: does the somatosensory cortex track behavioral performance? What would you expect to find?
Task: frequency judgement task, compare two stimulations with different vibration frequency, and indicate which stimulus was higher or lower in frequency.
Result: linear (monotonic) relationship between firing rate and stimulus frequency - this means that the information is preserved - the single neuron in S1 (somatosensory area 1) does a very good job matching behavioral choice.
Heterogeneous (different) response types within PFC neurons concerning working memory. Just showing that different neurons respond to different frequencies, information separated by frequencies → may be a basis of a working memory signal, not a motor plan (because information is not separated by if neuron thinks its high or low, it is separated by a lot of different and all possible frequencies – the complexity of frequencies indicate we are storing information and is using working memory)
48
What is top-down vs. bottom-up attention?
,odality specificity and direction specificity. Top-down attention is voluntarily directing attention (looking for red hat in crowd). Bottom-up attention is having it captured without any intention (flash on screen).
49
How would you measure the basic mechanism of visual attention in V4? (what, exactly, does attention do?) What would you expect to see?
Task: maintain fixation at the center, release a bar if the Sample stimulus matches the Test stimulus (match to sample). On either side of fixation point, there's a colored circle and a striped patch. The monkey already knows whether it's supposed to match the patch orientation or the circle orientation.
Method: this is a UNIQUE opportunity to compare the same stimulus (inside receptive field) when attended or not attended. (it's the same sensory stimulation inside the RF, but sometimes unattended, and sometimes attended, in a single orientation-selective V4 neuron.
Result: a gain change (multiplier effect - for orientation selectivity, when it's preferred direction of motion, more increased than w/ non-preferred direction of motion) in visual tuning function (volume knob). When attending, a neuron fires much more - responses are strongly gated by attention.
Attention creates multiplication effect, gain change on the tuning curve (change relative to control)
- Testing different hypothesis to see if this is most consistent explanation:
o Not additive
o OR, tuning width (signal to noise is getting sharper)
o OR, testing for orientation selectivity (shifting the curve left or right)
50
How would you test if motion direction selective responses are modulated by attention? What would you expect to see?
Measure activity in the MT (middle temporal) neurons (which are motion sensitive).
Method: maintain fixation at center, one dot appears first (instructing the animal to attend), and an animal presses the lever. Then, another dot appears, and both dots start moving. One moves inside the RF, other moves outside RF. The task is to release lever when ATTENDED dot changes SPEED of motion. The same stimulus is inside the RF, but only sometimes attended. The neuron received identical sensory information, the isolated difference is whether the animal was attending or not.
Result: the activity is reduced when not attending in a single MT neuron. The neuron's activity is gated by attention
51
Testing the basic mechanism of visual attention in the MT/MST. How would you test whether there's attention selectivity within the RF, too? Results?
For example, take a MT neuron with an upward motion preference. Both up/down motion stimulation is going on inside RF, although attending only one.
Result: response is strongly gated by attention! The neuron's action would be highest with upward motion inside the RF, and at baseline with the downward (nonpreferred motion) inside the RF. But, when paying attention to motion outside the RF, even with identical upward motion happening inside the RF, the neuron is at baseline activity. This shows that activity in the MT is strongly gated by attention.
52
What about the data shows us that attention is a gain shift (multiplier effect)?
It is an increase in amplitude (gain) - essentially a multiplication. There is no change in tuning width, or orientation selectivity (which would be a horizontal shift) - it is truly a multiplicative function.
53
What is the salience map theory of the lateral intraparietal sulcus (LIP)? How would you test it? Results?
Question: what aspects of visual stimulus engage LIP most effectively? Hypothesis: Behaviorally salient stimulus is the most effective stimulus (i.e., LIP is saliency map in the brain).
Method: two tasks. First task, just move eyes from one position to another (a fixation point with many symbols, task is to move fixation point upward). Task-irrelevant stimulus in the RF (what stimulates the RF has no behavioral meaning). Versus, a second task, pay attention to a cue (outside receptive field), move eye from one position to another, then saccade to match the symbol with the previous cue. This then becomes a task-relevant stimulus in the RF. (exact same visual information in RF as the first task, only now, it's attention relevant).
Result: LIP neuron only responds to salient stimulus. Compare 1) a baseline graph, when a stimulus flashes in a receptive field (captures attention); 2) a task-irrelevant stimulus in RF (not salient); 3) task-relevant stimulus in RF (salient). The graphs for attention capturing and salience show a spike at the stimulus onset, for the task irrelevant, it has no spike. It has to be behaviorally-relevant (saliency argument) - there is no stimulus drive.
54
How would you test to see if you can bias attention in the Frontal eye field (FEF)? Why would you choose FEF? What results would you expect?
FEF is known for saccade generation - microstimulation evokes saccades to a specific location.
Method: microstimulate FEF at subthreshold level (lower than evoking a saccade) to see if you can BIAS spatial attention. The motor field (MF) - where the saccade would go to, if it were over threshold.
Task: monkey fixates, then a target enters motor field (after microstimulation). Distractors come on/off constantly on the rest of the screen. When the target dims, monkey releases a lever.
Result: sub-threshold stimulation improves attentional sensitivity (lower luminance change required). This shows how closely related attention and eye movement are.
55
How would you test the difference between voluntarily (endogenous - top-down, internally driven) and involuntarily controlled attention (exogenous, stim-driven, bottom-up), from a network perspective? What results would you expect?
Method: cued spatial attention - told which cue (left or right arrow) to pay attention).
The Result (from network perspective, whole brain scanning): very different regions for topdown versus bottom up attention. In the case where subjects voluntarily directed attention (based on a cue), superior frontal, inferior parietal, and superior temporal all lit up. Versus, when it's stimulus driven, very dif regions light up (parietal).
56
What is spatial hemineglect?
Half of visual field is neglected. It's the defective ability of patients with unilateral brain damage to explore the side of space contralateral to the lesion (contralesional), and to report stimuli presented in that portion of space. Left hemineglect is more common, because human visual processing is lateralized: right side is processed by both L/R hemisphere, but left side is more exclusively processed by R hemisphere.
It's interesting to see the lateralization in human brains (which only starts with apes, not found in mice). This lateralization is for increased brain capacity, bc lateralization means that neural circuits do not have to be duplicated in each hemisphere.
57
What is functional connectivity?
Intrinsically coupled brain networks - inter-regional coupling at the global brain level. During typical attention tasks, some regions enhance BOLD activations ("task positive" regions), whereas other regions decrease activations ("task-negative" regions). At rest, there are a group of regions that are correlated with one another, and another regions that are anti-correlated. They do not have to be anatomically connected regions - they can be connected through fiber tracts. Anatomically different brain regions share tasks and functions. They could also just be responding separately to common stimuli.
58
How would you test which direction the information flows in the frontoparietal attention network, for top-down vs. bottom-up attention? What result would you expect?
Method: two tasks - a visual pop-out (engaging bottom-up attention), and visual search (engaging top-down attention). Confirm that the task works, because the reaction time is longer for top-down. Use neurons' target location selectivity to index their roles in attentional control.
Measure specific cells in the LIP (parietal) and prefrontal (FEF and lPFC) regions, to see which "come online" first / reach significance first, e.g., which regions first signal target location.
Result: for top-down (search), the prefrontal (lPFC and FEF) discriminated target location FIRST. For bottom-up (pop-out), parietal neurons (LIP) signaled target location first. (oriented around time from saccade, looking @ spikes before saccade)
59
How would you test the neuronal coupling across the frontoparietal network for control of attention? What result would you expect? (HINT - look @ coherence)
Measure COHERENCE between local field potential (LFP) from a parietal and prefrontal neuron.
LFP coherence: each neuron oscillates periodically, to look at the communication between groups, look at how the oscillations are "coherent" - how systematically related in phase they are. (this is looking at the gradient, not the action potential, which would be merely yes/no).
(spikes are high frequency, >700 HZ, LFP is low frequency, <250 hz - more related to fMRI signal).
Takeaway: when we are looking at top-down(search task) versus bottom-up(pop up task), there are differential patterns of the LFP coupling. Middle frequency is more related to search task. High frequency is more related to pop-out task.
60
For neuronal coupling/coherence, what would you expect to see for whether the communication differs for the top-down/bottom-up attention in the frontoparietal attention network?
Pop-out uses high frequency band (bottom-up, makes sense, because it's faster). Versus search (top-down), which uses low frequency band (slower).
Result: different frequency band in LFP signals carry the top-down vs. bottom-up attention signals across the frontoparietal attention network.
61
How would you test the interaction between attention and reward in LIP neuron? Result?
Attention and reward are confounded.
Method: record neuronal activity in LIP while varying the juice amounts earned for the animals (probability that the reward is delivered).
Task: two targets come up (one inside RF, other outside RF), then either central fixation point or one target goes off to signal the monkey to make a saccade.
Result: when low-probability reward, less activity. When high-probability, MORE activity. This makes sense, because LIP neurons measure salience.
62
How would you test the causal relationship between dopamine processing in the FEF and the top-down attention?
Method: focal injection of a D1R antagonist, in the FEF site that matches the RF location with the V4 neuron.
Task: monkey fixates, two targets appear asynchronously (one of them in the FEF RF, the asynchronous biases for selecting the first target), monkey saccades. Record V4 neuron, look @ behavior.
Result: dopamine processing in FEF improves attention. It establishes that the dopamine system IS attached with attention. Dopamine amplifies attention: The V4 tuning is sharper - gain change. Biases attention toward the V4 RF target with free choice.
63
For what purpose would you want to test sea slug gill reflexes? Why would you choose a sea slug?
PURPOSE: studying memory at the neural level
STRUCTURE OF STUDY: When touched, the gill shows withdrawal reflex. After conditioning (direct tactile stimulus + tail shock), a memory is formed. They figure out that you can condition withdrawal by pairing tactile stimuli with a tail shock. The shocked group means the stimulus takes on (and maintains) a very different meaning in the slug's brain.
You would choose to study a sea slug because it has a very very simple system: tactile stimulus -> sensory neuron -> motor neuron -> gill retracted. With sensitization, the stimulus also feeds into the motor neuron. It's a question of what happens at this synapse, in terms of EPSP and IPSP.
RESULT: This is memory at the synaptic level. The pairing of the tactile touch to siphon with the shock to tail results in a facilitation of EPSP in the synapse between the sensory neuron and motor neuron, which lowers the threshold for action potential, gives a lower signal a higher ability to activate the neuron.
64
What is EPSP and IPSP, and what role do they play in memory at a synaptic level?
Excitatory Postsynaptic Potential (EPSP) is depolarization in the postsynaptic neuron, makes it easier for neuron to be activated. Inhibitory Postsynaptic Potential (IPSP) is hyperpolarization in the postsynaptic neuron, makes it more difficult for the neuron to be activated.
These are important because they are ground 0 for synaptic plasticity. THESE are where memory based learnings are grounded - they show memory at the synaptic level. The synaptic potential conditions the threshold for action potential - increases or decreases it.
65
If you wanted to study the causal role of different sections of the medial temporal memory system, how would you do it? What results (and where) would you get?
Medial temporal memory system.
Method: use a delay non-match-to-sample (DNMS), testing with monkeys with aspirated lesions in the medial temporal memory system.
1. If monkey moves one object (key), obtains food underneath.
2. After variable delay (hours to days), 2 objects come up.
3. Food is under the NEW object - must use memory to recognize the first object as familiar, to successfully get the food under second object.
Result: the more the lesions included the brain areas in the medial temporal lobe memory system, the worse the behavior in the DNMS task became. However, lesions to the rhinal cortex produced much more severe drops in performance than lesions to the hippocampus. The rhinal cortex is the key node in the medial temporal memory system, not the hippocampus (as was previously hypothesized).
66
Does an exitotoxic lesion to the hippocampus / amygdala lead to deficits in the DNMS task, in the same way that an aspirated lesion did? What do these results suggest about previous studies with lesions?
Result: selective lesions to the hippocampus and amygdala result in intact visual recognition memory. However, the rhinal cortex lesion DID remain damaging to memory. This suggests that previous results using the aspiration technique might have been driven by damaging fibers of passage (rather than that the hippocampus and amygdala themselves are critical).
67
What experiment would you use in rodents, in order to determine the causal responsibility of the hippocampus in spatial memory? What result would you expect?
Morris water maze: tests how animals use external or internal spatial cues to find a platform (how long does it take?). Big tub, hidden platform under surface of milky white water, with stable visual cues around. Mouse must swim until it finds the platform. Repeated, measuring the directness of the path/time it takes to find platform after learning.
Result: hippocampal lesion impairs spatial navigation (route to the platform roundabout, vs. cortical lesion/control - relatively straight)
68
What is long-term potentiation (LTP)? Long-term depression (LTD)?
LTP - an increase in a cell's firing potential after brief, rapid stimulation (delivering repetitive electrical stimulation, find that certain stimuli pattern create long term changes in synaptic potential, which changes threshold for action potential.
LTD - long-term decrease in the excitability of a neuron to a particular synaptic input.
A persistent weakening or strengthening of synapses based on recent patterns of activity.
Significance: Believed to be a neural basis for learning and memory. If you used an antagonist to block the NMDA receptors in mice, you'd see a lack of LTP, which then results in spatial memory deficits.
69
Place cells in hippocampus, grid cells in entorhinal cortex, how do they interact?
Place cells in hippocampus, active at distinct places, encoding locations in space. Like receptive fields in visual neurons, they have 'place fields.'
Grid cells encode space-like, equal distanced grid-like pattern in hexagonal lattice, in the entorhinal cortex. They are thought to provide inputs to the hippocampus to form 'cognitive map' (place cells). A single grid cell fires at equal distanced intervals, and if there's a higher connection between grid cells, a place cell is formed
70
What is the difficulty with the term "executive/cognitive control?" How is it often used?
It's non-precise, since it's hard to distinguish from top-down processing. It's often used in tandem with task related processing - selecting, monitoring, and regulating for a goal-based behavior
71
How would you measure how the ACC (anterior cingulate cortex) activation reflects conflict in executive control, and how this conflict is used to adjust behavior (hint: across trials, look at task switch effect)?
Use a Stroop task.
To get task switch effect (post-conflict adjustment effect), compare the interaction with the previous trial type (current trial type = incongruent or congruent, previous trial type = congruent or incongruent). The currently congruent trials are faster overall (and it doesn't matter what the previous trial was, measured in terms of reaction time). However, with currently incongruent trials, there is this task switch effect - C->I is slower than I->I (going from Congruent to Incongruent is particularly hard - takes longest reaction time). Try to draw this out! This is also measured with BOLD activity (not just reaction time) - %signal change on Y axis, time on X axis, c->I with higher curve than i->I (another way to measure task switch effect cost).
72
What is a Stroop task, and what is it used to test? What are the basic findings?
Conflict-monitoring related activity in executive control. Measuring RESPONSE TIME - for Red, Green (colored words, word represents a different color to text color, e.g. "Red" would be colored green). Obviously, response time is faster for congruent than incongruent trials (congruent "Red" is colored red).
73
For conflict monitoring in the ACC, with a Stroop test, what would it mean to split the participants into two trial types, high adjustment trials, and low adjustment trials? What does this measure?
High adjustment trials were ones in whcih people showed faster RT on i-I trials than the median. They adjusted well to the incongruent trials (adjusted behavior well). The low adjustment trials were those in which people showed slower RT on i-I trials than the median. They didn't adjust as well to the incongruent trials (adjusted behavior poorly). These differences in brain activity show how well one implement the control (adjust to the previous test).
74
How would you compare the role of the ACC and the PFC in executive control (as related to conflict monitoring and adjusting?)
ACC - involved in detecting conflict (brain lights up for incongruent trials and error trials). PFC - involved in implementing the adjustment. It shows greater activity for high-adjustment (reacting and responding well to previous condition)than for low-adjustment trials (where they don't implement any adjustment well).
75
How would you test the role of SEF (supplementary eye field - medial frontal cortex) and performance monitoring? What test would you use, and what would you expect to see?
SEF is medial frontal structure involved in high-level processing of eye movements (gaze gets triggered by microstimulation)
Method: saccade countermanding task, which is a basic cue/saccade task, but which sometimes includes a second screen with a stop cue (sometimes the conflict is detected, and the animal successfully countermands the impending saccade, and sometimes the error is made.
This task is great because it allows you to compare the same behavior (eye movement) - one case with error, one case correct (without stop).
Result: the neurons in SEF are divided into caring about ERROR, CONFLICT, or REINFORCEMENT/REWARD (increased activity after successful no-stop trial, and after successfully cancelled trial).
76
What is the role of the ACC (medial frontal cortex) in performance monitoring? How does it compare to SEF (supplementary eye field)
Phasic (short activation), vs. tonic (longer term) - signal error. They show higher activation for non-cancelled trials (error on the stop trials) than on the no-stop-signal (comparing two cases where the saccade was made in either event). But, UNLIKE SEF, these ACC cells do not signal conflict (no-stop-signal and cancelled trials). Possibility: ACC may monitor the consequences of the actions, while SEF is more actively determining them in the first place (perhaps SEF sends error information to the ACC)
77
If you were asked to compare the medial pre/frontal structures (SEF, ACC), versus lateral prefrontal areas, in terms of their role in cognitive control?
All are engaged during various cognitive operations that require control. Monitoring of conflict may arise in SEF, ACC (medial frontal structures) - performance monitoring and signal conflict.
Lateral prefrontal cortices (dlPFC), regulatory/implementation-related processes - encoding rules and strategies, including abstract ones.
78
What is an overview of rule learning and rule based control for executive functioning?
The brain needs to program behavioral responses based on rules, which are context dependent (not stable over time). This requires flexible mapping.
79
How would you design an experiment to test the role of PFC in rule-/strategy-based control? What results would you expect?
Method: rule-based delay match-to-sample task. (according to the rule, release the bar if the stimulus either MATCHES or DOES NOT match the sample stimulus). Different rules are communicated by a jop of juice / no juice, or a high tone/low tone.
Measuring the firing rate (Hz) of a single prefrontal neuron, shows rule selectivity - different activity for match vs. non-match rule. This is regardless of how the rule was conveyed (juice or tone), and regardless of the sample stimuli. This neuron differentiates match rule to non-match rule
80
How would you measure/design an experiment to understand how PFC neurons map action-response based on rules? Results?
Method: Stay vs. Shift task (this is the rule), with saccades. Stay cue (either by orientation of a bar or color) means choose the same target as in the previous trial, whereas Shift cue means to change to the alternative target compared to the previous trial. These cues ask monkeys to change strategy for earning juice reward
This is important because the rule is based on the internal representation of the rule (compared to previous action), versus the study with rule-based delay match-to-sample (where the rule was externally represented through juice or a tone).
Result: there are cells selective for both strategy (either for shift or stay), and response (increased activity for either leftward or rightward response) - neurons are selective for a combination of strategy type (stay/shift) AND saccade type/action (left/right). This is consistent with dlPFC's role in implementing strategy with respect to a particular response - the dlPFC deals flexibly with rule implementation, whereas other parts are more about mapping or correctness or strategy
81
What is a bivariant stimuli? What is it useful for measuring, in terms of rule-/strategy-based control in humans?
Univariant targets always provoke the same response. With bivariant targets, the response depends on the context.
Bivariant targets are more difficult when switching context (measured by percent errors - switch cost!).
Using non-stationary mapping between stimulus and response - IF house is preceded by circle, saccade to left, IF house is preceded by triangle, saccade to right; IF tree is preceded by circle, saccade to right, IF tree is preceded by triangle, saccade to left.
82
How are representing rules and changing/adapting to rules separable in the brain? How would you measure this using bivariant and univariant saccade test?
Rule representation is in lateral PFC (bivariant > univariant for repetition and switch - it signals rule representation, since it doesn't matter whether it's repetition or switch).
Dorsal-medial PFC is for rule updating representation -> adapting behavior (bivariant > univariant only on switch - is task-set reconfiguration).
Result: representing rules (lateral PFC) and changing/adapting to rules (switch, medial frontal cortex) are separable in the brain.
83
What is a task block, and what is it useful for measuring?
A task block is a repeated task, with two tasks interleaved (e.g. either identify man-made vs. nature made, OR large/small). For the task, you either can repeat the same task, or switch the task.
You can analyze differences within the block.
A mixed block would have both rules present, a single-task block would only have one. This allows you to measure:
Mixing cost effect - compare task repeat in a mixed block with a single task block.
Switch cost effect - within a mixed block, compare task-repeat or task-switch. It allows you to divide up brain regions for sustained task control (which experience mixed cost effect, anterior PFC), versus transient task control (which experience switch cost effect - lateral superior parietal cortex)
84
How would you test the global effects of emotional control in the brain, in terms of arousal? What results would you expect to find?
A study scanned people while they viewed erotic films, asked them to inhibit their arousal to examine how the brain mediates self-regulation.
Result: regions activated for sexual arousal can get "shut down" with emotional control. Emotional control / attempted inhibition activates a different set of regions.
85
Describe a study which examines increase or decreasing negative emotions. How would you design this study, and what would you expect it to reveal?
Method: a participant looks at a screen, receives an instructional cue to increase negativity, decrease negativity (either using a situation-focused or self-focused regulation strategy) or look at image and respond naturally. Then, a negative or neutral photo comes up, and the participant regulates their reaction. Then, they rate the strength of their affect (how subjectively successful the regulation was).
Result: behavioral assessments show that regulation works: they could increase negative emotion above baseline, as well as (with more effort) decrease negative emotions.
The PFC and ACC are recruited for both processes. (the lateral PFC is more externally focused, and the internally focused is medial PFC)
86
How would you create a study to test the regulation of fear in the brain? Why is this type of regulation preferable to the subjective rating of negative emotion?
It's preferable to study fear because it's not subjectively measured - it's measured by skin conductance response (physiological response).
A participant receives a signal (yellow cube = no shock, blue cube = shock).
Then, they're either told to attend (focus on natural feelings), or regulate (think of something calming). The group who down-regulated their negative emotions lowered their SCR!
Emotion regulation activates similar parts of the brain - similar patterns of activation - as extinction learning (where you rewrite the cue mapping, learning that the blue cube longer gives a shock). THERAPUTIC potential (cross a bridge 100x, unlearn the association)
87
How would you test the effect of reward - reward modulation (and flexibility) - in the basal ganglia? (hint - these neurons are not direction specific! the result will be that they re-code based on the rewarded direction...)
Method: a memory-guided saccade task with 1-direction rewarded (out of four directions). Learning is required for the monkeys to figure out which is the rewarded direction. They record information in the caudate nucleus.
Result: expectation of reward modluates the basal ganglia neurons - evidence for how reinforcement taps into the basal ganglia control circuit. The same cell dramatically shifts tuning to recode the rewarded target in the 1DR task. (amazing bc activity is based on reward, not direction). This would be a facilitation type neuron. A suppression type neuron encodes the suppression of visual response. Both of these neurons reveal a strong gating based on reward - the activity goes up or to ZERO, flexibly modulated
88
How does the basal ganglia relate to cognitive control?
The basal ganglia is a critical structure for adaptive behavior based on reinforcement/reward.
