ABSOLUTE TAKEAWAYS cram sheet Flashcards
1
Q
why are neuromodulators well positioned to modulate internal states?
A
- neuromodulators modulate SYNAPTIC and CELLULAR function over LONG TIME SCALES
^ because they influence BIOCHEMICAL SIGNALLING and ION CHANNEL FUNCTION
- neuromodulator effects can SCALE with the magnitude of neuromodulator released
- can act LOCALLY or send signals across MULTIPLE BRAIN REGIONS
^ these properties make them well-suited to FLEXIBLE, SCALABLE and PERSISTENT control of behaviour that’s central to internal states
2
Q
4 themes that have been gathered from examining neuromodulators as behind internal states
A
- internal states influence multiple circuits and cell types in parallel
- neuromodulators act in concert
- state transitions engage mutually-exclusive neural populations
- persistence through recurrent dynamics
3
Q
2 mechanisms for persistent changes in neural activity
A
- NEUROMODULATORY SIGNALLING (long time scales, fan-out effects)
- RECURRENT NEURAL CIRCUITS (excitatory neuronal circuits)
4
Q
what does extinction training do (hint: suppresses what? activated what?)
A
EXTINCTION TRAINING SUPPRESSES REACTIVATION of contextual fear engram cells
while ACTIVATING A SECOND ENSEMBLE (putative extinction engram)
5
Q
what governs the suppression and relapse of fear after extinction?
A
HIPPOCAMPUS generates FEAR EXTINCTION REP
INTERACTIONS between hippocampal fear and extinction representations GOVERN the SUPPRESSION and RELAPSE of FEAR after EXTINCTION
6
Q
extinction training suppresses reactivation of what?
A
contextual fear engram cells
7
Q
extinction training activates what?
A
a putative extinction engram
8
Q
what does optogenetic inhibition of neurons active during extinction training do?
A
increases fear (more freezing)
9
Q
optogenetic silencing of neurons active during fear training does what?
A
leads to reduced spontaneous recovery of fear
10
Q
3 phenomena that show that fear memory can RETURN after extinction
A
- reinstatement
- renewal
- spontaneous recovery
11
Q
engram
A
physical or biochemical changes in the brain
formed when an experience occurs and persists over time
ALLOWS MEMORIES TO BE LATER RECALLED
12
Q
what does extinction to do the number of fear acquisition tagged cells?
A
extinction REDUCES the number of fear aqcuisition tagged cells that are active
compared to mice who didn’t undergo extinction
means that EXTINCTION REDUCES ACTIVITY OF FEAR ENGRAMS
13
Q
“no extinction” group will have a higher proportion of what?
A
reactivated cells
because extinction reduces the activity of fear engrams
14
Q
what does observing the reactivation of fear (ie. conditioning tagged) and extinction engrams suggest?
A
suggests that during extinction…
the FEAR ENGRAM IS SUPPRESSED
in favour of the EXTINCTION ENGRAM
15
Q
how do they silence extinction engram cells?
A
tag cells that were active during EXTINCTION with HALORHODOPSIN
halorhodopsin: inhibitory opsin stimulated by green light
16
Q
what happens when silencing extinction memories during spontaneous recovery?
A
no effect
because during spontaneous recovery the fear engrams are more active
16
Q
findings suggest that neuronal ensembles in the ________ _______ ______ play what kind of role in the expression of _____ and ______
A
HIPPOCAMPAL DENTATE GYRUS
play a CAUSAL ROLE
expression of FEAR and EXTINCTION
17
Q
LeDoux’s 2 system model
A
(contrasts with the ‘FEAR CENTER’ MODEL)
threat >
sensory system >
COGNITIVE CIRCUIT and DEFENSIVE SURVIVAL CIRCUIT
cognitive circuit leads to FEAR (fearful feelings)
survival circuit leads to DEFENSIVE RESPONSES
18
Q
where do fear memories reside?
A
unlikely to reside in a single brain region
engrams are likely DISTRIBUTED ACROSS BRAIN
19
Q
what did Lashley conclude about memories after his lesion studies?
A
memories (engrams) aren’t localized in specific parts of the brain
they’re highly distributed
20
Q
paper predicts what about the magnitude of increased network activity?
A
magnitude of increased network activity during extinction learning
will be CORRELATED
with EXTINCTION RECALL in a LATER SESSION after a delay
21
Q
measure of learning in the lamp study
A
skin conductance response (SCR) as measured by electrodes
22
Q
lamp study - what does a region with high connectivity on a trial indicate?
A
indicates the activity of this region is STRONGLY CORRELATED to the REST OF THE BRAIN compared to other trials
23
Q
lamp study - main broad takeaway result
A
brain connectivity changes during fear conditioning, extinction and extinction recall
(changes in brain connectivity induced by extinction learning are different for the CS+ versus the CS-)
(because extinction is happening to the CS+, but nothing is happening to the CS-)
24
lamp study - within the CS+ type, what did they observe about connectivity?
significant change (INCREASE) in connectivity for CS+ from EARLY CONDITIONING to EXTINCTION
(NO SUCH DIFFERENCE WAS OBSERVED FOR CS-)
25
lamp study - 133 brain regions generally exhibited larger connections with other regions in late extinction learning during what processing?
during CS+ processing
(as opposed to during CS- processing - because in CS+ processing this connectivity reps the extinction training, and there was nothing to go extinct in the CS- so therefore no accompanying increase in connectivity)
later analyses looked at average activity of the 133 regions and IDENTIFIED THEM AS IMPORTANT FOR LATE EXTINCTION
26
lamp study - was there a significant change in connectivity for CS+ during fear conditioning?
no
but in LATE CONDITIONING, CS+ connectivity was LOWER than CS- connectivity
this is the OPPOSITE to the observed connectivity INCREASE for CS+ across extinction
27
lamp study - what do they conclude about the brain connectivity changes induced by extinction learning?
conclude that these changes induced by extinction learning TARGET BRAIN NETWORKS INVOLVED IN:
emotion regulation
memory storage
conscious attention processing
(fronto parietal network, default mode network, ventral attention network)
28
lamp study - changes in activity seem to occur when?
seems like brain activity modulation MAINLY OCCURRED during EARLY EXTINCTION and EXTINCTION RECALL
whereas connectivity changes were more prominent in LATE EXTINCTION
29
lamp study - what could the changes in either connectivity or activity be related to?
could be related to CONSOLIDATION OF THE EXTINCTION MEMORY
ie. high activity during early conditioning, early extinction, then it drops off and increases again during extinction recall
ie. connectivity increases in late extinction
30
lamp study - looking at the ERI, they showed that what correlates with extinction?
CONNECTIVITY
but not activity
CORRELATES WITH EXTINCTION
(overall increase in mean connectivity across extinction-associated regions from early to late extinction was positively correlated with ERI for the CS+E)
31
lamp study - how may connectivity and activity reflect diff processes?
ACTIVITY may reflect LOCAL PROCESSING
whereas brain CONNECTIVITY may reflect NEURAL PLASTICITY and connectivity changes may be important for MEMORY CONSOLIDATION
32
PV neurons and engram size paper - summary of the effects
1. stress >
2. increased LA AEA signalling >
3. decreased GABA release >
4. increased excitatory activity >
5. increased engram size >
6. threat generalization
33
parvalbumin interneurons
GABAergic interneurons (inhibitory)
powerfully control the activity of pyramidal neurons
they're the specific cell type that's important in MAINTAINING SPARSE CODING
34
PV neurons and engram size paper - what does acute restraint stress increase?
the stress hormone corticosterone (CORT)
CORt increases during restraint stress, and foot shock etc
35
PV neurons and engram size paper - CORT has been shown to increase what?
threat memory generalization
(administering CORT results in similar increases in fear generalization to those seen in restraint stress manipulations)
36
what does CORT do to activity of LA PV+ inhibitory neurons during threat conditioning?
CORT DECREASES the activity of these cells
37
footshock does what to PV neurons?
increases their activity
but this effect is BLUNTED in CORT treated mice
38
CORT didn't change what things?
1. baseline activity without shock
2. activity in other interneurons
(so CORT is specifically blunting the PV increase after the footshock - specific neurons, specific timing)
39
how does CORT produce a less sparse engram?
1. decreasing inhibitory LA PV+ neuron activity
2. increasing excitatory pyramidal neuron activity
40
how does stress suppress PV activity?
we know that CORT suppresses PV neuron activity - but CORT is just a proxy for stress that was directly manipulated in the experiment
how does this work in real life?
ENDOCANNABINOIDS
41
what effect does eCB signalling have on PV neurons in other brain areas?
eCB signalling DISRUPTS GABA RELEASE from PV neurons in other brain regions
so it disrupts inhibitory signalling (just like CORT does)
42
what ligands bind CB1 receptors?
2 endogenous ligands
1. ANANDAMIDE (AEA)
2. 2AG
retrograde signalling to pre-synaptic CB1 receptors mediate changes in NT release
CB1 receptors = located on pre-synaptic neurons
so, 2AG and AEA bind to CB1 and affect NT release
43
what does inhibiting AEA in CORT treated mice do?
it REDUCES GENERALIZATION
inhibiting AEA restores memory specificity!
44
CORT-induced memory generalization is mediated by...
increased AEA levels in the LA
(can mimic effect of CORT by increasing AEA, and can block effect of CORT by reducing AEA)
45
CORT's effect on pyramidal activity and eCB signalling
CORT increases pyramidal activity
CORT increases eCB signalling
46
what does knocking down CB1 in LA PV+ neurons do?
1. sufficient to RESTORE MEMORY SPECIFICITY
1 a) knocking down CB1 receptors in PV prevents CORT-induced generalization in freezing
1 b) also restores the sparsity of the threat engram
2 c) also prevents stress induced generalization in freezing
47
cardiogenic control paper makes use of what technical advances?
ChRmine
and tiny mice vest that emits light
48
cardiogenic control paper - what do they conclude from the behavioural tests?
1. increasing heart rate increases anxiety in anxiogenic contexts
(so not in the room preference test, but yes in the other two)
2. stimulation alone has no effect - must be elevated HR
(the same manipulation within a baseline heart rate range (660 bpm) had no effect on behaviour)
49
cardiogenic control paper - activity of what area was found to be required for heart-rate induced-anxiety like behaviour?
posterior insular cortex
in animals where they suppress pIC activity, there’s an attenuated effect of the heightened HR
50
cardiogenic control paper - did they find that PFC activity was required for heart-rate induced anxiety-like behaviour?
no - it wasn’t
there was NO EFFECT when inhibiting PFC acitivity
51
cardiogenic control paper - is inhibition of the posterior insular cortex alone enough to alter anxiety like behaviour?
no
inhibiting the posterior insular cortex doesn’t inhibit anxiety
52
module
a neural subsystem that's PERMANENTLY DEDICATED to mediate ONLY ONE AFFECTIVE FUNCTION
or at least only one specific valence
could be:
subregion of the brain
projection pathway
neuronal population
individual neuron
53
what can hedonic affective reactions be modulated by?
by factors that alter hedonic impact
1. PHYSIOLOGICAL STATE (hunger/satiety)
2. PSYCHOLOGICAL FACTORS (learned preferences/aversions)
3. NEUROBIOLOGICAL MANIPULATIONS (neural inhibition or stimulation)
54
mode
a given neural module can exist in MULTIPLE NEUROBIOLOGICAL STATES
that give it DIFFERENT AFFECTIVE FUNCTIONS
suggests that affective function/valence mediated by a neural module need not remain permanently stable
it can CHANGE DYNAMICALLY ACROSS VARIOUS SITUATIONS
55
how can 'liking' and 'wanting' be separated?
1. behaviourally
2. in their neural mechanisms
56
what could an affective mode correspond to?
1. particular pattern of NEUROCHEMICAL stimulation
2. pattern of GENOMIC activation
3. temporal pattern of ELECTRICAL EXCITATION of the module
4. particular NEUROBIOLOGICAL STATE of the larger circuitry within which the module is embedded
57
when might affective modes shift?
with changes in:
1. INTERNAL PHYSIOLOGICAL STATES (hunger, stress)
2. EXTERNAL STIMULI/ENVIRONMENTS
58
evidence for existence of both hedonic and motivational modules in the nucleus accumbens comes from what?
dopamine signalling
it mediates WANTING (but not liking)
59
example of a wanting module in the NAc
dopamine signalling
(it's a neurobiological unit whose function is to encode wanting)
60
hedonic hotspots
(totally separate to the role of the NAc in wanting)
hedonic hotspots were discovered that control the LIKING of sweet tastes
they are SMALL REGIONS that play a PREDOMINANT ROLE in controlling LIKING
61
network of hedonic hotspots may rep WHAT?
a 'LIKING MODULE'
62
beyond the hedonic hotspot, the NAc appears to be organized HOW to MEDIATE WHAT?
organized along a ROSTROCAUDAL GRADIENT
to mediate APPETITIVE-DEFENSIVE reactions (motivational aspect)
(so NAc = implicated in both liking and wanting)
63
2 projection pathways from the NAc
1. DIRECT: to the VTA
2. INDIRECT: to the ventral pallidum (VP)
64
what does increased activity in the direct and indirect pathways do to VTA activity?
direct: decreases VTA activity
indirect: increases VTA activity
65
2 ways to inhibit NAc projection neurons
1. disrupting glutamatergic projections from PFC, hippocampus, amygdala
2. introducing GABA agonists
66
depending on the _____ _____ _____ of a manipulation of either glutamatergic or GABAergic signaling in NAc, what happens?
depending on the ROSTRAL-CAUDAL LOCATION
of a manipulation of either glutamatergic or GABAergic signalling in the NAc
INTENSIVE APPETITIVE (eating food) of AVERSIVE (defensive burying) is elicited
67
what do rostral manipulations do?
1. increase APPETITIVE EATING
2. can induce PLACE PREFERENCE
68
what do caudal manipulations do?
1. generate intense DEFENSIVE BEHAVIOURS (burying, distress calls, attacking)
2. can induce PLACE AVERSION
69
do NAc 'motivational modules' maintain invariance across conditions?
non
the rostral-caudal gradient is modified by ENVIRONMENTAL conditions
to bias the function of INTERMEDIATE 'MODULES' towards appetitive or aversive motivational behaviours
70
in a more comfortable, low stress testing environment, appetitive behaviours can be elicited where?
at more caudal locations
(more caudal than when compared to the stable condition findings)
71
in a high-stress, bright testing environment, aversive behaviours can be elicited where?
at more rostral locations
(more rostral than when compared to the stable condition findings)
72
the fact that most of NAc is highly sensitive to environmental manipulations, to extent that valence of responses can be reversed, demonstrates what?
this demonstrates that by and large,
affective modules in the NAc are NOT dedicated to a singular valence
but they DYNAMICALLY ENCODE valence in MULTIPLE MODULES
73
what have neuroscientists done to reconcile proposal that amygdala is involved in POSITIVE and NEGATIVE valence?
proposed amygdala is divided into TWO OPPOSING MODULES
2 main ways this has been considered:
1. amygdala projection modules
2. neuronal population modules in the amygdala
74
pleasure and pain interact in what?
in our SUBJECTIVE INTERPRETATION of our CURRENT HEDONIC STATE
75
what do we know about how pleasure and pain interact in the brain?
painful and pleasant sensations are PROCESSED in DISTINCT SPINAL and PERIPHERAL CIRCUITS
(but less is known about how these signals are integrated centrally in the brain in affective experience)
76
choc capsaicin study: gap this paper tries to fill
most of the evidence on pleasure and pain comes from studies that have looked at EITHER pain or pleasure,
but NOT BOTH in the SAME INDIVIDUAL
how do brain regions that are activated both by both pain and pleasure encode these experiences?
77
choc capsaicin study: their axiomatic approach
AXIOM 1:
^ brain regions encoding information for sustained pain or pleasure must SIGNIFICANTLY PREDICT RATINGS of subjective pain or pleasure
AXIOM 2:
^ of the brain regions identified in axiom 1, those encoding AFFECTIVE INTENSITY should predict both pain and pleasure ratings, IRRESPECTIVE of the POLARITY of the rating
AXIOM 3:
^ of the brain regions identified in axiom 1, those encoding AFFECTIVE VALENCE should predict the DIRECTIONAL SIGN of both PAIN and PLEASURE ratings
78
choc/capsaicin study: they developed predictive models of INTENSITY and VALENCE based on what hypothesis?
hypothesis that INTENSITY and VALENCE are 2 fundamental affective dimensions SHARED ACROSS PAIN AND PLEASURE
79
choc/capsaicin study: what does the fact that they found almost no blue mean?
even though these areas are implicated in BOTH intensity and valence, there are specific SUB-REGIONS involved in encoding valence and intensity separately
80
choc/capsaicin study: among the 7 brain regions, affective intensity and valence seem to be repped how?
by distinction sub-populations of voxels
INSULA preferentially predicts INTENSITY
VENTROMEDIAL PFC preferentially predicts VALENCE
81
ventral attention (salience) network correlated with what?
intensity model
(so intensity is associated with a specific functional connectivity network)
82
default mode network correlated with what?
valence model
(so valence is associated with a specific functional connectivity network)
83
while valence and intensity are encoded in brain regions that process both pain and pleasure, the evidence suggests that...
the SPECIFIC SUB REGIONS as well as broader FUNCTIONAL CONNECTIVITY networks are distinct
this suggests that intensity and valence are processed in DISTINCT BRAIN CIRCUITS
so it should be possible to DISSOCIATE these processes
84
sensitive period strict definition
"a sensitive period is a temporally restricted window during development when a particular part of the brain is especially responsive to environmental stimuli and can undergo significant, enduring changes during this period"
85
critical period
extreme version of a sensitive period
if brain doesn't receive the typical inputs during a critical period, it can lead to LASTING DYSFUNCTION
some functions are EXTREMELY DEPENDENT on the appropriate experience for normal development
86
what delays the onset of the critical period in the visual system?
1. DARK REARING (binocular deprivation)
2. manipulations that PREVENT the DEVELOPMENTAL INCREASE in GABAergic INHIBITION delay the onset of the critical period plasticity
87
what NT is needed for onset of critical period development?
GABA
88
perineuronal nets
PNNs
part of the extracellular matrix
they form AROUND PARVALBUMIN GABA neurons
form a STRUCTURAL BRAKE on further plasticity
89
manipulations that remove the structural barriers to plasticity or strong modulation of the excitatory-inhibitory balance can do what?
partially restore plasticity
after the critical period
90
lateral amygdala's (LA) role in pavlovian fear learning and extinction
receives info about CS and US
plasticity in LA after repeated pairings results in FORMATION OF FEAR MEMORY
CS presentations evoke LA riding to activate amygdala, to then trigger fear responses via projections to hypothalamus and brainstem
91
LA responses to CS are ______ - so the prelimbic cortex...
LA responses to the CS are TRANSIENT
so the PRELIMBIC CORTEX (PL) is necessary for SUSTAINED FEAR EXPRESSION
92
what does contextual learning of fear involve?
the hippocampus
hippocampus generates CONTEXTUAL REPRESENTATION, which becomes associated with the US in the amygdala
93
PL cortex is critical for...
sustained fear response
94
IL is critical for...
learning, consolidation and retrieval of EXTINCTION
following extinction learning, activity of IL neurons modulates fear expression via the intercalated masses (ITC)
ITC = inhibitory cells that inhibit LA signalling to the CeA
95
PL in humans
dACC
96
IL in humans
vmPFC
97
what does P10 coincide with in rodents?
onset of fear learning
and the onset of synaptic plasticity in the amygdala
98
when does contextual fear learning emerge in rodents? when is it later suppressed?
emerges at P24
suppressed between P29-P48 (adolescence)
99
is extinction also developmentally regulated?
yes!
in pre-weaning rats, EXTINCTION is PERMANENT (may represent unlearning)
doesn't show normal phenomena of renewal, reinstatement and spontaneous recovery
100
what happens to extinction during adolescence?
it's suppressed
adolescent rats show reduced extinction - which is reflected in lack of involvement of IL and PL
101
developmental changes in fear learning circuits are likely...
adaptive!
1. adolescence is a TIME OF EXPLORATION and animals must leave safety of home environment
^ suppressing contextual fear may be adaptive
2. yet specific dangers remain, so cued fear learning is adaptive, and resistance to extinction may further increase caution
102
paper: "________ ______ integrity is related to early life adversity and _____ ______"
ACCUMBOFRONTAL TRACT integrity is related to early life adversity and FEEDBACK LEARNING
103
corticostriatal circuit
plays key role in MOTIVATION, REWARD RESPONSES and REWARD LEARNING
involves ventral striatum (VS aka nucleus accumbens), VTA, vmPFC
104
effects of ELA on mPFC and VS...
have been reported
(and these are part of the corticostriatal circuit)
ELA = associated with reduced mPFC volume and changes in VS activity
what if ELA alters the CONNECTIVITY BETWEEN mPFC and VS?
105
accumbofrontal tract
(part of the corticostriatal circuit)
WHITE MATTER TRACT BETWEEN THE mPFC AND VS
individual diffs in structure/functional connectivity of mPFC and VS = related to diffs in IMPULSIVITY, REWARD LEARNING and FEEDBACK SENSITIVITY
106
how did study investigate if accumbofrontal tract and altered reward processing were both affected by ELA?
used validated MEASURE OF CHILDHOOD ADVERSITY (quantified individual differences in ELA)
and 'state-of-the-science' technique for QUANTIFYING WHITE MATTER INTEGRITY (quantitative anisotropy)
107
quantitative anisotropy (QA)
measure used in diffusor tensor imaging (DTI)
(DTI = similar to DWI but more complex and detailed)
QA provides sensitive measure of structural integrity
low QA = low white matter structural integrity
108
QA ELA - study's primary hypothesis
higher ELA will be related to lower white matter integrity in accumbofrontal tracts
109
QA ELA - study's secondary hypothesis
lower white matter integrity will be related to MALADAPTIVE DECISION MAKING
as measured by ALTERATIONS IN POSITIVE and NEGATIVE FEEDBACK SENSITIVITY
110
results: association between ELA and QA
it was significant in both the LEFT and RIGHT hemispheres
high QA = low ELA
high ELA = low QA
SUPPORTS PRIMARY HYPOTHESIS
111
white matter tract integrity and feedback sensitivity
LOWER white matter integrity was related to NEGATIVE FEEDBACK SENSITIVITY for both left and right hemispheres
(this remained significant when controlling for general cognitive ability)
NO ASSOCIATIONS for white matter integrity and POSITIVE FEEDBACK
112
what's the relationship between ELA, accumbofrontal tract integrity and negative feedback sensitivity?
performed a MEDIATION ANALYSIS to get at this
NO DIRECT RELATIONSHIP between ELA and negative feedback sensitivity
but there is an INDIRECT RELATIONSHIP:
ELA > accumbofrontal tract integrity (QA) > negative feedback sensitivity
113
computational psychiatry
uses computational modelling to capture the UNOBSERVABLE UNDERLYING PROCESSES that drive observable changes in behaviour
114
data-driven approaches (data analysis) capture _________ relationships in the data, and remain _________
capture STATISTICAL relationships
they remain DESCRIPTIVE
(they don't inform us about mechanism)
115
goal of mathematical modelling and information processing modelling
to understand the RELATIONSHIPS WITHIN THE DATA by applying a formal characterization of the UNDERLYING PROCESSES through which these data might have arisen
116
what should a useful computational model be able to do?
1. REPRODUCE the data
2. PREDICT data from other experiments
117
2 types of questions computational models can be used to answer?
1. what's the BEST EXPLANATION for the data?
^ comparing BETWEEN models
2. what are the values of the FREE PARAMETERS of the same model that BEST FIT THE DATA of each individual? (parameter estimation)
^ comparing WITHIN models
118
MDD may result in different weighting of ________
effort (WEIGH EFFORT MORE HIGHLY)
people with MDD weigh effort costs more highly than non-depressed control
antidepressant drugs reduce effort costs
119
changes in learning in depression
(suggested by computational modelling)
1. reduced ability to learn about NEGATIVE OUTCOMES could make AVERSIVE EVENTS MORE SURPRISING
2. increased surprise INCREASES 'PREDICTION ERROR'
3. prediction errors CORRELATE with MOOD - this has been linked to NEGATIVE MOOD
120
what do reinforcement learning models of decision making do?
1. estimate the VALUE of a STATE or ACTION to GUIDE CHOICES about which actions should be taken to MAXIMIZE REWARD
(concept of prediction error = foundational)
121
Rescorla-Wagner learning rule
describes how animals LEARN THE VALUE OF A STIMULUS or ACTION by experience
change in predictive value after an outcome depends on HOW SURPRISING the outcome is (prediction error)
prediction error is SCALED by a LEARNING rate
122
should individuals always choose the action with highest predicted value?
no - they should explore other options
exploration = important for OBTAINING INFO about other options (which could be of higher value)
123
higher beta results in...
results in choices that CLOSELY MIRROR ACTION VALUES
less exploration
124
lower beta results in...
choices that are MORE RANDOM
more exploration
125
role of beta parameter in modulating extent to which values influence actions can be thought of as parallel to what?
to how ANHEDONIA in MDD impairs ability to use info about REWARD in decision making
some evidence that B parameter is LOWER in people with MDD (associated with MORE EXPLORATION and MORE RANDOMNESS)
126
DDMs applied to judgment bias tasks - what can it reveal?
use DECISION TIME and CHOICE ACCURACY to estimate a threshold for decision-making for diff choices
can REVEAL how DIFF MANIPULATIONS MODULATE the THRESHOLD for certain kinds of decisions
ie. can show that a NEGATIVE AFFECTIVE STATE INCREASES the THRESHOLD for an optimistic decision
127
Bayesian decision-theoretic modelling in judgment bias
another computational framework
estimates EXPECTED VALUE of EACH CHOICE given the ANIMAL'S PERCEPTION of the stimulus
reveals AFFECT-INDUCED CHANGES in OUTCOME VALUATION
ie. mild stress increased valuation of reward
SHOWS THAT JUDGMENT BIASES AREN'T ONLY DRIVEN BY OPTIMISM/PESSIMISM, but also are influenced by CHANGES IN HOW OUTCOMES ARE EVALUATED (models can dissociate impact of related but distinct processes)
128
foul smelling room study
no effect on judgment bias was found using standard metrics
but BAYESIAN MODELLING found the manipulation made Ps:
WEIGH LOSS MORE HEAVILY, which in turn led to MORE OPTIMISTIC DECISION MAKING
so the effect on valuation cancelled out the effect on judgment bias
(we wouldn't be able to have this specificity without the computational models)
129
what happened to CS+ connectivity from early conditioning to exinction?
it INCREASED across extinction learning
(no such difference was observed for CS-)
130
what does fact that CS+ connectivity increases from early conditioning to extinction suggest?
suggests that EXTINCTION LEARNING induced CHANGES IN BRAIN CONNECTIVITY that are SPECIFIC to a CS+
131
during conditioning, was there a sig change in connectivity for CS+?
no
but in late conditioning, CS+ connectivity was lower than CS- activity
(opposite of observed connectivity increase for CS+ across extinction)
132
brain regions modulated during extinction display what in terms of their connectivity during CS+ conditioning?
they have REDUCED connectivity during CS+ conditioning
(broadly consistent with previous reports that regions important for fear extinction have decreased neuronal activity during fear conditioning)
133
how are moods different from emotions?
1. they last longer
2. their triggers are less clear
3. they reflect cumulative impact of multiple events/stimuli
134
mood depends on the ______ impact of _______ outcomes
mood depends on the CUMULATIVE impact of UNEXPECTED outcomes
135
CR
value of certain reward, if chosen
136
EV
expected value of gamble
137
RPE
diff between actual outcome and gamble EV
138
w
weights
capture influence task variables on momentary happiness
139
y
exponential decay parameter that captures forgetting
140
found that what weight parameter contributed more to happiness?
RPE parameter (more than EV parameter)
(Ps levels of happiness were best explained by model based on RPE and not overall earnings)
141
happiness depends not simply on how well things are going, but on...
whether they are GOING BETTER THAN EXPECTED
(likely involves dopamine signalling and BOLD activity in ventral striatum)
142
participants who won Wheel of Fortune chose to replay which game?
game 2
and those who experienced a loss preferred game 1
SUGGESTS that mood manipulation INFLUENCED HOW they experienced the two games that were otherwise the same
143
mood is an efficient way for general environmental factors to...
influence reward expectations
without having to DIRECTLY LEARN the impact of these factors on EVERY STATE
144
mood can capture _________ in sources of ______ to increase what?
mood can capture INTERDEPENDENCIES in sources of REWARD to increase LEARNING EFFICIENCY
145
when is a standard learning algorithm optimal in estimating expected reward?
if reward is determined by the state we're in
and if each state is independent
146
a mood-informed learning algorithm
standard learning algorithm doesn't cut it
states are not independent - multiple states are affected by GENERAL ENVIRONMENTAL CHANGES
so it's more efficient to UPDATE EXPECTATIONS OF ALL STATES TOGETHER
147
mt, nt and ft in the mood informed learning algorithm
mt = tracked average of all recent RPEs
nt = a learning rate
ft = scaling factor (degree to which mood factor influences updating)
148
positive and negative moods are adaptive and useful insofar as they persist only...
only until expectations are updated to align with changes in environment
149
potential outcome learning mechanism of depression
failure to learn about negative outcomes (overly optimistic expectations)
leads to negative RPEs
leads to negative mood
150
hypothetical mechanism of mood instability in bipolar
slightly lower negative learning rate
leads to negative surprises > lower mood > biased perception
bias perception = overly low expectations
overly low expectations leads to positive suprises
positive surprises > higher mood > biased perception
feedback loop
151
paper: "depression is associated with reduced __________ ______ in a dual valence, magnitude ______ ______"
depression is associated with reduced OUTCOME SENSITIVITY in a dual valence, magnitude LEARNING TASK"
152
how can RL frameworks probe depression and decision making
can identify SPECIFIC CAUSES of changes in decision making
ie. look into OUTCOME SENSITIVITY and CHOICE STOCHASTICITY
153
how is estimated value of a choice updated?
based on the difference between the PREDICTED VALUE of the outcome and the ACTUAL OBSERVED OUTCOME
^ the RPE
(and the RPE is scaled by a LEARNING RATE)
154
what kind of outcomes do most studies use in probing MDD and decision-making?
1. BINARY OUTCOMES
ie. 1 or 0
2. only probe learning about POSITIVE and NEGATIVE outcomes on SEPARATE TRIALS (or even separate tasks)
155
reinforcement learning in MDD - central hypothesis and primary prediction
CENTRAL HYPOTHESIS:
^ reinforcement learning is DISRUPTED in depression
PRIMARY PREDICTION:
^ current MDD will show IMPAIRED REWARD SENSITIVITY but NO DIFF in learning rate
STATE or TRAIT:
^ no prediction
156
computational modelling results
1. NO EFFECT ON LEARNING RATE
2. SENSITIVITY to REWARD and LOSS is significantly LOWER in Ps with CURRENT MDD
3. no diff between remitted MDD and healthy controls
^ suggests this is a STATE effect (not a trait effect)
157
what could the interpretation as a reduction in outcome sensitivity alternately be explained by?
change in CHOICE STOCHASTICITY
(increase in randomness)
158
paper: different components of ______ ______ ______ affect specific _______ _______"
different components of COGNITIVE BEHAVIOURAL THERAPY affect specific COGNITIVE MECHANISMS:
applies computational modelling approach to understand underlying mechanisms of diff types of therapeutic interventions with a focus on psychotherapy
159
CBT - behavioural and cognitive therapies are administered...
TOGETHER
as parts of the same treatment program
yet they DIFFER in underlying theory
do BEHAVIOURAL and COGNITIVE APPROACHES work through COMMON or DISTINCT mechanisms?
160
behavioural approaches emphasize what?
primary importance of behavioural change
ie. planning action (does this work on reward sensitivity or outcome sensitivity?)
161
cognitive approaches emphasize what?
primary importance of cognitive change
ie. attributions
162
what does behavioural activation therapy do?
helps individuals INCREASE ENGAGEMENT in REWARDING ACTIVITIES
to target LOW MOOD
ie. acting according to a PREDETERMINED PLAN increases likelihood of acting and therefore of experiencing rewards
163
how could behavioural activation therapy be boosting activity?
1. by making potential rewards SEEM MORE SALIENT (increasing sensitivity to rewards)
2. by REDUCING THE PERCEIVED EFFORT required to obtain the reward (reducing effort sensitivity)
164
role of goal-setting in behavioural activation therapy: created a task that probes what?
probes "how to decide when rewards are worth effort"
administered task BEFORE and AFTER planning/goal setting intervention
used comp modelling to examine extent to which each construct (reward sensitivity/salience and effort sensitivity) is ALTERED by the INTERVENTION
165
goal-setting intervention
based on behavioural activation therapy for low mood
1. read TEXT describing importance of setting goals
2. completed comprehension quiz about the text
166
results: goal-setting intervention did what?
REDUCES EFFORT SENSITIVITY
1. Ps in goal-setting condition showed reduced effort sensitivity pre and post intervention
2. controls didn't show this
3. no effect on reward sensitivity
167
researchers did what to probe effects of cognitive therapy?
1. designed task to PROBE ATTRIBUTIONS
2. administered the task before and after a COGNITIVE RESTRUCTURING INTERVENTION
3. used computational modelling to examine extent to which Ps CHANGED THEIR CAUSAL ATTRIBUTIONS after intervention
168
cognitive therapy experiment: wanted to see to what extent improved mood reflected...
1. DECREASED 'DEPRESSOGENIC' ATTRIBUTIONS
(internal/global for neg events)
or
2. INCREASED SELF-PROTECTIVE ATTRIBUTIONS
(internal/global for pos events)
169
cognitive intervention led to a reduced probability of what?
of endorsing INTERNAL ATTRIBUTIONS for NEGATIVE EVENTS
170
does each intervention (cognitive and behavioural) only impact the process it was designed to modulate?
YES
1. goal setting, but NOT COGNITIVE RESTRUCTURING, reduces effort sensitivity
2. cognitive restructuring, but NOT GOAL SETTING, reduces internal attribution of negative events
171
do magnitude of intervention effects relate to individual symptom profiles?
Ps completed a number of measures to study this
hypothesized there are 2 LATENT TRAITS in symptom data that will differentially INFLUENCE SENSITIVITY to interventions
1. behavioural amotivation (anhedonia and apathy)
2. negative cognition (neg self beliefs)
172
higher behavioural amotivation predicted what?
1. higher effort sensitivity
2. higher response magnitude to goal-setting intervention
SUGGESTS INDIVIDUALS HIGHER IN AMOTIVATION HAVE HIGHER BASELINE EFFORT SENSITIVITY
and are MORE RESPONSIVE TO EFFECT OF GOAL-SETTING INTERVENTION in reducing effort sensitivity
(but this didn't fully replicate so need more work)
173
higher negative cognition predicted what?
1. lower tendency to attribute positive events to internal causes
2. higher negative cognition DID NOT SIGNIFICANTLY PREDICT MAGNITUDE OF CHANGE FOLLOWING INTERVENTION
174
there's some evidence that higher behavioural amotivation predicts what?
1. greater response to goal setting
2. attenuated response to cognitive restructuring
SO BEHAVIOURAL AMOTIVATION MAY BE RELATED TO BOTH behavioural and cognitive effects
