working memory Flashcards
1
Q
Milner 1963 / Dias 1997
A
- impairment on WCST is classic sign of PFC damage in humans
- unable to exert cognitive control / flexibly alter their responses to same stimuli depending on which rule is currently in effect
- Dias = primate analogue
2
Q
Funahashi (1993)
A
- PFC lesions impair spatial delayed response tasks where cue is briefly flashed at a location and monkey must direct eye movement to remembered location
- cannot maintain location and knowledge of task demands in WM
3
Q
Asaad (2000)
A
- recordings from PFC neurons in 2 monkeys during 3 types of task: spatial, object and associative
- for many PFC neurons, activity was modulated by task being performed e.g. different baseline activity
- some neurons always active to same object regardless of task, others active to object only when certain behavioural response was required (so both sensory/perceptual and higher-level decision-making processes)
- increased neuronal activity in delay during object task: could represent impending choice / preparation, or could be inhibition of actions involved in the associative task for same object
4
Q
Duncan (2001)
A
- adaptive coding model
- context-specific task parameters directly shape tuning profile of PFC
- PFC neurons adapt tuning profiles to represent input according to task relevance
so changing task parameters shifts the response properties of the network and alters the way stimuli are coded and behaviour is produced
5
Q
Stokes (2013)
A
- used dynamic pattern analysis to determine how PFC establishes, maintains and uses flexible cognitive states for task-dependent decision making
- coding initially reflected physical properties of choice stimulus, then differentiated between two alternative decision values: “go” vs “no-go”
- neuronal activity in delay period represents distinct neurophysiological state triggered during cue processing, where the tuning profile of PFC neurons have temporarily been set according to task demands i.e. tuned to classify each choice stimulus as “go” or “no-go” response signal
- data consistent with prospective coding model, with memory state configured for future task demands rather than anticipated sensory target specifically
- via activity-dependent short-term plasticity / modification of synaptic weights = tuning according to recent responses (Buonomano 2009)
6
Q
Buonomano 2009
A
- activity-dependent short-term plasticity in PFC could be mechanism for flexible tuning and maintaining info in WM
- patterned activity leaves behind a patterned change in synaptic weights of network, so subsequent stimulation will be patterned according to recent stimulation history
- in trials where stimulus was fixed, the population response was patterned according to identity of previous cue (reflection of memory according to synaptic weights / recent patterned activity)
- so PFC doesn’t maintain previous stimuli info but has access to this info and can reliably encode whether subsequent stimuli are targets or distractors
7
Q
Wallis (2001)
A
- single-cell recordings from PFC (dorsolateral, ventrolateral and orbitofrontal) of monkeys trained to use two abstract rules: indicate whether two images were same or different depending on current rule
- performed task with new images = shows general rule had been abstracted from previous experience
- 50% neurons showed higher activity for match rule and 50% for non-match = selective for specific rule
- ability to abstract rules from direct experience allows these rules to be applied to general situations = intelligent behaviour
8
Q
Riggall + Postle (2012)
A
- fMRI study of elevated neural activity of PFC during WM tasks
- didn’t encode stimulus-specific info
- instead extracted trial-specific task instructions (current rule)
9
Q
Curtis + D’Esposito 2003
A
- DLPFC does not store representations of past sensory events or of future representations/responses
- instead, its activation reflects top-down biasing control over posterior lower-level regions that actually store the representations
10
Q
Ester 2009; Emrich 2013
A
- fMRI showed feature-specific activation in early visual regions during working memory maintenance e.g. orientation-specific patterns in visual areas V1-4 when remembering orientation of a grating
11
Q
Postle (2006)
A
- retention of info in WM occurs alongside sustained activity in same brain regions responsible for representation of info in non-WM situations, so PFC is not simply a substrate for storage during WM
- instead, role of PFC is to employ control processes (e.g. attentional selection, flexible control etc) that are also required when performing WM task
- so PFC actively focuses attention on relevant sensory representation, selects info and performs executive functions necessary for cognitive processing of selected info
12
Q
Lara + Wallis (2014)
A
- complex task where monkeys had to remember colour of one or two coloured squares: large set of colours and shade discriminations could be very small
- majority of PFC neurons didn’t encode colour
- instead, encoded spatial location and passage of time. so played role in organising behaviour during the task, but didn’t reflect contents of WM
- when task was more difficult, monkeys made microsaccades which reflected covert attention so item was represented more precisely in WM
- so PFC neurons encoded attentional control signals (during covert attention) which extract sensory representations from lower visual areas and improved their performance
13
Q
Roux (2012)
A
- task where participants remembered 3 red discs, 3 red discs while ignoring 3 blue discs or 6 red discs
- increased oscillatory gamma activity in both PFC and parietal areas, but only PFC reflected task-relevant information (parietal also encoded distractors)
- so increased ongoing oscillatory activity during WM tasks in both frontal and posterior areas
- supports theory that PFC and posterior areas interact via coupling of oscillatory activity in both areas
14
Q
Pasternak (2015)
A
- delayed match-to-sample task with random dot stimuli of varying motion coherence
- lesion to lateral PFC impaired ability of monkeys to remember direction of motion, but didn’t depend on motion coherence (i.e. stimulus properties)
- also more impaired for same stimulus in certain locations
suggests PFC plays role in attending to stimuli and shifting attention to motion in other areas
15
Q
Liebe (2012)
A
- monkeys performing visual WM task
- simultaneous recording from lateral PFC and V4
- increased synchrony between theta phases in PFC and V4 during delay period
- stronger synchrony when monkeys successfully maintained info in WM / remembered stimulus
- provides explanation of mechanism for how info is shared between both areas during maintenance of WM
