#10622 - The Neural Basis Of Amnesia 2 Lec 4 - Neuropsychology of Memory

The Neural Basis of Amnesia (2)

Episodic Memory

• Tulving (1972) first coined the phrase ability to remember personally experienced events – what, where and + when an event occurred

• Tulving (1983) elaborated original definition – proposed that EM necessarily depends upon autonoetic awareness (conscious re-experiencing of an event)

• Machphail (1998) – also maintained that consciousness is a crucial prerequisite for all forms of explicit memory

• This def present major problems – impossible to demonstrate pure EM in animals as they cannot communicate consciousness through language (Griffiths et al., 1999)

• This issue been resolved to an extent through the introduction of the term episodic like memory (ELM) has all the characteristics of EM – what, where + when and event took place – with the exception that there is no requirement to demonstrate autonoetic awareness (Clayton & Dickinson, 19898) – therefore can study in animals

• Tulving (1983)

• EM = memory which “receives + stores info about temporally dated episodes or events (what), and temporal-spatial relations (where-when) between them”

RECALL V FAMILIARITY

• Episodic memory is declarative in nature – it involves calling to mind + recalling the specific event

• Changes in behaviour caused by previous experience does not need episodic memory e.g. – procedural learning – same may be said for recognition memory based on familiarity – remembering something merely because it is familiar

Aggleton & Brown (2006)

• Recent findings reinforce the view that recognition memory comprises at least 2 independent processes -> one recollective + other using familiarity detection Only recollective recognition seems to depend on episodic memory

• Investigated the crucial issue of whether episodic + recognition memory reflect the same underlying processes + extent to which certain brain structures work as a single unit to support these processes

• Attempts to find the neural basis of these functions indicate that these two components depend on separate but interlinked structures

• Authors support two-process models of recognition that allocate different processes to separate but different, interconnected, brain structures.

• THERE ARE TWO SEPARATE COMPONENTS OF RECOGNITION MEMORY (FAMILIARITY + RECOLLECTIVE) – THESE HAVE DIFFERENT NEURAL SUBSTRATES

• Recall/recollection = hippocampus + anterior thalamus

• Familiarity = perirhinal + parrahippocampal cortex + medial dorsal thalamus

Evidence for this:

• Event related potentials

• 2 different anatomical populations have been identified that are functionally + temporally dissociable

• First pop = indexed by activity over the frontal scalp from 300-500ms post stimulus - neural correlate of familiarity

• Second pop = evident over the parietal scalp from 500-800ms post stimulus + might index recollection

• fMRI

• number of fMRI studies support dual processing accounts of recognition memory because dissociable patterns of activity are found for measures of familiarity + recollection

• Increased hippocampal activity has been correlated with reports of recollection of the learning episode but not with familiarity

• Some research has linked hippocampal activity with recollection but anterior parahippocampal cortex activity with familiarity

The MTLMS doesn’t contain just one memory circuit (Squire)– but two components

Suggested that episodic memory is recall based + semantic memory is familiarity based:

• Hippocampus + surrounding cortex has become known as the medial temporal lobe memory system (MTLMS)

• If familiarity based judgements reside in the parahippocampal + perirhinal cortex + recall judgements reside in the hippocampus what happens when you lesion the surrounding cortex (means they cannot use familiarity - semantics to solve the DMTS task) Eacott, Gaffan & Murray (1994)

• DMTS can be aided by both recall + familiarity (episodic + semantic memory normally interact strongly) – so this is only seriously impaired by damage to the hippocampus + other components of the MTLMS

• Considering previous research in these terms

• HM = severe amnesia + damage to the entire MTLMS recall + familiarity impaired

• RB = milder amnesia + only select lesion to CA1 cells in hippocampus – episodic only

• NA = damage restricted to the thalamus – probs in both recall + familairty because (see below)diff parts of thalamus involved in recall + familarity

• Squire suggested that the more MTLMS you damage the greater the amnesia – this is supported by Aggleton + Brown (2006) – as they show that the severity of amnesia demonstrated on many tasks will depend on the severity of impairment in both episodic + semantic memory (because of the two different neural substrates they possess – hippo + extra cortex)

What about diencephalic amnesics? E.g. Korsakoffs + NA

• Diencephalic amnesics have damage residing in the thalamus

• The thalamus receives connecitons from both the hippocampus + other components of the MTLMS

• Therefore damage to the diencephalon can disrupt function of the recall + familiarity circuits

• Different parts of the thalamus involved in recall + familiarity

• Recall = anterior thalamus

• Familiarity = medial dorsal thalamus

MUST BE AWARE THAT SOME TASKS GIVEN TO ANIMALS MAY JUST TAP FAMILARITY

DMTS Tasks

• This task only requires animals to recall one aspect of a particular episode.

• Doesn’t assess the ability of the animal to recall an integrated percept of what, where and when an event took place (Griffiths & Clayton, 2001).

• Tasks which require the animal to recognise only one aspect of a stimulus are most likely solved by familiarity processes rather than episodic recall (Griffiths et al., 1999).

• Zola-Morgan et al (1993) lesioning several different combinations of areas in the monkey

• They demonstrated that their H++ lesion which includes the hippocampal formation, perirhinal cortex, - but sparing the amygdala produces the greatest memory deficit in the amnesia sensitive task of DMTS

Evidence for ELM in animals

• Problem of familiarity being used to solve ELM task resolved to an extent with Clayton & Dickinson (1998)

• demonstrated that food caching scrub jays form an integrated memory for what, where an when details of their unique caching episodes.

• Birds allowed to cache two types of food worms (preferred but perishable) + nuts (non-preferred and non-perishable).

• After training, animals became aware of these differing decay rates.

• Then allowed to recover their caches after either a short or long delay.

• Birds preferentially recovered the preferred wax worms after short delays between caching and retrieval and when a longer interval elapsed, they preferentially searched for the nuts

• This alternative recovery preference = evidence that these animals form episodic like memories of unique caching episodes encoding info about the content, location and timing of their cache into an integrated what, where, when representation of the event (Clayton et al., 2003).

This fulfils Clayton, Bussey & Dickinson’s (2003) behavioural criteria for ELM:

1. Content recollecting what happened, where + when – on the basis of specific past experience

2. Structure forming an integrated – what, where, when memory representation

3. Flexibility set within a declarative framework – so involves the flexible deployment of info – e.g. different temps leads to differing decay rates – they realised this

DOES THIS ABILITY RELY ON THE HIPPOCAMPUS?

Pravosudov & Clayton (2002)

• Tested the hypothesis that accurate recovery is more critical for birds that live in harsh conditions - food supply is limited and unreliable

• Compared food caching memory + the hippocampus of black- caped chickades from Alaska (harsher) + Colorado

• Alaskans cached sig more food + were more efficient at cache recovery

• Alaskans also performed more accurately on one trial associative learning tasks in which birds had to rely on spatial memory - didn’t differ on non-spatial version of the task

• Alaskans had sig larger hippocampal volumes containing more neurons that the colaradian

• Population differences reflect adaptations to the harsh environment – increased hippocampal to deal with greater memory demands

• Suggestive that this ability for EM resides in the hippocampus

Problems:

• Clayton & Dickinson’s “what – where-when” model is good for birds but rats + monkeys cannot do this

• dependence on hand reared birds that possess a natural propensity to cache food for future use, makes extending this paradigm to other animals problematic (Eacott & Norman, 2004).

• Research demonstrating this form of episodic like memory in other species has had limited success.

• Experiments in both rats (Bird et al.,2003) and monkeys (Hampton, Headstead & Murray, 2005) have successfully demonstrated memory for what and where but have failed to document a temporal (when) memory component

• Been argued that animals which do possess memory for when an event happened, may simply remember when an event was more or less recent, enabling them to establish how long ago the event occurred without recalling an absolute time period (Eacott & Easton, 2010).

• Also been argued that animal’s ability to remember when an event occurred is more likely explained by the use of circadian cues (Hampton & Schwartz, 2004).

• Many humans may have a detailed memory of an event in terms of what happened and where it occurred, but rarely possess exact knowledge about the time and date of this event (Friedman, 2007).

• This questions whether a detailed and explicit representation of when an event occurred should be a prerequisite for ELM in animals.

• ...