week 4 - neuroimaging of memory Flashcards
What are strengths of using fMRI to study memory?
- Capture dynamic, spatiotemporal changes (encode, store, retrieve)
② Applicable to study both groups and individuals in health and disease
③ Provides a whole-brain measure of brain structure and function
④ Test hypotheses about specific regions (hippocampus) and their networks
⑤ Neuroimaging can create hypotheses where cellular and molecular probes can be
applied or for invasive animal studies
What are the types of memory
Sensory Memory (< 1 second)
Short-Term (Working) Memory (< 1 minute)
Long-Term Memory (Lifetime)
- Implicit (Unconscious)
- Procedural (Skills, tasks)
- Explicit (Conscious)
- Declarative (Facts, events)
- Episodic (Events, experiences)
- Semantic (Facts, concepts)
What are weaknesses of using fMRI to study memory?
① Proxy measures of neuronal activity
② Not a measure of “structure” or “connections” per se - instead looking at correlatins
③ Results subject to methodological variation (Botvinik-Nezer et al. Nature 2020)
④ High costs associated with data collection, storage, and processing
⑤ Multiple reasons for a structural MRI signal change
What is evidence (fmri) that the hippocampus is related to spatial memory
Maguire et al, 1998.
Got mice to navigate through a virtual maze. Looked at where the brain was using glucose the most on average during this task.
This showed activation in the hippocampus
What did Moser et al. (1995) show about the hippocampus and spatial memory?
Moser et al. trained rats in a water maze and found that lesions to the hippocampus impaired their ability to learn and recall the platform location.
The study demonstrated that the hippocampus is critical for spatial learning and memory, particularly for flexible navigation strategies.
Rats with intact hippocampi could form spatial maps, while lesioned rats relied on non-spatial cues, indicating a specific hippocampal role in spatial encoding.
What did Lever et al. (2002) contribute to our understanding of hippocampal function?
Lever et al. recorded from neurons in the rat hippocampus and identified place
cells that fired selectively when the animal occupied specific locations in its environment.
They showed these cells form part of a spatial representation (cognitive map) of the environment, dynamically updating as the animal explores.
This provided strong neurophysiological evidence that the hippocampus encodes spatial information at the single-neuron level.
What did Maguire et al. (1998) discover about the hippocampus and spatial navigation in humans?
Maguire et al. found that London taxi drivers had increased posterior hippocampal volume, correlated with years of navigation experience.
This showed structural brain plasticity linked to long-term spatial memory.
Supporting studies also showed increased glucose utilisation in the human hippocampus during virtual maze navigation, indicating heightened metabolic activity during spatial tasks.
Together, these findings confirm that the hippocampus is both structurally and functionally specialised for complex navigation.
What does neuroimaging of the DMN tell us about memory?
robin et al 2015
- Spontaneous activity within the hippocampus is synchronized (functionally connected) with a
network of brain regions known as the default mode network (DMN). - The functional association of DMN with the hippocampus may vary based on cognitive
demands.
What is evidence (task based fmri) that the hippocampus is related to spatial memory
ryan et al 2009
- fMRI shows preferential involvement of hippocampal activity during episodic and spatial
memory tasks - Activity defined as increased blood flow (perfusion) or BOLD signal contrast increase
- Yet this tells us nothing about relevant functional connections and wider network changes
How can fMRI study structural plasticity in memory (maguire et al 2000)
Include criticism and their response to this
Classic taxi driver study
Hippocampus enlarges after the knowledge
criticism:
Does this reflect training or innate expertise?
* Performance in a virtual navigation task amongst an unselected group of people was not predictable based on hippocampal volume
* The “taxi-driver effect” may be a specific consequence of extensive training
* Not explainable as an aspect of population variance (underpowered?).
Outline why neuroimaging studies may not replicate / why might evidence be ambiguous
- Sample size - using a small sample can be non replicable (Masouleh et al, 2019)
- Variability in analysis - They might not take into account covarience
- Variability in statistical correction - leniently or stringently correcting for multiple comparisons
- Varaibility in analysis of hippocampus anatomy
Outline clark et al (2020) on hippocampal volume and memory
tested the relationship between hippocampal grey matter
volume and scene imagination, autobiographical memory,
future thinking and spatial navigation task performance
n = 217 (large)
Little evidence that hippocampal grey
matter volume is related to task
performance
* Irrespective of methodology or
statistical approach
* Hippocampal volume may not
significantly influence performance
on tasks requiring the hippocampus
in healthy people
Outline strengths of small animal neuroimaging when studying memory
Address the lack of continuity
between levels of analysis in
animal models and clinical
research
② Test hypothesis regarding
influence of genetic, of
environmental risk factors or
putative confounds
③ Link neuroimaging signal
changes to their molecular and
cellular correlates
Describe Lerch et al’s 2011study using small animal neuroimaging to study memory
READ THIS PAPER
Lerch et al. Neuroimage, 2011
Spatial maze variant led to specific growth of the mouse hippocampus
* Cued version (non-spatial) led to striatal growth
* Replicates findings in humans based on different navigation strategies
* Compare in mouse models for familial Alzheimer’s disease?
When they looked at other brain regions, they did see that there was activation in other regions too
IS THIS TRAINING OR EXCERCISE?
Learning content
differed across
protocols
* Swimming content
(proxy exercise?) also
differed
* What if we compare
exercise matched
groups?
* Key volume differences
in striatum and
hippocampus remained
Possible cellular hypotheses to account for
the volume changes :
(1) alterations in neuron numbers/sizes
(2) alterations in astrocyte numbers/sizes,
(3) increased neurogenesis
(4) remodeling of neuronal processes.
Carried out immunostaining and quantification
of:
* NeuN (neurons)
* GFAP (astrocytes)
* DCX (new-born neurons)
* GAP-43 (axon marker
Axon/growth marker correlates with Δ volume
Lerch et al. Neuroimage, 2011
* Hippocampal GAP-43 staining
correlated positively with behaviour
* All structure volumes correlated with
GAP-43 staining
* GAP-43 implicated in memory storage
* Binds to cytoskeletal proteins actin and
fodrin
* Causes presynaptic morphology
changes
How can neuroimaging track memory decline in ageing and dementia?
Cognitive deficits in normal aging have been associated with hippocampus pathology
① Accumulation of pathological proteins
② Loss of synapses
Hippocampus volume, when controlling for age, can be a predictor of dementia
Individuals with dementia also show different functional connectivity signatures
How can you use neuroimaging to measure the therapeutic benefits of excercise
Erickson et al. PNAS, 2011
Exercise enhances learning and improves
retention
* RCT of exercise intervention in 120 older
adults
* Exercise training increased hippocampus
volume vs. non-exercising controls
* Positively associated with Spatial memory
scores and serum BDNF levels
