MRI and fMRI Flashcards
1
Q
MRI Basics
A
- the technique used in radiology and research to form pictures of the anatomy and physiological processes of the body
- use strong magnetic fields, magnetic field gradients, and radio waves to create these images
- widely used in medicine for diagnoses, staging of disease and follow-ups without exposing the body to radiation (e.g. in a CT)
- basically maps the location of water and fat in the body (because hydrogen atoms - which are what is being detected - are abundant in water and fat)
2
Q
Clinical Relevance
A
- preferred over CT for investigating neurological cancers as it offers better visualisation of certain brain regions such as the brainstem and cerebellum
- provides contrast between grey and white matter —> helpful for diagnosing demyelinating diseases, epilepsy, and dementia among others
- used in surgery for treatment of brain tumours, malformations, and other surgical brain procedures
3
Q
MRI for molecular imaging
A
- MRI has high spatial resolution and can be used for both morphological imaging (anatomy) and function imaging (in fMRI)
- MRI can be used to do molecular imaging of disease biomarkers using contrast agents (basically like dyes or so that can be visualized)
- Recently, a new type of contrast agents have been developed that are able to trace cells with a specific mRNA via DNA labeling (was done with DTI not just normal MRI)
4
Q
MRI Safety
A
- generally, a very safe technique but can result in injury due to human error
—> certain implants are unsafe including cochlear implants and cardiac pacemakers
—> any type of metal in the body such as implants from fracture surgeries
- can cause magnetic materials to move at great speeds —> you could literally hit someone if you did not remove the objects
- some people may feel claustrophobic
5
Q
fMRI Basics
A
- uses BOLD response to look at brain activation (blood oxygenation levels)
- relies on the idea that blood flow and neuronal activity are coupled
- according to the oxygenation hypothesis, changes in oxygen usage in regional cerebral blood flow during can be directly related to the cognitive or behavioural task being done
- scanner allows subjects to
- be presented with different stimuli on a screen
- make actions e.g. pressing a button or moving joystick
—> can be used to reveal brain structures and processes associated with perception, thought, and action
6
Q
fMRI Clinical Uses
A
- changes after stroke
- used to identify critical region in patient’s brain before surgery
- pharmacological: whether and how a drug affects brain activity
7
Q
fMRI general limitations
A
- much more difficult to scan clinical populations than healthy volunteers
- e.g. tumours, lesions, and certain medication such as antihistamines can affect the blood flow but in ways that are not related to neural activity
- often difficult to lay still e.g. using head restraints or bite bars may injure epileptics if they have a seizure in the scanner
- it takes about 6s for a BOLD response to be measured after neuronal activity and takes about 10s to return to baseline —> limitation of temporal resolution
- BOLD signal is relative and does not have a true zero value such as number of spikes per second
—> tasks need to create a strong variation in BOLD signal over time
8
Q
fMRI Issues - Design (types and compare)
A
- Block design*
- two or more conditions alternate
- subtraction design
- only difference between conditions should be cognitive process of interest
Block Design Limitations
- very sensitive to noise e.g. head movements
- poor baseline means poor experiment
- no randomisation possible –> very predictable to participant
- Event-related design*
- allows ‘real world testing’
- low statistical power
- Parametric Design*
- varying levels of stimuli e.g. levels of reward
- no control needed
- variation of neuronal activity can be analyzed
- computational modeling: construct model of how brain processes stimuli and look for voxels that vary accordingly
Limitations
- less statistically powerful
- makes assumptions of relationship between stimuli and brain activity
9
Q
fMRI- baseline
A
Baseline = Baseline?
- there are always neurons firing
- Squire et al showed that resting state was associated with greater activity in memory task
- others also showed greater activity in other tasks –> maybe unconstrained thought? Wandering?
- -> better to constrain activity during resting period
10
Q
fMRI - BOLD response relating to neural activity?
A
There have been concerns about how BOLD signal responds to neuronal activity
- *Evidence from Nature papers which combined fMRI and ephys**
- BOLD correlated with local field potentials –> means they reflect input and processing, not output
- BOLD response can occur without spiking or stimulation –> thought to be an anticipatory response to tasks
- optogenetic stimulation of neurons leads to BOLD signals
- *BUT**
- still requires massive neuronal activity for response to be detected –> may not be detected
11
Q
fMRI and the neural code
A
Repetition Suppression
- if A is presented successively, the second time will elicit a reduced response
- -> can test which stimuli are considered equivalent by the brain
example: mirror neurons
- mirror neurons respond to own movements and those of others
- -> if suppressed the second time, then that area contains mirror neurons
Multivariate Analysis
- machine-learning data analysis where a computer is trained to discriminate stimuli based on patterns of activity across voxels
example:
- computer was able to make accurate predictions of participants’ responses to spontaneous changes in visual stimuli
- -> learned this from pattern of responses to stimulus features
- *Combination with Bayesian encoding/decoding**
- can be used to reconstruct stimuli from brain activity!!
