Seminar 1 - Neuroimaging: Structural & Hemodynamic Methods Flashcards
The triangle of cognitive neuroscience
Computational model
Neural architecture:
- neuro-physiology
- neuro-anatomy
- neurology
Cognitive architecture:
- neurology
- behaviour
- psycho-physics
- cognitive psychology
Empirical methods - experimental confirmation, exportation and debunking
Overview over most important methods
Behaviour:
- self-report
~ questionnaires
~ interviews
~ free association
- observable
~ reaction time
- eye-gaze or pupil dilation
- errors
Brain:
- hemodynamic
- electrophysiological
- anatonomic/neurophysiological
- brain stimulation
Body:
- psychophysiology
~ heart rate
~ electrodermal activity
~ respiratory rate
These are all important even when considering neuroimaging, as we may be doing some of these while carrying out neuroimaging techniques
Brain imaging methods
Hemodynamic and anatomical/neurophysiological methods
Magnetic resonance imaging (MRI)
Invented in 1980s
Can scan band create images of brain structures
Cabin test structural differences in brain of different groups of people (e.g. do taxi drivers have bigger hippocampus than non tax drivers - brain structure that is involved in navigating and spatial memory)
How do we take images of a 3D organ
Slices
There are different planes can be sliced on
- coronal plane - through the brain parallel to the face
- horizontal/axial plane - horizontally y through the Brain
- Sagittal plane - through the brain perpendicular to the face and parallel to the ears
Slices are then used to reconstruct a 3 model
Used interpolation and averaging
The method:
Scanning from the bottom of the bran to the top, one at a time, meaning that there is a 2 second difference between the first photo and the second photo, which can cause issues as that is a long time with the function of the brain - which is when interpolation and averaging is done
How do we get from static images to brain function?
Short answers: we are taking a lot of functions in rapid succession while people are doing things
Functional MRI - the brain during cognitive tasks
Enables us to look at brain function
Utilise what we know about cell metabolism to make inferred about neural function
Bold activity = blood-oxygen level dependent
MRI scanners
MRI machine is powerful electromagnet
- 1.5-7.0 Tesla
- Earth electromagnetic field is 0.25-0.6mT
- so even the weakest MRI scanner is 30.000 times stronger than the electromagnetic field of the earth
Can be rather dangerous, you can have absolutely no metal near by or in the body because of how strong they are - you can’t have any metal on your body or have metallic implants
What is BOLD?
Neurones need oxygen
Oxygen is transported by haemoglobin from the lungs to the brain cells via blood flow
The MRI signal is sensitive to deoxyhemoglobin
During brain activity, the level of deoxyhemoglobin in the blood changes
Why is a magnet sensitive to blood flow?
Red blood cells carry iron
Iron is stabilised though oxygen but destabilised when there is no oxygen, hence why MRIs are sensitive to deoxygenated blood
The hemodynamic response function (HRF)
With fMRI we can’t measure neurone spiking
We can only measure the energy cravings of the neurone after it did its thing
Neural activity -> increased blood flow -> decreased deoxyhaemnoblin -> decrease in local magnetic field inhomogeneity -> local increased signal: the blood oxygen level dependent (BOLD) signal
Positron Emission Tomography (PET)
Very similar to fMRI, same machine, but uses a contrast agent
- radioactive tracker substance injected to make blood flow more visible due to positron emission - hence why is seen too be very bright
- no worries, very safe
- more direct measure of blood flow rather than magnetic field differences based on deoxygenating differences, making it potentially more accurate - seeing overall blood function
- however you see the blood vessels and not the brain, so you can’t see small brain images
PET vs fMRI
PET:
- better signal to noise ratio
- higher sensitivity to blood flow
- slower data acquisition (scans per minute)
- spatial resolution of 5-10mm
- movement less problematic - because it already so blurry
- scans independent - you aren’t taking slices and adding them together
- faithful map of rCBF
FMRI:
- clear image
- lower sensitivity to blood flow
- fast data acquisition (one image can be acquired in as little as 150ms)
- spatial resolution of 1-2mm
- very prone to movement artefacts (=noise) - need to breathe, sneeze, tremors, can’t stay still for such a long period of time
- scans correlated
- some brain regions susceptible to spatial distortion or signal loss
What is the best imaging technique?
Depends on your research question
The considerations are:
- speed - PET much slower than fMRI, tradeoff between temporal and spatial revolution
- sensitivity - PET more sensitive than fMRI, signal changes are small, not equal representations of different parts of the brain
- accuracy - fMRI is more accurate than PET. Better to detect locations, especially if small brain areas
How do we get to see specific brain functions?
The subtraction method
- Franciscus Donders (Dutch physiologist)
If you measure 2 variations of the same mental function that only differ in one expect, the observed difference between these two variations will be attributed to that 1 aspect in which they are different
Td=t2-t1
The fMRI experiment
Averages of activity were taken over 100 tasks, the the same with the second variation and subtract
Tasks they were asked to and the resulting PET scan activity:
- look at fixation cross -> activity related to sitting upright and stating at computer screen
- silently read word (hammer) -> activity related to sitting upright, staring at a computer screen and reading a word
- read word out loud -> activity relates to sitting upright, staring to a computers screen, reading a word and speaking this word out loud
Say related verb (e..g hit) -> activity related to sitting upright, staring to a computer screen, reading a word, searching and selecting an associated verb and speaking this word out loud
Note: the brain is always 100% active, there is never an area in the brain that is resting, hence why we need to do the subtraction method, so we can see what areas are affected in a certain task
