fMRI: L8 Flashcards
what does oxygenated blood, oxyhemoglobin (Hb) do to the signal?
it is diamagnetic, enhances the signal
what does deoxyhemoglobin (Hb) do to the signal?
introduces field distortions, decreasing the signal
what do neurons need?
highly oxygenated blood for glucose metabolism
what is neural activity accompanied by?
a local oversupply in oxygenated blood and therefore a better blood oxygen-level dependent (BOLD) signal
What is being measured?
what does this conclude?
- the oxygen content NOT the neurons themselves
- something about neurons in brain regions
what do large blood vessels cause in scans
brighter areas = better signal
- areas of enhanced activity have…
2. this can be…
- more oxygenated blood
2. mapped onto a structural image of the brain called statistical parametric mapping
what can statistical parametric mapping say about the brain?
that there is a significantly strong activation in a particular region during a cognitive task or for task A compared to task B
red indicates
blue indicates
- more oxygen
- less oxygen
what is measured during a cognitive task?
BOLD signal = blood oxygen-level dependent
what is required for the whole brain while performing experimental tasks and why?
repeated measures because the signal is very noisy
what does the BOLD signal tell us?
-> doesn’t tell us
whether a brain region is engaged in the task or not
-> doesn’t directly measure neural activity, it is an INDIRECT measure of neural activity
what does enhanced neural activity impact? (3)
- how much oxygen is in the blood
- how fast the blood flow is
- how much oxygen can be extracted
when blood flow is faster how much oxygen can be extracted?
LESS because in total more oxygen is extracted per unit time
what is the temporal lag? & how many s?
lag between neural activity and the peak of the BOLD response - 8 seconds
how long does the BOLD signal need before reaching baseline?
baseline reduction is therefore…
- 16 seconds
- slow
Why can we not compare signals between different regions of the brain?
the signal change is different therefore it would be invalid to compare them
the measured response is described by the what?
what does it look like across regions?
- Haemodynamic response function (HRF) -> line on the graph
- looks v similar across different regions
the BOLD signal DOESN’T reflect
neurons firing/ neuron output
what does Logothetis (2008) describe local cortical excitation-inhibition networks (EIN) as?
small and highly interconnected functional microunits, which show massive recurrent feedback
What does feedback processing in EINs account for?
most activity measured by BOLD fMRI
Limitations of BOLD fMRI (2)
- we do not see the full networks involved during cognitive functions (we only see the rip of the iceberg)
- fMRI might not always map the units that matter
Limitations of BOLD cont. (1)
- poor temporal resolution = HRF is slow & fast processes are difficult to image. it takes 2s to measure the brain (we cannot explore processing in this time). We have to wait for HRF to reach baseline every time (16s)
Limitations of BOLD cont. (2)
- smallest measurement unit is a “voxel” = 3D pixel. 1 voxel contains >100,000 neurons!
- risk of rejecting the Ho for each test is 1% ( t-test conducted), we may therefore have to calculate loads of tests (est. 50,000 x) to find false positives
what is the bonferroni-correction?
divide the significance level by number of tests -> use for each test
0.01/50,000 = p < .0000002