BOLD Flashcards
Name an example of early research on mental activity and blood flow.
Angelo Mosso’s experiments: Participants lied down on a scale and performed mental tasks. When tasks where performed, the scale tipped towards their heads, implicating blood flow to the brain.
Name two reasons, why it is possible to infer neural activity from fMRI, as stated in the lecture.
- fMRI utilizes magnetic properties of hemoglobin
- Neurovascular coupling occurs
Name three different relaxations effects found in MRI, why they occur and what they are mostly used for.
T1 -> slow recovery of longitudinal magnetization -> structural imaging
T2 -> dephasing of spins due to spin-spin interactions -> imaging of “soft structures”
T2* -> same as T2 but magnetic field inhomogenities are also reason for dephasing -> measurement of BOLD signal
What two types of hemoglobin are discussed in the lecture? What are their magnetic properties?
Oxygenated Hemoglobin -> diamagnetic (weaker magnetic properties)
Deoxygenated Hemoglobin -> paramagnetic (stronger magnetic properties)
How are the T1 and T2* relaxation rate modulated through the presence of deoxy-hemoglobin?
Basing your explanation off an experiment where test-tubes were used can help!
T1 relaxation is basically not affected by the concentration of oxygen in the blood. Varying the oxygenation does not modulate the T1 relaxation rate.
T2* relaxation rates have a linear relationship with the oxygenation rate of blood. Deoxygenated Hb results in much faster relaxation, the more oxygen is present the slower the relaxation rate.
Why is that so? Deoxygenated Hb is paramagnetic. The presence of Deoxygenated Hb alters the local magnetic field and produces inhomogeneities. The less oxygen is present, the faster the dephasing of the transverse magnetization occurs. This again leads to a faster decay of the signal and a less intense (darker or weaker) signal.
How do rodent brains compare visually if one is fully oxygen saturated and the other receives a regular dose of 65% oxygen?
At full saturation no vessels can be observed, as strongly oxygenated blood appears hyperintense on MRI and cannot be distinguished from the background.
“Less oxygen -> shorter T2* -> darker on image”
Bonus question: How would these differences be expressed on a T1 image?
How does a SpinEcho T2 image (a sequence that yields “pure” transversal dephasing w/o field inhomogenties) compare to a T2* image when observing oxy and deoxy-Hb in a testtube?
In addition to describing the findings, feel free to explain how they come about!
For oyxygenated blood, the T2 and T2* sequences produce clear images of hyperintensity (looks light greyish.)
For deoxygneated blood T2 also appears rather clear cut, T2* however produces a “phased out, unclear” image.
Why?
T2* takes inhomogeneties of the local magnetic field into account. As paramagnetic deoxy Hb produces inhomogenites, it presents in the way described above. In BOLD measurements, this artifacts - usually considered undesirable - can be helpful for location deoxy-Hb!
Why does T2* pick up magnetic field inhomogenities and how is this applicable to BOLD?
Local magnetic field inhomogenties arise, as the magnetic deoxygenated Hb in the blood distorts the field localy. These local inhomogeneties are responsible for additional phase changes, that occur due to the differences in resonance frequencies experienced by protons at different locations. (remeber the frequency is B0 dependent!) These additional phase changes contribute to the faster decay of the MRI signal, leading to shorter T2* relaxation time.
This effect of faster decay can be picked up in the BOLD signal!
If you do not understand this, just listen to this banger of a song by Greg Crowther from the University of Washington : Twinkle Twinkle T2 * https://www.youtube.com/watch?v=uu7Ph25EhLQ
How is the BOLD signal generated?
Neurovascular coupling: neural metabolism up -> cerebral blood flow up -> deoxygenated blood is displaced -> deoxy-Hb down -> hydrogen spin relaxation slower, T2* up
What is an HRF? Describe its dynamics over time
HRF is a Hemodynamic Response Function. It measures the changes in BOLD signal intensity
HRF:
1. baseline — no stimulation, baseline firing (occasional spikes)
2. stimulus — initial dip, neurons fire (0-2s, local oxygen consumption, less oxygen in the area)
3. rise (overshoot) — stimulus switched off, no firing (rise, peak at 4-6s, decay, significantly more oxygen, still some deoxygenated blood)
4. decay
5. post-stimulus undershoot (10-20s, oxygen level is below the baseline)
6. Back to baseline
=> total length 20-30s
Give a rough estimate of how many:
* neurons
* synapses
* dendrites
* axons
* capillaries
1mm^3 of cortex contains
- 10^4-10^5 Neurons
- 10^8-10^9-Synapses
- 300 m Of dendrites
- 4000 m Of axons
- 0.4 m of capillaries
NB! 3x3x3 mm = 27 mm^3 — what fMRI measures — really far from measuring a single neuron activity
Is the summation of BOLD responses a problem?
Trials sum to each other and the last trial have more signal. But we can deduct previous signals from the later one, so there is no violating of linearity and the summation is not actually a problem
What is a negative BOLD response?
A negative BOLD response (NBR) was observed beyond the stimulated regions of visual cortex, associated with local decreases in neuronal activity below spontaneous activity
Is BOLD data in line with the electrophysiological data?
Yes, it is
* fMRI responses in human V1 and MT are proportional to average firing rates in monkey V1 and MT
* spikes vs. BOLD in humans: BOLD is a good prediction of human brain activity
What is neurovascular coupling?
It is a coupling between neuronal activity and cerebral blood flow.
It is a fundamental mechanism that ensures the brain receives an adequate supply of oxygen and nutrients to support its increased metabolic demands during neural activity.
Multi-Unit Activity (reflects action potentials/spiking) and especially Local Field Potentials (summation of post-synaptic potentials) are good predictors of NVC
