fMRI Method Flashcards
The advantage of fMRI compared to PET
PET involves administering a radioactive isotope to the patient, thereby exposing the patient to a significant amount ionizing radiation
fMRI (originally called NMRI - Nuclear Magnetic Resonance Imaging) does not involve radiation
Only harm is metal
What is the function of head coil in fMRI
Send radio frequency pulses and also function as a receiver
The premise of MRI
More than 70% of the human brain consists of water Hydrogen atoms (H+ protons) can be thought as small bar magnets, “precessing” like a spinning top about an axis
How to align the random spin directions of protons
Externally applied very strong magnetic field in the MRI scanner
They are not perfectly aligned – and they are also not static, but they still keep “precessing” in a random fashion
The precession frequency of protons depends on the strength of the magnetic field
What is Z-axis?
The axis along which the magnetization is build up in the scanner;
The magnetization vector
How to measure the Z-axis
We need to tilt the magnetization vector to measure it as a signal
A radio frequency (RF) pulse is applied perpendicular to magnetic field; Its frequency is matching the precession frequency of the protons
The RF pulse causes the protons to absorb energy, which tilts the magnetization vector to the transversal plane and aligns the precession of the the spins
What is relaxation and its effect
Switching off the RF pulse and the transversal magnetization decays
The effect is that the longitudinal magnetization is re-established independently; the protons emit the excess energy and lose phase coherence very quickly
What does structural brain image depend on
The summed effect of many protons recorded during relaxation
Why the transversal magnetization decays with different speeds
depending on the tissue, the differences in the density of protons:
Protons lose coherence because they will be influenced by other protons in their environment and begin to cancel each other out
How to reconstruct the source of the measured signal
By using gradients
Because protons will absorb energy from RF pulses only when the frequency of the RF pulse matches the proton’s precession (resonance) frequency, by causing the magnetic field to vary linearly, we can cause the resonance frequency to vary throughout the brain
The first step of reconstruct brain images
Slice selecting gradient
Divide the brain into “slices”: vary the gradient field along the z-axis and know that different slices were exposed to different field strengths
By exciting one slice at a time, we get the z-coordinate of all resulting signals
The second step of reconstruct brain images
Frequency encoding gradient
Use a second gradient to change the magnetic field within this slice: vary the gradient along the y-axis
We get the x-coordinates of the measured signal
The third step of reconstruct brain images
Phase encoding gradient
Briefly using a gradient along the y-axis causes protons to “speed up” their precession according to the strength of the magnetic field.
When switching off this gradient, all protons are back to the same precessing frequency (i.e. “locked” into precessing at the same speed) – but they are “out of phase” with each other
We get the y-coordinate of the resulting signal
The fourth step of reconstruct brain images
Use a technique called Fourier transformation to reconstruct the entire space
We can measure slices in ascending order, descending order, or interleaved, until we have a full 3D image of the brain
Usually, measuring one full 3D image of the brain takes 1-3 seconds
The steps of MRI
Externally applied very strong magnetic field in the MRI scanner (Z axis)
The RF pulse causes the protons to absorb energy, which tilts the magnetization vector to the transversal plane
Switching off the RF pulse and the transversal magnetization decays
By using gradients on the:
Z-axis (Slice selecting gradient)
X-axis (Frequency encoding gradient)
Y-axis (Phase encoding gradient)
Use Fourier transformation to reconstruct the entire space
What does MRI image
the structure of the brain
What does fMRI image
the functioning brain
Difference between oxygenated and deoxygenated blood
Oxygenated blood, oxyhaemoglobin (Hb), is diamagnetic, enhancing the signal;
Deoxyhaemoglobin (dHb) is paramagnetic, decreasing the signal
What does BOLD fMRI make use of
The fact that all neurons need oxygen, supplied from the blood
Neural activity is accompanied by a local increase in blood oxygenation, which is needed for glucose metabolism; also a local oversupply in oxygenated blood, therefore a better BOLD signal
How BOLD signals are analysed
Statistical Parametric Mapping:
General Linear Model is fitted to brain activity at each measurement point (voxel)
How does a typical fMRI experiment work
BOLD signal within a region is measured while participants engage in a cognitive task
Repeated measurement of brain activity is required for the whole brain while performing experimental tasks (e.g., A vs. B)
Significantly stronger activation in region X for task A compared to task B is interpreted as involvement of the region in task A
This is a Reverse inference: drawing conclusions about cognitive processes from the presence of activation
Why we have to be careful when interpreting differences in BOLD signal
Blood Oxygen-Level Dependent (BOLD) signal indirectly measure neural activity
First, enhanced neural activity bring more oxygen in blood but also extracts oxygen more slowly
Second, there is also a substantial temporal lag between neural activity and the peak of the BOLD response – in the order of 8 seconds. The BOLD signal further needs ~16 seconds before reaching baseline again
Third, it is also not valid to compare signals between different regions of the brain because the signal change is different
What is Heamodynamic Response Function (HRF)
The measured response looks very similar across regions.
What neural processes drive the BOLD signal
Logothetis (2008) describes local cortical excitation-inhibition networks (EIN), small and highly interconnected functional microunits, which show massive recurrent feedback
Feedback processing within these EINs (net excitation or inhibition) affect the regional metabolic energy demands, measured by BOLD fMRI.
Limitations of BOLD fMRI
Brain is not necessary modular, fMRI might not always map the functional units that matter
fMRI has a poor temporal resolution
The spatial resolution of fMRI is good but not good enough, a voxel contains many neurons
The multiple comparisons problem, we run a t-test for every voxel, which leads to too much false positives.
We need to do a statistical correction (Bonferroni-correction)