Basics of fMRI

fMRI: Blood Oxygen Level-Dependent Signal

This note discussed the Blood Oxygen Level-Dependent (BOLD) signal, the basis of functional magnetic resonance imaging (fMRI).

Basic Principles:

Our brains have capillaries, arteries, and veins supplying oxygenated blood.
During rest, a steady ratio of oxygenated vs. deoxygenated hemoglobin exists.
Active brain regions require more oxygen for processing, leading to a local increase in deoxygenated hemoglobin.
Deoxygenated hemoglobin disrupts the magnetic field compared to oxygenated hemoglobin. fMRI Signal Measurement:
Radio waves nudge protons in the brain, causing them to precess (spin).
Relaxation time is measured - the time it takes for protons to return to alignment.
T2* relaxation reflects local magnetic field variations caused by oxygenation levels.
BOLD fMRI measures changes in T2* relaxation to indirectly assess brain activity.

BOLD Signal :

BOLD fMRI and Oxygen Depletion

Astrocytes absorb oxygen to replenish oxygen and glucose metabolism in firing cells.
Hemoglobin dephasing, caused by oxygen absorption, changes the MRI signal.
Deoxygenated blood causes more signal distortion than oxygenated blood.
BOLD signal allows us to infer brain activity by detecting changes in oxygenation.

The Hemodynamic Response Function

Brain activation initially reduces the MRI signal due to oxygen depletion.
Blood supply increases in response to activation, causing the BOLD signal to rise.
When the stimulus stops, oxygenation and the MRI signal overshoot and then undershoot before returning to baseline.
This entire cycle is called the hemodynamic response function.

BOLD fMRI Does Not Measure Neural Activity Directly

BOLD fMRI measures metabolic demands, specifically oxygen consumption of active neurons.
The hemodynamic response function reflects the change in the fMRI signal triggered by neural activity.

Physiological Basis of the BOLD Signal

Action potentials in presynaptic terminals cause neurotransmitter release.
Neurotransmitters bind to postsynaptic ion channels, allowing ions to flow into the cell.
Astrocytes reuptake glutamate and pump out ions to restore ionic gradients.
These processes require glucose and oxygen, hence the BOLD signal.

BOLD Signal and Correlation with Neuronal Activity

Previous Thought:

BOLD signal was believed to correlate with the number of action potentials.
Studies combined BOLD activation in monkeys with single-cell recording of action potentials.

Recent Findings:

Logothetis et al. (2001) conducted a more extensive study using BOLD signals and electrophysiological data.
They measured:
- Multi-unit activity (MUA) - activity from multiple neurons.
- Local field potentials (LFPs) - summation of postsynaptic potentials reflecting overall neuronal activity.
BOLD signal showed the strongest correlation with LFPs, not MUA or action potential count.
A follow-up study showed a close match between predicted BOLD signal based on LFPs and the actual BOLD response. Interpretation:
BOLD activity likely reflects synaptic input and information processing within a neural population.
BOLD is more closely related to LFPs than action potentials or MUA.

Limitations:

Summary:

BOLD signal reflects changes in blood oxygenation due to neural activity.
LFPs provide a better representation of local neuronal activity than BOLD signal.
BOLD signal can be used to estimate local field potentials and neuronal activity indirectly.
fMRI utilizes 3D BOLD signal measurements to create activation maps of brain function.

Summary of BOLD fMRI

Neural activation removes oxygen from blood to support local cognitive processing.
This changes the magnetic properties of the blood.
Influx of oxygenated blood further changes magnetic properties.
BOLD signal reflects these magnetic property changes, which correlate with local field potentials and neuronal activity.
By measuring BOLD signal in 3D, we can generate activation maps of brain activity.