Neural activities lead to metabolic responses, such as increased blood flow and oxygen supply, which are detected and used in fMRI by the following methods:
Contrast fMRI: involves an injection of contrast agents such as iron oxide, coated with sugar or starch. The agents reach the brain and generate a signal that is proportional to the cerebral blood volume (CBV). This method provides relatively strong signals; however, due to the introduction of agents into the bloodstream, it is considered a semi-invasive method.
Blood-Oxygen-Level Dependent (BOLD) fMRI: this widely used technique detects differences in magnetic susceptibility between oxygenated hemoglobin (diamagnetic) and deoxygenated hemoglobin (paramagnetic). It is considered an indirect measure of neuronal activity because it measures changes in blood oxygenation rather than direct neuronal firing. Evidence strongly correlates BOLD signals with neuronal activity (Logothetis et al., 2001). The method still relies on the magnetic properties of hydrogen atoms in water molecules, which are influenced by the local magnetic environment altered by oxygenation levels.
Perfusion fMRI: this technique employs ‘arterial spin labeling’ (ASL) to magnetically label hydrogen nuclei in arterial blood and image their distribution in the brain. This method is sensitive to cerebral blood flow (CBF), which is considered a good correlate of neuronal activity and does not require any contrast agents. Compared to the BOLD responses, the signal in perfusion fMRI is more stable and the noise is whiter. However, it has a weaker signal and higher image acquisition time, so it’s less used.
A major challenge with the BOLD response is that hemodynamic responses are still relatively slow, weak, and noisy.
Following a single stimulus event, the response rises after 1-2 seconds, peaks at 4-6 seconds and ends at 12-16 seconds. To increase statistical power, many repetitions of the same stimuli must be made. For this reason, the event-related design was introduced. In this technique, long intervals (12-20s) were applied to avoid overlaps between the BOLD responses.
However, in research, slow event-related designs were considered boring for the subjects and therefore resulted in an inefficient data collection. Rapid event-related designs are now predominantly used, wherein different event types are pseudo-randomly mixed, maintaining intervals between stimuli.
The data Analysis is calculated through the general linear model (GLM), aiming to explain the variance in the BOLD time course (in terms of a linear combination of explanatory variables and an error term).
fMRI is primary used to map functional connectivity in the brain, this can be used to investigate brain networks like language, memory and emotion. In clinical settings, it is often used to study disorders like epilepsy, depression or schizophrenia. Additionally, it can evaluate the effects of drugs on brain activity in pharmacological studies.
The advantadges of fMRI include its high spatial resolution, providing millimeter-level detail, and also the capability to image the entire brain in a single scan. Another advantadge is its non-invasiveness, as there is no need for contrast agents nor radiation exposure.