Functional Magnetic Resonance Imaging (or fMRI) describes the use of MRI to measure hemodynamic signals related to neural activity in the brain or spinal cord of humans or other animals. It is one of the most recently developed forms of brain imaging

It has been known for over 100 years (see Sherrington ) that hemodynamic activity is closely linked to neural activity. When nerve cells are active, they consume oxygen supplied by local capillaries. Approximately 4-6 seconds after a burst of neural activity, a haemodynamic response occurs and that region of the brain is infused with oxygen-rich blood.

It turns out that oxygenated haemoglobin is diamagnetic, while deoxygenated blood is paramagnetic, so give off slightly different MR signals. An MR scanner can be used to detect this signal difference, which is known as BOLD contrast. The precise relationship between neural signals and BOLD is under active research using fMRI in monkeys with simultaneous electrical recording of neurons. So far, it appears that BOLD is well correlated with both local field potentials (caused by electrical activity in dendrites) and with action potentials (spiking), but the correlation with LFPs is slightly better (see Logothetis 2001 and 2003). This suggests that the BOLD signal is more of an indicator of input and neuronal processing (which occurs in dendrites) than the output activity of an area (which is conveyed by spiking).

BOLD effects are measured using a T2 related imaging process (actually T2*), which is different from the T1 scan taken in ordinary structural MRI images (the former measures the rate of change of spin phases, while the later detects the half-life of inverted spins). T2* images can be acquired with moderately good spatial and temporal resolution; scans are usually repeated every 2-5 seconds, and the voxels in the resulting image represent cubes of tissue approximately 3 millimeters on each side. Other non-invasive functional medical imaging techniques can improve on one of these figures, but not both.

The science of applying fMRI is quite complicated and multi-disciplinary. It involves:

• A good understanding of the physics of MRI scanners.
• Statistical analysis of results. Because the signals are very subtle, correct application of statistics is essential to both "tease out" observations and avoid false-positive results.
• Psychological study design. When conducting fMRI on humans, for example, it is essential to employ carefully designed experiments which allow the precise neural effect under consideration to be separated.
• For a non-invasive scan, MRI has moderately good spatial resolution, but relatively poor temporal resolution. Increasingly, it is being combined with other data collection techniques such as EEG or MEG, which have much higher recording frequencies.
• Integration with other areas of neuroscience in order to better understand the location (and role) of the signals which fMRI is able to detect. This includes a great deal of neuroanatomy but also other sub-fields such as neurochemistry and neuropathology.

Aside from BOLD fMRI there are other ways to probe the brain activity using MRI:

• By using a injected contrast agent, e.g., MION, causing a local disturbance in the magnetic field that is measurable by the MRI scanner. The signal associated with these kind of contrast agents are proportional to the cerebral blood volume . Other methods of investigating blood volume which do not require an injection are a subject of current research.

• By using what is called arterial spin labelling ASL. The associated signal is proportional to the cerebral blood flow , or perfusion.

Magnetic resonance spectroscopic imaging (MRS) is another, NMR-based process for assessing function within the living brain. MRS takes advantage of the fact that protons (H) residing in differing chemical environments depending upon the molecule they inhabit (H₂O vs. protein, for example) possess slightly different resonant properties. For a given volume of brain (typically > 1 cubic cm), the distribution of these H resonances can be displayed as a spectrum. The area under the peak for each resonance provides a quantitative measure of the relative abundance of that compound. The largest peak is composed of H₂O. However, there are also discernable peaks for choline, creatine, n-acetylaspartate (NAA) and lactate. Fortuitously, NAA is mostly inactive compound within the neuron, serving as a precursor to glutamate and as storage for acetyl groups (to be used in fatty acid synthesis) -- but its relative levels are a reasonable approximation of neuronal integrity and functional status. Brain diseases (schizophrenia, strokes, certain tumors, multiple sclerosis) can be characterized by the regional alteration in NAA levels when compared to healthy subjects. Creatine is used a relative control value since its levels remain fairly constant, while choline and lactate levels have been used to evaluate brain tumors.

diffusion tensor imaging (DTI) is a related use of MR to measure anatomical connectivity between areas. Although it is not strictly a functional imaging technique, because it does not measure dynamic changes in brain function, the measures of inter-area connectivity it provides are complementary to images of cortical function provided by BOLD fMRI. As protons are directed along certain axes in the brain (for example, as water flowing down a neuronal axon within a bundle of nerve fibers in cerebral white matter), this directionality can be measured. Connectivity between brain regions may be inferable from diffusion images, and illnesses that disrupt the normal organization or integrity of cerebral white matter (such as multiple sclerosis) have a quantitative impact on DTI measures.

Scanning in Practice

Subjects in a fMRI are asked to lie still, and usually restrained with soft pads to prevent small motions from disturbing measurements. It is possible to correct for some amount of motion with postprocessing of the data, but significant motion can easily render these attempts futile.

Is fMRI worthwhile?

Ever since its inception, fMRI has been critised for only asking "where" brain activity occurs. Some authors, such as Uttal, go so far as to suggest that fMRI is just a modern-day phrenology and is therefore destined to fail and fundamentally uninformative. Other functional imagers offer counter-arguments (e.g. Donaldson 2004, Henson 2005).

See also

• statistical parametric mapping