2004;75;6-12 J. Neurol. Neurosurg. Psychiatry P M Matthews and P Jezzard Functional magnetic resonance imaging http://jnnp.bmj.com/cgi/content/full/75/1/6Updated information and services can be found at: These include: References http://jnnp.bmj.com/cgi/content/full/75/1/6#otherarticles4 online articles that cite this article can be accessed at: http://jnnp.bmj.com/cgi/content/full/75/1/6#BIBLThis article cites 56 articles, 22 of which can be accessed free at: Rapid responses http://jnnp.bmj.com/cgi/eletter-submit/75/1/6You can respond to this article at: serviceEmail alertingtop right corner of the article Receive free email alerts when new articles cite this article - sign up in the box at theTopic collections (1501 articles) Other imaging techniques (3652 articles) Other Neurology Articles on similar topics can be found in the following collections Notes http://www.bmjjournals.com/cgi/reprintformTo order reprints of this article go to: http://www.bmjjournals.com/subscriptions/ go to: Journal of Neurology, Neurosurgery, and PsychiatryTo subscribe to on 22 January 2007 jnnp.bmj.comDownloaded fromNEUROSCIENCE FOR NEUROLOGISTSFunctional magnetic resonance imagingP M Matthews, P Jezzard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .J Neurol Neurosurg Psychiatry 2004;75:6–12Blood oxygenation level dependent (BOLD) functionalmagnetic resonance imaging (fMRI) is a powerfulapproach to defining activity in the healthy anddiseased human brain. BOLD fMRI detects local increasesin relative blood oxygenation that are most probably adirect consequence of neurotransmitter action andthus reflect local neuronal signalling. The methodallows localisation to volumes of the order of a few toseveral cubic millimetres and can be used in serialstudies of individual subjects. Basic approaches toexperimental design and analysis are reviewed briefly, aswell as potential clinical applications. The latter includethree broad areas: anatomical characterisation of normalor pathological patterns of brain functioning;distinguishing pathological traits; and monitoring treatmentresponses. New research is emphasising the integration offMRI with other techniques, particularlyelectrophysiological. In conjunction with MRI methodsfor characterising pathological load, fMRI promises arefined understanding of when disease processesbegin and how they can be modified by newtreatments.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .See end of article forauthors’ affiliations. . . . . . . . . . . . . . . . . . . . . . .Correspondence to:Professor P M Matthews,Centre for FunctionalMagnetic ResonanceImaging of the Brain, JohnRadcliffe Hospital,Headington, Oxford OX39DU, UK;[email protected] 14 May 2003In revised form14 August 2003Accepted 20 August 2003. . . . . . . . . . . . . . . . . . . . . . .Avariety of methods have been developedover the past few decades to allow map-ping of the functioning human brain. Twobasic classes of mapping technique have evolved:those that map (or localise) the underlyingelectrical activity of the brain; and those thatmap local physiological or metabolic conse-quences of altered brain electrical activity.Among the former are the non-invasive neuralelectromagnetic techniques of electroencephalo-graphy (EEG) and magnetoencephalography(MEG). These methods allow exquisite temporalresolution of neural processes (typically over a10–100 ms time scale), but suffer from poorspatial resolution (between 1 and several centi-metres). Functional MRI (fMRI) methods are inthe second category. They can be made sensitiveto the changes in regional blood perfusion, bloodvolume (for example, using injected magneticresonance contrast agents), or blood oxygenationthat accompany neuronal activity. Blood oxyge-nation level dependent (BOLD) fMRI, which issensitive primarily to the last of these variables,allows an image spatial resolution that is of theorder of a few millimetres, with a temporalresolution of a few seconds (limited by thehaemodynamic response itself). An accessibleand more detailed introduction to the techniquethan is possible in this brief review is found in arecent book.1PRINCIPLES OF FUNCTIONAL MRIContrast in a magnetic resonance imageThe contrast in a magnetic resonance image(which determines the apparent structure inwhat we see) depends on how it is acquired. Byadding radio frequency or gradient pulses, andby careful choice of their timings, it is possible tohighlight different characteristics of the tissuebeing imaged. While it is generally true that MRImaps the distribution of water in the brain, theuseful contrast in MR images comes not justfrom spatial variations in the density of waterbut also from differences in fundamental nuclearmagnetic processes known as relaxation, whichare characterised by distinct rates or ‘‘relaxationtimes.’’ There are three relaxation times that areof primary interest in MRI—T1, T2, and T2*.These describe the time constant for the return ofthe magnetisation to its equilibrium positionaligned along the static magnetic field of thescanner whenever it is disturbed (T1 relaxation)and the time constants associated with loss ofsignal once the magnetisation has been sampled(T2 and T2* relaxation). T2* is the most relevantrelaxation time for understanding contrast infMRI images.The physiological basis of BOLD fMRIMost of the energy used for neuronal activity isexpended as a result of the postsynaptic neuronaldepolarisation and, to a lesser extent, the actionpotentials generated.2The energy cost thereforearises from information transfer and its integra-tion postsynaptically. Substrate delivery forenergy metabolism is increased with increasedlocal blood flow. However, it is not the increasedenergy use itself that directly drives the increasein blood flow.3Instead, increased blood flowappears to be a direct consequence of neuro-transmitter action and thus reflects local signal-ling. Electrophysiologically, increases in theBOLD signal are correlated most clearly withthe local field potential rather than the neuronalfiring rate.4Blood flow in fact increases over awider volume and to a greater extent than isnecessary simply to provide oxygen