What is the Difference between MRI and Fmri

Is MRI a Functional or Structural imaging technique?

MRI is a versatile imaging method and can be used for imaging both complex structures or it can be used for functional imaging; this is referred to as functional MRI (fMRI). The main difference between MRI and fMRI is whereas structural MRI imaging shows the difference between types of tissues at high resolution with respect to space, fMRI shows the difference between the tissues with respect to time[1].

Structural MRI is used to view anatomical structures, most commonly neurological and other soft tissue structures, MRI takes advantage of the different factors within the body such as the presence of different chemical bonds, paramagnetic ions and rate of flow of fluids that produce different Magnetic Resonance (MR) signals and hence show contrast between tissues not possible with other methods of imaging such as Computed Tomography (CT)[2]. The structural contrast can be shown in four main ways with respect to the relaxation constants, T1 or T2, proton density or diffusion weighted images:

The relaxation constants are defined by processes which cause the MR signal to decay:

T1 relates to the time constant of the ‘spin-lattice relaxation’ this refers to the interaction of the protons spin with its surroundings into which energy is released as protons return to a lower state of alignment.

T2 relates to the time constant of ‘spin-spin relaxation’ this is caused by the loss of precession synchrony among the protons; prior to the radio frequency pulse that aligns the protons, the protons precession is random and out of phase, the radio frequency pulse aligns the protons and brings them into phase however when the pulse has passed, the protons revert to random states due to interactions with neighbouring protons whom they swap energy through random collisions.

There is also a T2* decay like T2 it is the dephasing of protons but due to inhomogeneity of the static magnetic field[3].  The relaxation (or decay) time is always shorter for T2 than T1, though T1 tends to be longer in solids and shorter in liquids whereas T2 tend to be longer in liquids and shorter in solids.[4] Different contrasts can be obtained by weighting the images; this is done by adjusting the pulse parameters to reduce one type of relaxation, in a typical MRI scan, an initial pulse is sent out to phase the protons to an angle, and then a second ‘rephasing pulse’ is applied at 90° to the original pulse to eliminate field inhomogeneity, pulse sequences can be selected to reduce one relaxation parameter and ‘weight’ the image toward the other, for example to create a T1 weighted image the time between pulses would be equal or less than the specific T1 time for that tissue, and echo time less than the tissue specific T2 time (the echo time is a time at which the original pulse echoes)[5], conversely a T2 weighted image would be the opposite to that. Both types of weighted images have different advantages for the types of tissue being imaged, for example T2 weighted images show water to be very bright since the T2 time for liquids is longer whereas water would appear dark in a T1 weighted image.

Proton density contrast image is an image created by minimising the T1 and T2 factors which leaves an image dependant on the density of the protons, within the image the greater the density of protons the larger the magnetisation and hence the brighter it appears on the image.

Diffusion weighted images are sensitive to the diffusion of water within the area, in this type of imaging the protons are pulsed with a gradient, the protons are then pulsed again, at 90° to rephrase them, however protons that have moved within this process will be at a different part of the gradient and hence will not be rephrased completely from the difference left by incomplete rephasing on some of the protons an image can be constructed[6].

These types of contrast are the four main types of structural MRI imaging and can be used to provide a non-invasive insight into many anatomical areas. MRI is advantageous over CT for many reasons, CT uses ionising radiation, which increases patients lifetime cancer risk and other effects are relatively unknown, whereas MRI has no known health risk, although there is some debate over whether small induced currents by magnetic fields (such as those used in MRI or power lines) can act as a carcinogenic, there is no evidence this occurs in the case of MRI.[7] MRI also has a much greater contrast difference than that of CT and uses smaller amounts  and safer contrast[8]. The main disadvantage of MRI is the cost and the time of which it takes to image a patient, CT is far quicker and is hence used in emergency situations.

Functional MRI is used to show the difference between tissues at different times, currently fMRI is limited to neurological use although it is being researched to looking at cardiac function, currently modern day MRI is not advanced enough to provide imaging of cardiac function, there are a number of technological obstacles to be overcome before it can be achieved, most notably a frame rate of around real time at a fairly high resolution.[9] Neurological fMRI work on the principle that oxygenated blood is diamagnetic, however deoxygenated blood is paramagnetic, this is because although both oxygenated and deoxygenated blood contains iron, within the oxygenated blood (oxyhemoglobin) the outer shell iron electrons are transferred to the oxygen molecules, whereas in deoxygenated blood (deoxyhemoglobin) four of the irons 6 electrons are unpaired, meaning the iron is in a ferrous Fe2+ state which is paramagnetic. It is known that neurological activity causes localised blood flow, volume and oxygen delivery and hence increases the oxyhemoglobin and decreases the deoxyhemoglobin at certain areas within the brain, fMRI can distinguish between the magnetic resonance of the two which affects the T2 and T2* relaxation rates and hence map the level of blood flow[10]. Typically a high spatial resolution MRI image(T1 weighted) is used to map the changes in oxygenated and deoxygenated blood ratio which is measured by a series of low spatial resolution fMRI scans(T2 weighted), of a relatively high temporal resolution in the order of one image every 2-3 seconds[12], although scans can be adjusted to incorporated higher temporal resolutions at the cost of a lower spatial resolution, for example a spatial resolution of 3mm may allow a temporal resolution as high as 50ms[13].

fMRI has advantages of other methods of neurological activity as it does not require radioactive tracers to be injected, it has a relatively short scan time around two minutes (dependant on the study) and it has a relatively high spatial resolution of around 1.5mm but can be better than 1mm.[14]

A good example of a study which can utilise both structural and functional aspects of MRI is that conducted on the difference between London taxi drivers and London bus drivers. Within this study MRI and fMRI were used to study the hippocampus a part of the brain which is linked with forming new memories and retrieving old ones [15] and is also involved with spatial navigation[16]. Firstly MRI was used to measure the grey matter volume within the hippocampus, fMRI was then used to measure neurological activity within the hippocampus and determine the response and test the ability to acquire new visuo-spatial information. The MRI led the study to determine that taxi drivers had greater grey matter volume in the mid-posterior hippocampi and less in the anterior hippocampi in comparison to bus drivers, and that greater navigational experience as a taxi driver led to greater posterior gray matter, and less anterior. The fMRI showed  that taxi drivers could not acquire new visuo-spatial informational as well as bus drivers, leading to the conclusion that an expert navigational ability and knowledge can be associated with a larger volume of posterior hippocampal gray matter but makes it more difficult to retain new knowledge and reduces volume of anterior hippocampal gray matter[17]. Although the results are somewhat irrelevant, the study shows how MRI and fMRI and both important tools that can be used together to provide a better understanding of a situation.

To conclude, MRI is both a structural and functional imaging technique, it can provide a range of differently contrasted high spatial resolution images, or alternatively it can provide a series of low spatial resolution at a relatively high temporal resolution to map neurological functionality due to the levels of oxygen within the blood.


http://epl.meei.harvard.edu/~keh/cd846/Lecture12.pdf p3.

[2] William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p368

[3] William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p361

[4] William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p364

[5] http://www.mr-tip.com/serv1.php?type=db1&dbs=T1%20Weighted%20Image

[6] http://spinwarp.ucsd.edu/NeuroWeb/Text/br-710dwi.htm

[7] William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p397

[8] http://www.medscape.com/viewarticle/488088_5

[9] http://ibd.nrc-cnrc.gc.ca/research/m_r_tech/3_cardiac_mri_e.html.

[10]William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p377

http://www.fmrib.ox.ac.uk/fslcourse/lectures/intro1.pdf p10-11

[12]William R. Hendee, E. Russel Ritenour, Medical Imaging Physics, 4th edition, Wiley-Liss, 2002, p377

[13] Mark F. Bear, Barry W. Connors, Michael A. Paradiso, Neuroscience: Exploring the Brain 3rd edition , Wolters Klewer Health, 2007, p178

[14] http://fmri.org/fmri.htm

[15] New York University (2004, May 13). Scientists Show Hippocampus’s Role In Long Term Memory. ScienceDaily. Retrieved March 24, 2008, from http://www.sciencedaily.com­ /releases/2004/05/040513010413.htm

[16] http://www.nsi.edu/index.php?page=iii_brain_regions

[17] http://www.fil.ion.ucl.ac.uk/Maguire/Maguire2006.pdf