Advanced MRI techniques, including diffusion weighted imaging (DWI) and perfusion weighted imaging (PMRI), MR spectroscopy (MRS), functional MRI (fMRI), and magnetoencephalography (MEG), are extending uses of MRI from a purely diagnostic tool that previously answered the question, “Is a disease present?” to a physiologic examination capable of monitoring water movement in the brain, blood flow, changes in chemical composition, and changes in neuronal activity. With these new techniques, imaging is useful in predicting patient outcomes and monitoring response to therapy.

For example, a combination of diffusion and perfusion imaging is used to detect potentially salvageable ischemic brain that can be saved with intervention to improve circulation locally or generally. Detection of this so-called “perfusion-diffusion mismatch” is now a vital part of the algorithm used to select patients for thrombolytic therapy.

This article reviews these MR techniques and discusses their leading role in patients with neurological diseases ranging from primary brain neoplasms to acute stroke.


Diffusion . With the application of the appropriate magnetic field gradients, MRI may be sensitized to the random, thermally driven motion (diffusion) of water molecules in the direction of the field gradient. This diffusion-weighted imaging technique is exquisitely sensitive to the random motion of water molecules in the brain parenchyma and is capable of determining the presence of infarction within minutes to a few hours after symptom onset.1 Infarction of cerebral tissue is characterized by increased intracellular water that diffuses less rapidly, resulting in greater signal intensity in the DWI. The characteristic “light-bulb bright” intensity in a DWI indicates acute infarction.

Diffusion-weighted MR images are most commonly obtained using single-shot echo-planar imaging (EPI) techniques, although experience with multi-shot fast spin-echo (FSE) sequences is gaining and appears to decrease the B0-related artifacts often seen with EPI-based DWI.2 The PROPELLER technique (periodically rotated overlapping parallel lines with enhanced reconstruction) developed by Pipe at the Barrow Neurologic Institute allows a multishot fast spin-echo sequence to be used for DWI.3 Data are acquired in a series of rotating “blades” in k space, allowing one to compare the blades for phase inconsistencies and correct for them. Because a FSE sequence can be used, the B0-related artifacts so notorious with EPI-based DWI are reduced and image quality is improved. Visualization of brain parenchyma in regions of local field inhomogeneity, such as along metallic clips, air-brain interfaces, and mixed blood products, is improved. At the University of Wisconsin, this technique is used to obtain DW images of the cervical cord and we hope to use it to obtain physiologic information in patients with demyelinating disease of the cord.4

DWI has other applications. It is effective at discriminating dermoid or epidermoid tumors from arachnoid cysts. These entities appear similar on T1- and T2-weighted images but are strikingly different on DWI. While the free-flowing fluid within the arachnoid cyst is isointense to cerebrospinal fluid and hypointense to brain, dermoids and epidermoids are hyperintense. DWI is also helpful in the early diagnosis of Creutzfeldt-Jakob disease, showing progressively hyperintense changes in the striata and cerebral cortices.5 In imaging of tumors, highly cellular tumors (which restrict the free flow of water molecules) can be differentiated from less cellular tumors. It is also useful in differentiating intracerebral abscesses, which restrict diffusion, from cavitating tumors, in which restricted diffusion is rarely seen.

Figure 1. The penumbra: region of decreased perfusion seen as red on the mean transit time map with corresponding increase in cerebral blood volume depicted in blue. This perfusion abnormality is larger than the infarct territory seen on diffusion-weighted imaging in this patient with left middle cerebral artery infarct and is the target for intervention.

Perfusion . Perfusion is the cerebral blood flow arriving at a particular brain region at a given moment in time. A time series of multi-phase echo-planar, T2*- or T2-weighted images obtained during the administration of contrast medium may be used to derive relative cerebral blood flow (CBF), mean transit time (MTT), and relative cerebral blood volume (CBV).6-8

In cases of hyperacute stroke, these perfusion images appear accurate for prediction of areas likely to undergo infarction.9 Areas with diminished perfusion that have not developed restricted diffusion represent regions that may be rescued if perfusion is restored (penumbra). Due to autoregulatory mechanisms, areas of ischemic, but not yet infarcted, brain may demonstrate increased CBV relative to CBF. The demonstration of the penumbra provides useful prognostic information about the amount of brain that may proceed to infarction if left untreated. This technique is now an integral part of the workup of stroke victims and helps determine who receives thrombolytic therapy (see Figure 1).

Perfusion imaging may play a supplementary role in the evaluation of dementia, neoplasms, trauma, epilepsy, migraines, and vasospasm.10-11 Early posttraumatic cerebral hypoperfusion is implicated in the evolution of secondary damage after traumatic brain injury and the quantification of such reductions may help predict long-term outcome.12 Perfusion imaging may be helpful in the preoperative assessment of tumor grade, with one series showing increased permeability values in high-grade glial neoplasms compared to the low-grade tumors.13 In the investigation of Alzheimer’s, the capability to assess regional blood flow changes may be useful. The role of PWI in the assessment of dementia is under investigation.


An important application of diffusion imaging is diffusion tensor imaging (DTI). Because axonal membranes and myelin sheaths present barriers to the motion of water molecules in directions not parallel to their own orientation, diffusion is anisotropic, or directionally dependent, in white matter (WM) fiber tracts. The direction of maximum diffusivity within WM tracts coincides with WM fiber-tract orientation.14 This information is contained in the diffusion tensor, a mathematical model of diffusion in 3D space.15 With DTI, both the degree of anisotropy and the direction in which water preferentially diffuses can be mapped voxel by voxel, providing a new and unique opportunity for studying WM architecture in vivo.

While many measures of anisotropy have been used, the most common anisotropic metric, fractional anisotropy (FA),16 derives from the standard deviation of the relative amount of diffusion in three perpendicular axes. The direction of maximum diffusion may be mapped using red, green, and blue color channels, in which the intensity represents degree of anisotropy and the color itself, direction (see Figure 2, page 109). DTI maps have been studied as a supplemental imaging modality in tumors, trauma, multiple sclerosis, and many other diseases. For surgical planning, the DTI map provides a means to determine the integrity of white matter tracts near a tumor and the approach that will result in the least damage to these tracts. DWI is being evaluated for the investigation of patients with epilepsy and patients with various neuropsychiatric disorders and in cerebral trauma.

DTI and tumors. At this year’s American Society if Neuroradiology (ASNR) meeting in Washington, DC, Jellison et al from the University of Wisconsin described a new diffusion tensor metric that may increase the specificity of MRI in discriminating vasogenic edema from infiltrating neoplasm in cerebral white matter tracts. We analyzed two patients with grade II oligodendrogliomas that infiltrated white matter tracts and compared them to four cases of vasogenic edema in patients with metastatic lesions. We found that while the more traditional measures of tensor shape varied little between the two pathologies, measures of regional tensor organization discriminated between the two.17 This promising application of DTI, distinguishing infiltrated versus edematous white matter tracts, may allow the radiologist to better direct the surgeon’s intervention and allow him to spare important white matter tracts.

DTI and trauma. Disruption of the cytoskeletal network and axonal membranes in cerebral white matter characterizes diffuse axonal injury (DAI) in the first few hours after traumatic brain injury. Histologic abnormalities seen in DAI hypothetically decrease the diffusion along axons and increase the diffusion in directions perpendicular to them. Investigators at the University of Wisconsin recently evaluated patients with mild traumatic brain injuries and found significant reduction of diffusion anisotropy in several regions compared with the homologous, normal contralateral hemisphere, suggesting that DAI might be detected in patients with mild traumatic brain injury through a reduction of the diffusion anisotropy. This detection, at the very early stages in patients with mild traumatic brain injury, could have significant implications for the diagnosis and treatment of trauma patients.18


Despite years of experience and the widespread use of conventional MR imaging in the evaluation of gliomas, the nonspecific nature of their enhancement still produces difficulty in defining the extent of these tumors’ involvement (for example, both treatment-induced necrosis and recurrent tumor can enhance).

Single-voxel proton MR spectroscopy has been used in the evaluation of brain tumors for many years. It requires placement of a volume of interest in the affected brain and the separate application of phase encoding to obtain two- and three-dimensional arrays of voxels. At specific frequencies, resonant peaks are identified from the presence of specific metabolites. For MRS in the brain, choline, creatine, N-acetylaspartate (NAA), lactate, and lipids are the most useful peaks to evaluate. The choline peak reflects membrane synthesis and turnover. The creatine peak reflects the energy requirements at the cellular level. NAA is a marker for adult healthy neurons. Lactate is specific for anaerobic metabolism. It is not found in normal brain, but is in regions of ischemia or mitochondrial dysfunction. The lipid peak is also not seen in normal brain but is seen in regions of necrosis.19 It has also been detected in the frontal lobes of patients with bipolar disorder, suggesting that MRS may be useful in assessing some neuropsychiatric disorders.

In the normal brain, the NAA peak is about twice the height of the choline or creatine peaks. In tumors, the choline peak is elevated with respect to creatine or NAA (see Figure 3, page 110). Regions of necrosis show spectra with increased lactate and lipids. In infarction, MRS shows acutely the presence of lactate and, later, lipid.

Figure 2. T2-weighted image, apparent diffusion coefficient map, fractional anisotropy map, and directional color map in a patient with a pilocytic astrocytoma. Note the deviation of the adjacent white matter tracts exquisitely depicted with color diffusion tensor imaging.

The use of MR spectroscopy has increased dramatically over the last decade. Many neuroradiologists routinely use MR spectroscopy and consider it a valuable diagnostic tool, capable of discriminating between tumor and necrosis in cases where the anatomic evaluation on routine MR imaging is not diagnostic. MRS may have a role in cerebral trauma to predict the brain’s potential for recovery. MRS has proven utility in AIDS patients, where it can distinguish infections from tumors. Its role in dementia, ischemia, hepatic encephalopathy, and mitochondrial disease is currently under investigation.20-22


Some radiologists use the term “functional MRI” in reference to any physiologic MR imaging study of the brain, distinct from conventional, anatomic MR imaging. This includes diffusion-weighted imaging, perfusion imaging, and MR spectroscopy. The term functional MRI, abbreviated as fMRI, refers specifically to the use of fast MRI to detect regional changes in the signal intensity in the brain due to the activity of neurons in that region. fMRI is a noninvasive means of mapping eloquent brain regions, such as the motor cortex, visual cortex, or language areas. Images obtained with fMRI are based on the increase in blood flow that accompanies the neural activity of specific brain regions. Because the increase in blood flow supplies a greater quantity of oxygenated hemoglobin than the neurons require, the ratio of oxyhemoglobin to deoxyhemoglobin increases. This relative decrease in deoxyhemoglobin causes less MR signal loss (a loss that deoxyhemoglobin normally causes secondary to spin dephasing), and produces an increase in MR signal. Cortical functions can therefore be imaged without exogenous contrast administration. This is the basis for blood oxygenation level-dependent (BOLD) contrast and makes up the majority of fMRI techniques. Patients are asked to perform specific tasks or are presented with images or sounds to attempt to activate particular brain regions. Sophisticated statistical analyses correlate the signal time curve obtained with BOLD to the task paradigm and allow identification of “active” areas of the brain (see Figure 4, page 111).

fMRI’s role in presurgical planning has been verified. It can estimate the postoperative risk of deficits, determine the need for intraoperative mapping, and help triage a patient for nonoperative therapy such as radiation or chemotherapy. However, its use is not limited to presurgical planning. Researchers have used fMRI to study how the brain develops, how humans learn, and why gamblers gamble.23,24 A noninvasive test, fMRI has the potential to replace the WADA test that requires carotid artery catheterization to localize language and memory function in patients selected to undergo resection of an epileptic focus or lesion. It may also be an excellent tool to help guide the rehabilitation of specific, unaffected brain regions in patients with recent stroke, facilitating their rehabilitation from this devastating illness.

Details of fMRI’s accuracy and reliability for specific paradigms are being elucidated. At the University of Wisconsin, investigators recently analyzed the reliability, test-retest precision, and accuracy of standard word-generation paradigms for fMR imaging and found it capable of producing accurate and reliable maps of Broca’s area.25 More studies will help determine the precision and accuracy of this powerful technique.26

fMRI is increasingly being applied to the study of neuropsychiatric disorders such as traumatic brain injury, Alzheimer’s disease, Huntington’s disease, and schizophrenia. More recently, research has focused on fMRI’s ability to monitor response to therapy, with one study showing a functional normalization (an increase in left frontal lobe and cingulate activity compared to baseline) after the addition of an atypical antipsychotic to patients’ medication regimen.27

Figure 3. Chemical shift imaging in a patient with a right hemisphere tumor. From the spectroscopic data, a color image indicating the distribution of choline is prepared. This image shows that the posterior portion of the tumor has the highest choline levels. A spectrum shows the elevation of choline with respect to creatine and NAA. (Some lipid is present in the spectrum as well, indicating necrosis.)


In contrast to fMRI, which relies on secondary hemodynamic measurements to localize functional regions, magnetoencephalography (MEG) is a brain mapping technique that measures the magnetic fields associated with neuronal electrical events.28 This technique is the MRI equivalent of the EEG, and many view the two techniques as complementary.

MEG requires sensitive detectors (capable of identifying tiny extracranial neuromagnetic fields at a femtotesla level) and a shielded room to protect the system from ambient noise. SQUID (superconducting quantum interference device) technology acts as a low-noise amplifier coupled to a primary detection coil.29 This technology, developed only recently, is the backbone of MEG and allows detection of the so-called time-varying magnetic field. Many methods are available to then localize the source of the detected magnetic signal, but the “single equivalent current dipole method” is the most frequently used. The magnitude of a modeled magnetic field is compared to the detected signals, and a least-squares optimization procedure is performed. The final source is overlaid on MR images and a magnetic source image (MSI) is obtained. When MSI is used to map the dipoles created by a specific task or stimulus, it maps the anatomic localization of the neurons involved in the task (see Figure 5, page 112). It is also used to map the location of spike activity, to localize epileptogenic regions in patients with epilepsy.

Compared with the gold standard of intraoperative recordings, MSI’s efficacy is similar to that of fMRI for presurgical mapping. In addition, MEG’s ability to map spontaneously occurring events, such as interictal electrical discharges occurring in epileptogenic tissue, provides a powerful, noninvasive means of mapping.

Roberts et al have proposed that the temporal morphology of neuronal activity may someday provide quantitative indices appropriate for the evaluation of learning disabilities and other neurologic disorders, such as dyslexia or attention deficit hyperactivity disorder. They suggest the possibility of a fusion of modalities in the future, where fMRI, EEG, and PET combine with MEG/MSI and provide the first complete evaluation of brain function with both excellent localization and a temporal dimension to activity provided by MEG. 28

Numerous researchers are evaluating MEG’s utility in the clinical realm. The temporal morphology of event-related fields is an important component in language-based learning disabilities.30-31 Similar to fMRI, one promising application of MSI is in evaluating cortical plasticity. Studies have shown the human cortex reorganizes after stroke, amputation, or even training. MEG may have a role in guiding rehabilitation of specific, unaffected brain regions.

Figure 4. Right hemispheric activation in a patient with a left hemisphere tumor, indicating mixed or right hemispheric dominance for language.

Figure 5. (a) fMRI study showing activation secondary to finger tapping in a patient with a large mass. (b) MEG study showing the location of the magnetic dipole secondary to finger movement in the same patient.


The techniques reviewed provide physiologic information to supplement the anatomic detail in MR images. These techniques can be implemented on 1.5T imagers. As 3T and other high-field magnets become available in clinical practice, these functional techniques will become more powerful due to the increased signal to noise in the stronger magnetic fields.  These imaging capabilities are changing the way in which patients are treated. The MRI techniques described in this article offer greater capabilities in predicting outcomes and monitoring response to therapy. The advancement in MR imaging also drives the development of MR machines with which a radiologist may perform procedures to obliterate lesions or restore blood flow.

Brian Jellison is a third-year radiology resident, Department of Radiology, University Hospitals, Madison, Wis.


  1. Provenzale JM, Sorensen AG. Diffusion-weighted MR imaging in acute stroke: theoretic considerations and clinical applications. AJR Am J Roentgenol. 1999;173:1459-1467.
  2. Forbes KP, Pipe JG, Karis JP, Heiserman JE. Improved image quality and detection of acute cerebral infarction with PROPELLER diffusion-weighted MR imaging. Radiology. 2002;225:551-5.
  3. Pipe JG, Farthing VG, Forbes KP. Multishot diffusion-weighted FSE using PROPELLER MRI [published correction appears in Magn Reson Med. 2002;47:621]. Magn Reson Med. 2002;47:42-52.
  4. Jellison BJ, Wu Y, Field AS, Alexander AL. PROPELLER MR for diffusion-weighted imaging in regions of magnetic field inhomogeneity: brains and cervical spinal cord. American Society of Neuroradiology 2003 Proceedings. 41st Annual Meeting; April 26-May 2, 2003; Washington, DC.
  5. Murata T, Shiga Y, Higano S, Takahashi S, Mugikura S. Conspicuity and evolution of lesions in Creutzfeldt-Jakob disease at diffusion-weighted imaging. AJNR Am J Neuroradiol. 2002;23:1164-72.
  6. Rosen BR, Belliveau JW, Vevea JM, Brady TJ. Perfusion imaging with NMR contrast agents. Magn Reson Med. 1990;14:249-265.
  7. Rosen BR, Belliveau JW, Buchbinder BR, et al. Contrast agents and cerebral hemodynamics. Magn Reson Med. 1991;19:285-292.
  8. Rosen BR, Belliveau JW, Aronen HJ, et al. Susceptibility contrast imaging of cerebral blood volume: human experience. Magn Reson Med. 1991;22:293-299.
  9. Sunshine JL, Bambakidis N, Tarr RW, et al. Benefits of perfusion MR imaging relative to diffusion MR imaging in the diagnosis and treatment of hyperacute stroke. AJNR Am J Neuroradiol. 2001; 22:915-921.
  10. Yoshikawa T, Murase K, Oku N, et al. Heterogeneity of cerebral blood flow in alzheimer disease and vascular dementia. AJNR Am J Neuroradiol. 2003;24:1341-1347.
  11. Yang S, Law M, Zagzag D, et al. Dynamic contrast-enhanced perfusion MR imaging measurements of endothelial permeability: differentiation between atypical and typical meningiomas. AJNR Am J Neuroradiol. 2003;24:1554-1559.
  12. Hendrich KS, Kochanek PM, Williams DS, Schiding JK, Marion DW, Ho C. Early perfusion after controlled cortical impact in rats: quantification by arterial spin-labeled MRI and the influence of spin-lattice relaxation time heterogeneity. Magn Reson Med. 1999;42:673-81.
  13. Provenzale JM, Wang GR, Brenner T, Petrella JR, Sorensen AG. Comparison of permeability in high-grade and low-grade brain tumors using dynamic susceptibility contrast MR imaging. AJR Am J Roentgenol. 2002;178:711-6.
  14. Moseley ME, Cohen Y, Kucharczyk J, et al. Diffusion-weighted MR imaging of anisotropic water diffusion in cat central nervous system. Radiology. 1990;176:439-446.
  15. Basser PJ, Matiello J, Le Bihan D. MR diffusion tensor spectroscopy and imaging. Biophys J. 1994;66:259-267.
  16. Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative diffusion tensor MRI. J Magn Reson. 1996;111:209–219.
  17. Jellison BJ, Wu Y, Field AS, Hasan KM, Alexander AL, Badie B. Diffusion tensor imaging metrics for tissue characterization: discriminating vasogenic edema from infiltrating tumor. American Society of Neuroradiology 2003 Pro- ceedings. 41st annual meeting; April 26-May 2, 2003; Washington, DC.
  18. Arfanakis K, Haughton VM, Carew JD, Rogers BP, Dempsey RJ, Meyerand ME. Diffusion tensor MR imaging in diffuse axonal injury. AJNR Am J Neuroradiol. 2002;23:794-802.
  19. Paley M. Proton spectroscopy of the human brain. In: Bradley WG Jr, Bydder GM, eds. Advanced MR Imaging Techniques. St Louis: Mosby-Year Book; 1997:309-332.
  20. Nelson SJ, Vigneron DB, Dillon WP. Serial evaluation of patients with brain tumors using volume MRI and 3D 1H MRSI. NMR Biomed. 1999;12:123-138.
  21. Spahr L, Burkhard PR, Grotzsch H, Hadengue A. Clinical significance of basal ganglia alterations at brain MRI and 1H MRS in cirrhosis and role in the pathogenesis of hepatic encephalopathy. Metab Brain Dis. 2002;17:399-413.
  22. Tarasow E, Wiercinska-Drapalo A, Kubas B, et al. Cerebral MR spectroscopy in neurologically asymptomatic HIV-infected patients. Acta Radiologica. 2003;44:206-12.
  23. Davidson MC, Thomas KM, Casey BJ Imaging the developing brain with fMRI. Ment Retard Dev Disabil Res Rev. 2003;9:161-7.
  24. O’Doherty J, Kringelbach ML, Rolls ET, Hornak J, Andrews C. Abstract reward and punishment representations in the human orbitofrontal cortex. Nat Neurosci. 2001;4:95-102.
  25. Brannen JH, Badie B, Moritz CH, Quigley M, Meyerand ME, Haughton VM. Reliability of functional MR imaging with word-generation tasks for mapping Broca’s area. AJNR Am J Neuroradiol. 2001;22:1711-8.
  26. Ugurbil K, Toth L, Kim DS. How accurate is magnetic resonance imaging of brain function? Trends Neurosci. 2003;26:108-14.
  27. Nahas Z, George MS, Horner MD, et al. Augmenting atypical antipsychotics with a cognitive enhancer (donepezil) improves regional brain activity in schizophrenia patients: a pilot double-blind placebo controlled BOLD fMRI study. Neurocase. 2003;9:274-82.
  28. Roberts PL, Poeppel D, Rowley HA. Magnetoencephalography and magnetic source imaging. Neuropsychiatry Neuro-psychol Behav Neurol. 1998;11(2):49-64.
  29. Clarke J. SQUIDs. Sci Am. 1994; 271(2):36-43.
  30. Tallal P, Miller SL, Bedi G, et al. Language comprehension in language-learning impaired children improved with acoustically modified speech. Science 1996;271:81-4.
  31. Merzenich MM, Jenkins WM, Johnston P, Schreiner C, Miller SL, Tallal P. Temporal processing deficits of language-learning impaired children ameliorated by training. Science. 1996;271:77-81.