Department of Diagnostic Radiology, Tampere University Hospital, Tampere, Finland.
1. Introduction
With the induction of magnetic resonance imaging (MRI) and computed tomography (CT) in the last two decades,
simple ands fast examination methods became available which has led to a separation of neuroradiology from its
parent fields. Further development of MRI with methods that allow to measure brain perfusion and visualize brain
function has led this diagnostic modality back to its origin, neurology: adequate evaluation of MRI is possible
in the context of all neurosciences (1). With advent of open magnets, which require close cooperation between the
neuroradiologist and the neurosurgeon it is now possible to perform interventional neuroradiology with endovascular
procedures of the brain. Using high resolution MRI and fMRI along with other imaging modalities like positron emission
tomography (PET) and single photon emission computed tomography (SPECT) in psychiatry, it is now possible to carefully
reassess the our sociological understanding to mental illnesses (2). Diffusion MR imaging serves to detect early
stroke and perfusion imaging has a clinical impact in the assessment of brain tumors and cerebral ischemia (3).
Fast spin echo imaging promises rapid MR imaging of the brain leading to better resolution (4) especially in small
pathological lesions of brain. Magnetization transfer imaging has helped detect changes in the structural status
of brain parenchyma that may not be visible with standard MRI techniques (5) in diseases like multiple sclerosis
and other white matter diseases. Several segmentation techniques now help in the quantitative analysis of brain
lesions and reliable three dimensional imaging (6) in diseases like multiple sclerosis, dementia, infarctions and
brain tumors. Fusion of data obtained from different imaging modalities (7) like CT, MRI, SPECT and PET have enhanced
the accuracy of differential diagnosis of different brain lesions. MR spectroscopy coupled with MR imaging techniques
allows for the correlation of anatomic and physiologic changes with changes in the metabolic and biochemical processes
occurring within a pathological brain tissue (8). This review shall discuss different MRI techniques in aiding
a complete pathological and functional analysis of brain lesions in different neurological and neurosurgical diseases.
2. New MRI techniques
2.1 Diffusion MR imaging
For the first time, now it is possible to noninvasively measure and depict molecular diffusion coefficients
in vivo with MR imaging. Studying molecular displacements over distances comparable to cell dimensions has provided
information abut the geometry and spatial organization of the tissue compartments and about water exchange between
these compartments in normal and diseased states.
Diffusion-weighted images are obtained by incorporating strong magnetic field gradient pulses into an imaging
pulse sequence. In a diffusion-weighted image, structures with fast diffusion are dark because they are subject
to greater signal attenuation, whereas structures with slow diffusion are bright. Quantitative diffusion images
are generated from a series of diffusion-weighted images. The term apparent diffusion coefficient (ADC) is used
to quantitatively describe the results of diffusion imaging in vivo. This can be used in the early diagnosis of
stroke, assessment of white matter diseases and monitoring of tissue temperature changes during hyperthermia or
laser surgery. principal headings.
In acute ischemia, ADC decrease significantly within minutes, and this decrease develops progressively within
the first hour. In subacute ischemia the ADC in white matter increases to about two- or threefold above its normal
value. The increase is due to the presence of vasogenic edema, in which bulk water motion occurs in the extracellular
space. In chronic ischemia low diffusion has been seen on diffusion images. Thus diffusion imaging offers the unique
opportunity of noninvasively study ischemia during its various stages. Diffusion imaging may also help detect encephalomalacic
cysts.
2.2 Perfusion MR imaging
By introducing a tracer into the circulation and monitoring its concentration in a tissue over time using MRI,
one can determine the rate of delivery of the tracer and hence the blood flow to this tissue. In clinical imaging.
Gadolinium-DTPA has been used as a tracer.
When the Gadolinium-DTPA reaches the brain capillaries it induces a difference in magnetic susceptibility between
the blood compartment and the brain tissue, where the contrast agent dose not penetrate because of the blood brain
barrier. Diffusion of water through the internal gradients produce a low signal attenuation. Ultrafast imaging
technique such as echo planar imaging (EPI) can monitor the first passage of contrast agents through the brain
tissue within a few seconds of time. Thereafter the quantification of blood flow is performed, a procedure not
so easy.
Perfusion imaging has widespread applications in tissue characterization, treatment monitoring in brain tumors
and function sudies. It has been used to differentiate active recurrent brain tumors from fibrous tissue occurring
secondary to operation or radiotherapy. Also its use has been important in evaluating infarctions and ischemic
conditions. Perfusion weighted imaging provides an inexpensive, safe, reliable, and accurate technique in evaluating
blood flow measurements in brain compared to PET.
2.3 Functional MR imaging (fMRI)
The first attempt to investigate brain activity was made in 1991 (9). Studies have been performed at 1.5 Tesla
MR units using EPI and conventional fast gradient echo images. The most common approach has been the Blood Oxygen
Level Dependent contrast technique (10). Using the cooperation of the patient, who is told to perform certain tasks,
different brain centers can be activated.
Initial fMRI studies have demonstrated activation in the primary cortices like the visual, sensorimotor and
auditory cortices. MR signal changes in these cortices have been correlated with the rate of stimulation. FMRI
detects activation during higher order cognitive functions such as language tasks, motor learning, motor ideation,
or visual mental imagery (11).
Functional MRI has been used to study certain clinical problems like presurgical mapping, imaging of the epileptic
foci, monitoring recovery after stroke or head trauma and following treatments using neuropharmaceutical agents.
In addition to being a tool to study anatomy of the brain, it is becoming a powerful functional tool to understand,
detect and manage functional disorders of the brain (12).
2.4 Magnetic resonance spectroscopy (MRS)
MR imaging and MR spectroscopy are basically one and the same technique, differing only in the manner in which
the data are processed. In MR imaging the signal obtained in the time domain is used to generate an image, whereas
in MR spectroscopy the Fourier transform of MR signal in the time domain is used to generate a frequency domain
spectrum of components that make up the image (13).
Two important factors in in vivo MRS are the volume localization technique used and how the signal measurement
is affected by the type of localization procedure employed. At present there are two types of single volume localization
techniques used in clinical MRS and they are the stimulated echo acquisition mode (STEAM) and point resolved spectroscopy
(PRESS) techniques.
MRS is used in the clinical field at present in differentiating brain tumors from brain infarctions or inflammatory
white matter diseases. Also it is used in the better understanding of the metabolite type produced in infarctions,
ischemia, and tumors. The potential clinical applications today are the differentiation of cancerous from non-cancerous
tissue, infarcted from the normal tissue and necrosis from recurrent tumor.
2.5 MR segmentation techniques
Segmentation is the process of assigning labels to pixels in 2D images or voxels in 3D images. The effect is
that the image is split up into segments also called as regions or areas. In medical imaging this is essential
for quantification of outlined structures and for 3D visualization of relevant image data (6).
The segmentation techniques can be classified into manual, semiautomatic and automatic techniques. Manual technique
is too tedious to be used in clinical trials. Automatic segmentation techniques suffer from the inaccuracies created
by the inpriori knowlegde fed into the technique. Semiautomatic segmentation is the best of all because of its
ease, fastness and manual intervention.
For clinical purposes segmentation techniques using MRI has been widely used in phase three trials in multiple
sclerosis, and in monitoring brain infarctions, brain tumors, dementia and brain atrophy. Reliable 3D images constructed
from the segmented images helps understand the relation between the lesions and surrounding normal brain structures.
2.6 Three dimensional MRI (3D MRI) and multimodal image fusion
Three dimensional imaging is now widely available and used often to aid in the comprehension and application
of volumetric data to diagnosis, planning and therapy. Models of the image data can be visualized by volume or
contour surface rendering and can yield quantitative information (14).
3D presentation of the bone and skin surface of the skull serves to orient the viewer , while planar reformations
and/or transparent projections can be applied for the assessment of the brain structures of interest. Overlaid
3D MRI image-guided neuronavigation techniques that allow navigation during operative procedures have demonstrated
their usefulness (15). Combined assessment of three dimensional anatomical and functional images (SPECT, PET, MRI
and CT) is useful to determine the extent of lesions in the brain (16).
Analysis, visualization, method characteristic image processing and image synthesis is needed not only for the
interpretation of the images but also for performing effective consultations with clinical colleagues and computer
supported therapy planning and control strategies in radiological and clinical practice (17).
2.7 Fast spin echo (FSE) imaging and other new MR sequences
FSE imaging is a modified rapid-acquisition relaxation–enhanced technique that affords rapid MR imaging while
retaining true spin-echo contrast features. By manipulating the echo train length, echo spacing and order of phase
encoding, images may be obtained many times faster than with conventional spin echo images.
Fast fluid attenuated inversion recovery (fast FLAIR) is being increasingly used in detecting subarachnoidal
haemorrhage in addition to detecting brain parenchymal lesions situated near the cerebrospinal fluid spaces. Magnetization
transfer images are being increasingly used to detect subclinical lesions in the white matter which are not detected
by conventional MRI (5)
The use of these new techniques are mainly focussed now in diseases like multiple sclerosis, AIDS and epilepsy.
The use of FSE has reduced the imaging times considerably, an advantage to the sick patient.
3. MR imaging in specific diseases
3.1 Epilepsy
The development of MR imaging with its outstanding sensitivity and specificity in identifying structural and
pathological abnormalities has made the neuroradiological contributions in the presurgical work-up of patients
with drug resistant epilepsy of extreme importance (18) . Multiplanar high resolution T1- and T2-weighted FSE images,
fast gradient echo or hybrid images, FLAIR, magnetization transfer, MR angiography, MR perfusion, MR diffusion
weighted imaging, fMRI and 3D acquisition images for morphological details and coverage of the entire brain has
made MRI invaluable in epilepsy. The various disease detected are mesial temporal sclerosis, focal migration disorders,
dysplastic lesions and low- grade tumors, phakomatosis, and vascular anomalies. MRI is being used for placing depth
electrodes. Never before epilepsy could be diagnosed so well as now.
3.2 Human immunodeficiency virus type 1 (HIV-1)
Advances in MR imaging techniques and MR spectroscopy provide noninvasive diagnostic approaches to this disease.
Cerebral inflammation appear as high intensity lesions on T2-weighted images and lactate, often present in macrophages
associated with inflammatory lesions can be visible on spectroscopy. Diffusion MR imaging shows the physiologic
changes associated with disease states. The typical focal inflammatory lesions are caused by toxoplasmosis, lymphoma,
cryptococcoma, tuberculoma, sypilitic gumma, bacterial abscesses, cytomegalic virus infections and progressive
multifocal encephalopathy. Both MRS and diffusion weighted imaging can be used to monitor these disease states
(19).
3.2 Pediatric diseases
Diffusion imaging holds great promise for the evaluation of hypoxic-ischemic injuries related to perinatal birth
injury (20). It can detect early changes as early as 1 hour after onset of the symptoms. A primary examination
in hydrocephalus patients is a detailed MRI and cerebrospinal flow studies using phase contrast MRI. All anomalies
of the brain optic pathways and spinal cord can be diagnosed using MRI.
4. Trends in MR imaging
A current trend in clinical MR imaging has been to move towards more open magnet designs. Open MR systems have
the advantage of cost, ease of use, and niche applications like trauma and interventional . Open systems are also
more accomodating to the claustrophobic and obese patients. Considerable progress has been made in MR imaging guided
interventional and intraoperative MR imaging (21). Functional MR imaging will offer images of the entire brain
in real time, and at a modest cost to the neurologists, psychologists and psychiatrists. Diffusion MR imaging is
being actively used in early stroke diagnosis.
MR spectroscopy is becoming a routine for several clinical assessments. Spectroscopy with other nuclei like
carbon-13, nitrogen-14, fluorine-19 etc. present major challenges associated with low natural abundance. Overall,
MRI with its versatility has been able to meet every challenge encountered in the field of neuroradiology.
References
[1] Scroth G, Ozdoba CH, Remonda L: Neuroradiology-past and future. Schweiz Rundsch Med Prax 84: 1490-500, 1995.
[2] Jabourian AP, Benhamou PA, Bitton R: Clinical imaging in psychiatry. Ann Ned Psychol (Paris) 154:74-7, 1996.
[3] Bruning R, Weber J, Wu RH, Kwong KK, Hennig J, Reiser M: Echo-planar imaging of the brain. Radiologie 35:902-10,
1995.
[4] Jolesz FA, Jones KM: Fast spin-echo imaging of the brain.Top Magn Reson Imaging 5:1-13, 1993.
[5] Grossman RI, Gomori JM, Ramer KN, Lexa FJ, Schnall MD: Magnetization transfer: theory and clinical applications
in neuroradiology. Radiographics 14:279-90, 1994.
[6] Suetens P, Bellon E, Vandermeulen D, Smet M, Marchal G, Nuyts J, Mortelmans L: Image segmentation: methods
and applications in diagnostic radiology and nuclear medicine. Eur J Radiol 17:14-21.
[7] Rolland Y, Lemoine D, Biraben A, Barillot C: Fusion of data. Multimodality in neurological imaging. Ann Radiol
(Paris) 36:375-80, 1993.
[8] Kwock L: Localized MR Spectroscopy. In Proton MR Spectroscopy of the Brain , Neuroimaging Clinics of North
America, Saunders Press, Philadelphia, 1998.
[9] Belliveau JW, Kennedy DN, McKinstry RC, et al.: Functional mapping of the human visual cortex by magnetic resonance
imaging. Science 254:716-19, 1991.
[10] Kwong KK, Belliveau JW, Chesler DA: Dynamic magnetic resonance imaging of he human brain activity during primary
sensory stimulation. Proc Natl Acad Sci , USA 89:5675-9, 1992.
[11] Le Bihan D, Turner R, Zeffiro A, Cuenod CA, Jezzard P, Bonnerot I: Activation of human primary visual cortex
during visal recall: A magnteic resonance imaging study. . Proc Natl Acad Sci , USA 90:11802-5, 1993
[12] Le Bihan D: Diffusion, Perfusion and Functional MRI- Functional MRI, Springer- Verlag, Berlin, 1995.
[13] Shaw D: The fundamental principles of nuclear magnetic resonance- Biomedical Magnetic Resonance Imaging, VCH
publishers, New York, 1988
[14] Kremer JR, Mastronarde DN, McIntosh JR: Computer visualization of three-dimensional image data using IMOD.
J Struct Biol 116.71-6, 1996.
[15] Iseki H, Masutani Y, Iwahara M, et al.: Volumegraph (overlaid three-dimensional image guided navigation). Clinical
application of augmented reality in neurosurgery. Stereotact Funct Neurosurg 68:18-24, 1997.
[16] Treves ST, Mitchell KD, Habboush IH: Three dimensional image alignment, registration and fusion. Q J Nucl Med
42:83-92, 1998.
[17] Engelmeier KH, Fink U, Hilbertz T: Visualization of multimodal image information in medicine. Proc Annu Symp
Comput Appl Med Care 25-9, 1992.
[18] Meiners LC, Valk J, Jansen GH, van Veelen CWM: MR contribution in surgery of epilepsy. Eur Radiol 9:493-507,
1999.
[19] Chang L, Ernst T: MR spectroscopy and diffusion weighted MR imaging in focal brain lesions in AIDS- Neuroimaging
of AIDS, Neuroimaging Clinics of North America, Saunders Press, Philadelphia, 1997.
[20] Philips MD, Zimmerman RA: Diffusion imaging in pediatric hypoxic- ischemic injury- Pediatric Neuroimaging,
Neuroimaging Clinics of North America, Saunders Press, Philadelphia, 1999.
[21] Rothenberg LN, Nath R, et al.: A Perspective on the New Millenium. Radiology 209.600-603, 1998.