History

The interaction of the intrinsic magnetic moment of the nucleus with an externally imposed magnetic field results in the phenomenon known as nuclear magnetic resonance (NMR). Two independent groups, Felix Bloch working with liquid water²² and Edwin Purcell working with solid paraffin,²³ detected the hydrogen nucleus resonance in 1946 in bulk matter. Bloch further described the processes and time constants (T1 and T2) by which the resonance would dissipate.²⁴ This set the stage for the eventual development of MRI 30 years later. Between 1946 and 1976, NMR became a useful laboratory tool for probing molecular structure. Laboratory NMR instruments had small spaces for the sample, usually a small test tube. Use of NMR for larger objects, such as humans, required the development of larger magnets with larger sample spaces. The term magnetic resonance imaging is now used rather than NMR to allay patient anxiety about a test that has the word “nuclear” in it.

Basic Physics of Magnetic Resonance Imaging

Creating the Signal

To begin, the sample is immersed in a strong, constant magnetic field. A magnet that may be one of three designs creates the field. First is the electromagnet, similar in principle to a washing machine solenoid. Bloch and Purcell used these magnets in their original experiments. However, the maximal field strength that can be achieved is limited in practical applications to around 0.4 T (1 T = 10,000 G). An electromagnet consumes large amounts of electricity, and its use in MRI has therefore declined.

The second magnet type is a permanent magnet assembled from ferromagnetic material. The field strength of this type of magnet is limited to 0.3 T. Permanent magnets are generally used for small, low-cost open designs. MRI systems that accommodate large (>300 lb) or claustrophobic patients tend to use this technology.

The third and most widely used magnet type is the superconducting magnet. This design is also similar to a washing machine solenoid. However, unlike the solenoid, the wire is an alloy that conducts electricity without resistance when kept at temperatures within 15 degrees of absolute zero. An electric current is slowly driven into the magnet. Once the current reaches the desired level, the ends of the magnet wire are connected, which forces the current to circulate continuously without loss. Magnets of this type can remain at field strength for many years without the addition of electric current. Field strengths of up to 8 T can be achieved in these magnets, which can accommodate human subjects. Most MRI systems now use superconducting magnets, usually 1.0 to 1.5 T, although an increasing number of hospitals and imaging centers are now using 3-T clinical MRI systems.

When immersed in the strong, constant magnetic field, the spins in the sample experience a slight polarization. This polarization, known as M0, increases with increasing magnetic field. At 1.5 T, this polarization is very slight, about 1 × 10⁻⁵. This translates to just 10 in 1 million nuclei being polarized.²⁵ The polarization competes with the randomizing effect of the thermal vibration (Fig. 17-9). It is only the polarized nuclei that contribute to the MR signal; hence, MRI systems with higher field strength produce better images.

FIGURE 17-9 Illustration of nuclear spins in the magnetic field B0. The net alignment of spins is along B0. This is the source of M0, or longitudinal magnetization. In reality, the net aligned versus unaligned fraction is about 1 in 100,000 in a 1.5-T field at body temperature.

We will use the classic model created by Bloch to describe the motion of the spins. Spins that align with the main field precess around the direction of the main field in a manner similar to the spinning of a toy top (Fig. 17-10). The rate of this precession is a product of the intrinsic magnetic moment (the gyromagnetic ratio) of the spin and the strength of the main magnetic field. The rate of precession is known as the Larmor or resonant frequency.²⁶ The spins aligned along B0 are rotated into a plane transverse to the direction of the main magnetic field by the action of a time-varying magnetic field called B1 (Fig. 17-11). The B1 field is created by the radiofrequency (RF) transmitter and the antenna, known as the RF coil. The frequency of the B1 field matches the Larmor frequency of the spins. The B1 field is of brief duration, typically 1 to 10 msec, and is thus referred to as an RF pulse. The angle through which the spin is rotated is called the flip angle. The flip angle depends on the duration and amplitude of the RF pulse.

FIGURE 17-10 “Spinning top” model of nuclear spin. The spin of nucleus “a” creates the magnetic moment of the nucleus, which causes the spin to precess along circle “b” around magnetic field B0. The rate of this precession around B0 is called the Larmor frequency.

FIGURE 17-11 The applied (transmitted) radiofrequency magnetic field B1 causes the spin to experience a torque, which twists (nutates) the M0 magnetization into the transverse plane x-y, where it is referred to as Mxy.

Detecting the Signal

To detect the MR signal, an RF coil is placed as shown in Figure 17-12. This may be the same coil used to transmit the B1 pulse. A time-varying magnetic field will be created at the coil by the magnetic field of the precessing spins (rotated in the transverse plane) as they pass by the RF coil. Magnetic induction (Faraday’s law) causes the RF coil to produce an electric current, which can then be amplified and detected (Fig. 17-13). An interesting and important disparity should be noted. The amount of power used to produce the RF pulse is in the range of 100 to 20,000 W. The signal that is received from the object is on the order of 10⁻¹² W. This received signal is no greater than the signals from radio or television stations, among other sources. To prevent interference from these external sources, MRI systems are enclosed in electrically shielded rooms, which are six-sided copper boxes.

FIGURE 17-12 The radiofrequency (RF) coil both transmits and receives the signal from the spins in the transverse plane.

FIGURE 17-13 Bloch diagram of the transmit and receive scheme used in magnetic resonance imaging that illustrates the large disparity between the high radiofrequency (RF) power transmitted into the patient and the tiny RF signal received from the patient.

Physics: Localizing the Signal

To this point, the sample has been polarized and excited and a signal detected, but the location of the spins that created the signal remains unknown. Suppose that two objects are in the magnetic field, both of which create a signal. To force each object to give off a unique signal, the magnetic field can be modified to vary as a function of position along the x-axis (Fig. 17-14). The resonant frequency of the spins is a function only of the magnetic field at that point in space. Thus, the spatial origin of the signal can be determined by the frequency of the received signal. In practice, this is done by creating a linear gradient that adds or subtracts to the main field as a linear function of offset from the origin. An MRI system has three gradients: x, y, and z. The gradient system serves two purposes in the MRI system. The first is to limit the excitation to a plane or slab (Fig. 17-15). If an RF pulse is transmitted during the time that a gradient is applied, only a slice or slab will be excited, the thickness of which depends on the amplitude of the gradient and the bandwidth of the RF pulse. The second purpose is to encode the spatial location of spins to form the MR image. In a slice, the two directions of encoding required are frequency and phase. Location along the frequency-encoding axis is accomplished by applying a gradient during the signal readout time. The orthogonal axis is encoded by applying a gradient somewhere between the time of excitation and reception and is called phase encoding. One unique value of phase encoding is applied every time that the readout gradient is applied. Thus, the excitation and readout must be repeated to make an image. The rate at which the excitation is repeated is the TR time.

FIGURE 17-14 Magnetic field gradients alter the strength of the main magnetic field. A single gradient is designed to change the magnetic field along only one direction (x, y, or z). Three such gradients are required to localize signals in three dimensions.

FIGURE 17-15 Slice selection illustrating a group of four excited axial slices relative to a patient.

(Radiofrequency head coil courtesy of Midwest RF LLC.)

The Origin of Image Contrast

The intensity of a voxel in an image arises from the three principal factors. The first is the number of protons in the voxel, known as proton density. If this were the only image mechanism, MRI would be little more than a CT scanner. However, intensity in the voxel also depends on the relaxation rates of the spins. The first is T1, or the longitudinal relaxation rate (Fig. 17-16A). After excitation, the magnetization in the slice returns along the axis of the main magnetic field by interaction with other nonmoving hydrogen spins, typically those attached to large molecules. The magnetization returns along B0 as an exponential function of the ratio of T1 and the rate at which the excitation is repeated, or TR. The second relaxation rate is T2 (Fig. 17-16B). This rate describes the rates at which spins that have been excited into the transverse plane lose coherence. Each spin precesses at a rate that is determined by the magnetic field at the location of that spin. Macroscopic and microscopic field gradients, created by differences in the magnetic susceptibility of tissue, cause some of the excited spins to precess faster and some slower. Eventually, the spins are spread out uniformly, which produces no signal in the RF coil used to detect the spins.

FIGURE 17-16 A, Recovery of longitudinal magnetization (Mz) versus time for three different substances shown with regrowing vectors and plotted. B, Decay of Mx versus time for three different substances. CSF, cerebrospinal fluid.

Three principal image intensity factors—proton density, T1, and T2—influence the appearance of every voxel on MRI. For example, cortical bone is hypointense on MRI because the proton density (of mobile spins) is quite low. Lung parenchyma is usually hypointense because the T2 of lung tissue is so low that the signal is gone before it can be sampled. The long T1 of liquid water causes CSF to be dark on sequences that use short TR times.

An MRI scan user has a choice of pulse sequence parameters such as echo time (TE) and repetition time (TR). Maximal contrast between structures of interest may be achieved by the appropriate choice of sequence parameters. Conversely, inappropriate choices may result in failure to detect a lesion. Other endogenous sources of contrast include blood flow (magnetic resonance angiography [MRA]) and water-macromolecular T1-based interactions (magnetization transfer).

Spin Echo

After the spins have been rotated into the transverse plane by the initial 90-degree RF pulse (Fig. 17-17A), they begin to lose coherence because of the effects of local inhomogeneities (contributed by changes in tissue), inhomogeneities secondary to imperfections in the B0 field, and diffusion of water molecules (Fig. 17-17B). If a second RF pulse is transmitted at twice the amplitude of the first pulse, the relative direction of the spins can be reversed (Fig. 17-17C). Then, at the echo time, the spins will have nearly regained coherence (Fig. 17-17D). The effects of magnetic field inhomogeneities are thus cancelled out. One loss, that caused by diffusion, cannot be reversed but is usually negligible in routine imaging. The resulting coherence produces the spin echo as described by Erwin Hahn in 1950.²⁷ If the 180-degree pulse is not used, the signal decreases with increasing echo time as T2*. With the 180-degree pulse included, the decrease is T2, with T2 always being greater than T2*.

FIGURE 17-17 The spin echo starts with nutation of longitudinal magnetization (Mz) into the transverse plane (A). The magnetization decays over time (B) until the radiofrequency (RF) 180-degree pulse (C) reverses the direction of the spins, which then reform Mx at the echo time (D).

The spin echo sequence is the mainstay of routine clinical imaging. By adjusting TE and TR times, proton density–, T1-, or T2-weighted images can be selected (Table 17-1).

TABLE 17-1 Weighting of Magnetic Resonance Images

TE, echo time; TR, repetition time.

Gadolinium Contrast

Exogenous contrast enhancement is now a routine part of MRI. The most widely used is a gadolinium chelate. The gadolinium atom is strongly paramagnetic and acts to shorten the T1 relaxation time of nearby water protons in blood.²⁸ The agent does not pass the blood-brain barrier and thus becomes a marker for abruptions in the blood-brain barrier caused by neoplasm, infection, trauma, and infarction. The T1-shortening effect is also used for rapid MRA, which is performed by using a fast scanning protocol after the bolus injection of a contrast agent. In the brain, bolus injection of gadolinium with repeated echo planar imaging (EPI) has been used to image perfusion²⁹ and the blood volume of tumors.³⁰

Fast Spin Echo

Fast spin echo (FSE) sequences decrease scan time by increasing the efficiency of data collection.³¹ The increased efficiency can be used to either decrease scan time or increase the signal-to-noise ratio of the resulting images. Improved scan efficiency has resulted in the frequent application of FSE sequences in radiology, particularly in imaging of the CNS.

To understand this improved efficiency, it is necessary to understand how the data are collected. In a conventional spin echo sequence (Fig. 17-18A), a single line of k-space samples (called a view) is collected. For images that are not fractional excitation, the number of views is equal to the matrix size in the phase that encodes image direction. The next view is collected at TR time later, when the sequence is repeated. The entire k-space matrix must be collected before the images can be reconstructed. The TR time is set in accordance with the desired contrast (Table 17-1). The FSE sequence collects multiple views in each TR time (Fig. 17-18B). The number of views collected per TR time is known as the echo train length (ETL). An FSE sequence with an ETL of 4 has a total scan time a fourth that of a conventional spin echo sequence with equivalent TR. The ETL can be equal to the number of views in the sequence. This allows the entire image to be collected in a single TR time.

FIGURE 17-18 A, Multiple-echo spin echo sequence in which a single phase encoding is applied for all four echoes and each echo fills in the same sample line in four separate k-space matrices. The sequence must be repeated a number of times equal to the resolution in the phase-encoding direction of the image. B, A fast spin echo sequence applies a different phase encoding for each echo of the same sequence as in A. In this way, a single k-space matrix is filled four times faster. This yields one image in a quarter of the time required to complete the sequence in A. ETL, echo train length.

The improved efficiency of FSE sequences comes at the expense of image contrast purity because each view is collected at a different echo time (Fig. 17-18B). The effect is to create image blurring that worsens with increasing ETL or decreasing tissue T2. The relatively long T2 times of tissues in the neuraxis allow ETL values of 16 to be used routinely. To image uncooperative patients, single-shot FSE sequences can be used, but with some increase in image blurring.

Inversion Recovery

Image contrast can be further manipulated by transmitting a 180-degree RF pulse before the pulse sequence. The effect of using this pulse is to rotate the spins from their orientation along +z to −z (Fig. 17-19A). The longitudinal magnetization (Mz) signal regrows from −z to +z at a rate determined by the T1 of the spins. The regrowth is plotted in Figure 17-19B for several tissues. If the 90-degree RF pulse is transmitted at the time at which Mz is 0, the tissue will produce no signal. Inversion recovery can be used to increase contrast between structures, such as gray and white matter, or to null signals that arise from protons of a known T1. A coronal image of a volunteer imaged at multiple T1 times is shown in Figure 17-20. The round object above the volunteer’s head is a jar of mayonnaise. Fluid-attenuated inversion recovery (FLAIR) imaging³² uses this method to null CSF while preserving signal from edematous tissue. The inversion pulse may be applied to a spin echo or gradient echo sequence.

FIGURE 17-19 Inversion recovery sequence showing both the motion of the spins (A) and a plot of longitudinal magnetization (Mz) versus time (B). CSF, cerebrospinal fluid.

FIGURE 17-20 Coronal inversion recovery magnetic resonance images of a volunteer and a jar of mayonnaise. The nulling effect of certain T1 times on various substances is illustrated: T1 of 110 msec nulls mayonnaise, T1 of 150 msec nulls white matter, and T1 of 375 msec nulls gray matter.

Gradient Echo

The pulse that rotates Mz into the transverse plane, nominally a 90-degree pulse, does not have to be followed by an 180-degree pulse to produce an echo. In this case the sequence is referred to as a gradient echo because a readout gradient is used to form the echo. By eliminating the 180-degree pulse, sequence time is saved and TR time may be made very short—as little as 6 msec. In so doing, k-space can be filled very quickly and the scan can be completed in a matter of seconds rather than minutes as is the case with conventional spin echo. At the TR times used in conventional spin echo sequences, 500 to 2000 msec, the MR signal is strongest when the first RF pulse is 90 degrees. When the TR time is very short, the peak signal is obtained at a flip angle of less than 90 degrees. The flip angle becomes an important determinant of image contrast in gradient echo. Low flip angles, typically 5 to 20 degrees, result in images that are more T2* weighted, whereas higher flip angles, 40 to 90 degrees, create more T1-weighted images. It is important to remember that gradient echo sequences do not yield the purity of contrast of a conventional spin echo sequence and that the T2 weighting in a conventional spin echo is replaced by T2* in gradient echo. The advantages of gradient echo are short scan times, utility for MRA, and the ability to visualize hemosiderin and ferritin. MRA does not use spin echo because flow is dephased and becomes invisible.

Echo Planar Imaging

The readout gradient used to form the echo in a gradient echo sequence can be repeated to collect additional k-space views in a manner similar to FSE.³³ Between repetitions of the readout gradient, a small phase-encoding gradient is applied that allows a different view to be collected during the subsequent readout gradient (Fig. 17-21). Only a single excitation is used to collect all the views in k-space. The entire image can be collected in approximately 40 msec. Applications that are highly sensitive to even minor patient motions use this sequence. Like FSE, EPI suffers from blurring caused by acquiring views at different echo times, but without rephasing of the RF 180-degree pulses. This effect limits the resolution obtainable. Although an FSE image may use a matrix of 256 × 256, EPI is limited to 128 × 128 or more typically 64 × 64 over the same field of view. Despite these limitations, EPI is essential for diffusion, perfusion, and functional MRI (fMRI).

FIGURE 17-21 Sequence diagram and image from echo planar imaging (EPI). The readout gradient repeatedly reverses, which fills the k-space matrix in a single excitation. EPI is similar in concept to fast spin echo imaging, except for the repeated 180-degree radiofrequency pulses. In the absence of these pulses in EPI, susceptibility artifacts are present in areas adjacent to bone and air.

Diffusion-Weighted Imaging

Diffusion of water in the brain depends on the integrity of the cellular membrane and the osmotic balance of the neurons and astrocytes.³⁴ In ischemic processes, the diffusion of water first decreases because of cellular swelling and thus decreases the extracellular space. During recovery and into the chronic stage, the loss of cell wall integrity causes the diffusion to increase.

In the section on the spin echo, it was noted that the 180-degree pulse could not recover signal lost because of diffusion. If the water molecule moves between the time of the gradient that is applied before the 180-degree pulse and the gradient after the 180-degree pulse, the spin will experience a different B0 field. The different B0 field is due to the applied gradient and the total motion of the spin in that direction. The spin precesses at a different rate before and after the 180-degree pulse, thereby resulting in an accumulated phase error at the time of readout and a loss of signal. The farther that the spin wanders (wander is the correct term here because the motion is brownian) between gradient lobes, the greater the signal loss. Hence, a chronic infarct is hypointense. Conversely, when the cells swell and the spin motion is reduced to less than that of normal tissue, the diffusion loss is reduced to less than that of normal tissue and the region becomes hyperintense.

Perfusion-Weighted Imaging

Two main approaches are taken to measure tissue perfusion. Both methods use EPI sequences because like diffusion, the effect can be easily degraded by minor patient motion. The first method for imaging tissue perfusion is to create an endogenous contrast. This is done by presaturating the blood flowing into the organ with an RF pulse. EPI of the organ is performed with the presaturation on and off. By subtracting the two images, a tissue perfusion image is created.³⁵ The second method is by bolus injection of contrast material while repeatedly performing EPI. Time course images from EPI are then fitted to a model of the response of the tissue to the presence of the contrast agent, and an image of perfusion is made. This method is also known as a dynamic susceptibility contrast technique and requires a rapid intravenous bolus injection of gadolinium contrast material at a rate of 4 to 5 mL/sec, usually through an 18-gauge intravenous catheter, followed by normal saline injection. Manual or automatic postprocessing is required to generate color maps of various parameters of cerebral perfusion, such as CBV, CBF, mean time to enhance, and MTT.

Perfusion MRI is used clinically to evaluate patients with chronic ischemic disease, acute stroke, vasospasm after SAH, or intracranial neoplasms. It can also be used to distinguish recurrent neoplasm from radiation necrosis in patients who underwent radiation therapy for primary brain tumors.

Spectroscopy

The resonant frequency of a spin is determined not only by the B0 magnetic field and its associated imperfections but also by the molecule of which the nucleus is a part. The effect is slight, but not negligible. This phenomenon is known as chemical shift; that is, the chemical microenvironment of a given nucleus results in a slight change in the resonance frequency of the nucleus from that expected in its pure state. The difference between the resonant frequencies of the protons in water and the hydrogen atoms in the methylene groups in lipid are about 3.5 ppm. This translates to a difference of about 220 Hz for operation at 1.5 T. Between the water proton and fat proton resonances lie resonances of other moieties of interest in the brain. These include myoinositol, a sugar phosphate; choline, a key indicator of membrane turnover; creatine and phosphocreatine, part of the energy pool; glutamate and glutamine, the primary excitatory neurotransmitter and its astrocyte-recycled counterpart; N-acetyl aspartate (NAA), a key indicator of neuronal health; and lactate, an indicator of a shift to anaerobic metabolism. A spectrum of the resonances of the protons in these compounds is shown in Figure 17-22A. These resonances are included in every MR image of the brain. However, the concentration of each is lower than that of water by a factor of 5000 to 10000. To detect the resonances of the metabolites thus requires much larger voxels than imaging does: 2 cm³ versus 1 mm³ in imaging.

FIGURE 17-22 Magnetic resonance spectroscopy. A, Spectra extracted from a (¹H) chemical shift imaging data set. There is increased choline (Cho) resonance and decreased N-acetyl aspartate (NAA) resonance in the spectrum labeled “lesion.” The spectra from the contralateral hemisphere show a normal pattern. Cr, creatine; Glx, glutamate and glutamine; Myo, myoinositol. B, Metabolite images made from the same data set used in A. The increased concentration of choline is seen in the lesion as the red area.

Two principal methods are used to collect spectroscopic data. The first, single-voxel spectroscopy, selects a voxel by using three successive RF pulses and then reads out the signal from the entire voxel. The second, chemical shift imaging, selects a larger volume or slice with one or more RF pulses. Phase-encoding gradients are then applied in one to three directions to yield 2D or 3D spectroscopy data sets. Reconstruction allows the selected region to be subdivided into separate voxels. An individual spectrum can then be extracted from a voxel, or a resonance may be selected and an image made of that resonance. Such 2D and 3D data sets can be postprocessed to generate color maps of metabolite distributions within the volume of brain evaluated with spectroscopy (Fig. 17-22B).

Spectroscopic data can be collected for several different nuclei present in the body, including phosphorus, carbon, and sodium. Exogenous substances that can be detected in the body include fluorine and lithium, both of which are found in psychoactive drugs. However, proton (hydrogen) spectroscopy is the most readily commercially available because of the high abundance of hydrogen, in the form of water and other hydrogen-containing molecules, in comparison to other nuclei. To perform spectroscopy using nuclei other than hydrogen also requires additional hardware and software to allow what is known as multinuclear spectroscopy. In the MRI literature that describes human imaging, the greatest literature volume exists for proton spectroscopy.

Clinical proton MR spectroscopy (MRS) has been commercially available for more than a decade. Nevertheless, MRS remains somewhat limited in use and is found in select academic medical centers and some large community hospitals because of the level of complexity in acquisition and postprocessing of data. However, in centers with active spectroscopy programs, clinical MRS can add value in the diagnosis, follow-up, and thus management of patients with various neurological diseases, including epilepsy, neoplasms, demyelinating disorders, metabolic diseases, and ischemic diseases.^36,37 The clinical applications of proton spectroscopy to neoplasms, stroke, and epilepsy are discussed later in this chapter.

Functional Magnetic Resonance Imaging

fMRI is used to image the change in the ratio of oxyhemoglobin to deoxyhemoglobin in cerebral blood in response to a stimulus.³⁸ Changes in the ratio within activated cortical structures are taken to be indicative of changes in the oxygen extraction fraction.³⁹ These changes are thought to occur in response to cortical function and the resulting upregulation of local metabolism. The change in the oxyhemoglobin-to-deoxyhemoglobin ratio alters the magnetic susceptibility of the blood because deoxyhemoglobin is less paramagnetic than oxyhemoglobin. This difference is manifested as a change in T2* in activated tissue.

The effect is slight; at 1.5 T the change in image intensity is less than 6%. To preclude interference from patient motion, the data are collected with EPI techniques. Images are repeatedly acquired at the same location over the course of the application of a stimulus pattern. By using the a priori knowledge of the stimulus pattern and cross-correlating it with the time course intensity data, a functional image may be made (Fig. 17-23).

FIGURE 17-23 Functional magnetic resonance imaging of a patient with a lesion near the left motor cortex. The functional data overlie a fluid-attenuated inversion recovery image.

Screening

The usual screening MRI of the brain begins with a sagittal localizer of the brain, which is then used to prescribe the remainder of the study. Typically, this includes T1- and T2-weighted axial images and, more recently, a FLAIR sequence, which results in increased lesion conspicuity, especially in the areas adjacent to CSF spaces (i.e., cortex, subcortical white matter, and periventricular deep white matter). At some centers, a gadolinium intravenous contrast agent is given routinely as a part of a screening examination. At other centers, gadolinium is administered at the neuroradiologist’s discretion. The contrast-enhanced sequence usually includes an axial or coronal T1-weighted image, or both. Sagittal contrast-enhanced images can also be added as needed.

Generally, screening MRI of the brain includes the entire brain, orbits, and maxillofacial and skull base regions, including the craniocervical junction. Variable slice thickness, planes, and slice intervals are used. Usually, axial and coronal images are obtained as 5-mm-thick slices, with or without an interslice gap of 1 to 2 mm. Sagittal images are generally 4 to 5 mm in thickness with a 1- to 2-mm interslice gap. Contrast-enhanced images are almost always T1 weighted because the gadolinium contrast material’s enhancement characteristics (T1-shortening effect) are best observed on T1-weighted images, although its effects may also be observed and quantified on T2-weighted images.

Tailored Magnetic Resonance Imaging

There are many different ways to perform a brain MRI examination. For patients in whom clinical suspicion of pathology is high, a more tailored examination instead of a screening study should be done to increase the likelihood of detecting relevant and clinically significant abnormalities. For example, in a patient with a suspected meningioma that involves the cavernous sinus, a more tailored study that includes thin 3-mm contiguous T1-weighted coronal images before and after contrast administration may better define extension of the tumor in relation to the optic chiasm, pituitary infundibulum, Meckel’s cave, and orbital apex. Such a tailored study requires close communication between the neuroradiologist and the neurosurgeon before the study.

Neoplasms

Gliomas

Gliomas are the most common primary brain neoplasms in both adults and children and represent approximately half to two thirds of all brain tumors, respectively, encountered in these populations. Gliomas are usually categorized into four distinct histologic subgroups—astrocytomas, oligodendrogliomas, ependymomas, and choroid plexus tumors, which include the more common papillomas and rare carcinomas and xanthogranulomas.^40,41

The MRI appearances of gliomas are quite varied, but in some cases these neoplasms can have a set of very distinct appearances and location that permit a correct preoperative radiologic diagnosis. In other cases, the MRI appearance is not specific and several differential diagnoses have to be entertained preoperatively. MRI appearances can also vary according to the neoplasm’s histologic grade: lower grade tumors usually exhibit little or no peritumoral vasogenic edema, whereas higher grade tumors demonstrate prominent vasogenic edema associated with a local/regional mass effect or herniation, or both.^42,43 However, the sensitivity and specificity of MRI in assessing the degree of histologic anaplasia of a given neoplasm are limited.

Astrocytomas represent the largest subgroup of gliomas (about 20% to 30% of all gliomas). The histologic subtypes of astrocytomas (e.g., fibrillary, protoplasmic, or gemistocytic) cannot be distinguished by MRI. However, certain subtypes, such as the cystic pilocytic astrocytomas typically found in the cerebellum in children and subependymal giant cell astrocytomas in patients with tuberous sclerosis, may be diagnosed preoperatively by MRI because of their typical appearance, location, and clinical features. Astrocytomas are usually hypointense to normal brain tissue (cortex and white matter) on T1-weighted images and mildly to significantly hyperintense to normal brain on T2-weighted and FLAIR images. Although not completely reliable, in general, low-grade astrocytomas are usually well circumscribed and exhibit no or minimal peritumoral edema, whereas the higher grade anaplastic or glioblastoma multiforme varieties are usually more infiltrative in appearance with less well defined borders and are associated with significant surrounding edema. Low-grade tumors are generally homogeneous in signal characteristics on both T1- and T2-weighted images, whereas higher grade tumors are more likely to be heterogeneous; this reflects the heterogeneity observed by histology associated with the cystic and necrotic changes typically found in these tumors (Figs. 17-24 and 17-25). Rarely, certain astrocytomas demonstrate signal intensity characteristics that are similar or identical to those of CSF on T1- and T2-weighted images.^42,43 Some of these tumors are in fact predominantly cystic in composition, which explains their signal characteristics, but some are solid neoplasms whose parenchymal composition and organization (e.g., microcysts) mimic CSF signals on MRI. Therefore, with MRI, unlike CT, lesions that appear to have the same signal as that of CSF may not be cystic.

FIGURE 17-24 A, Spin echo T1-weighted coronal image with contrast enhancement demonstrating a low-grade astrocytoma as a hypointense mass in the left temporal lobe and inferior frontal lobe that is causing superior displacement of the M1 segment of the left middle cerebral artery. B, Fast spin echo T2-weighted axial image demonstrating that the lesion involves the left basal ganglia and insula, with a mild midline shift to the right. It is homogeneously hyperintense.

FIGURE 17-25 A, Fast spin echo T2-weighted axial image demonstrating a heterogeneously hyperintense mass in the left temporal and occipital lobes. B, Spin echo T1-weighted contrast-enhanced image demonstrating a large cystic anterior component with thin peripheral enhancement and a heterogeneously enhancing solid component along the posterolateral aspect of this glioblastoma multiforme.

Brainstem gliomas, which are usually of a fibrillary histologic type and more common in children, appear as a poorly defined mass within the brainstem with diffuse enlargement. They frequently demonstrate exophytic growth with partial or complete circumferential encasement of the basilar artery (Fig. 17-26).^44,45 Gliomatosis cerebri, in which the neoplasm is diffusely infiltrating without a discrete mass, is another astrocytoma subtype. The patient may often have nonfocal neurological findings such as headache, decline in mental status, or changes in personality; on MRI, extensive involvement of one or both cerebral hemispheres is observed, with diffusely abnormal T2 hyperintensity involving the subcortical and deep white matter.

FIGURE 17-26 A, Spin echo T1-weighted sagittal image with contrast enhancement showing a diffusely enlarged pons and upper medulla. B, Fast spin echo T2-weighted axial image demonstrating heterogeneous signal characteristics in this brainstem glioma consisting of a mixture of hyperintense and isointense areas.

Enhancement of astrocytomas on gadolinium-enhanced T1-weighted images can be quite variable. In general, higher grade tumors demonstrate more prominent enhancement than do lower grade tumors; however, exceptions do occur.⁴⁶ Therefore, the presence or absence of significant enhancement in a neoplasm, by itself, should not be used to suggest the degree of anaplasia. Interlobar, interhemispheric, or transtentorial macroscopic extension of astrocytomas (or any combination of such extension) and associated vasogenic edema can readily be seen on MRI, especially on T2-weighted and FLAIR images. However, MRI is more limited in the assessment of microscopic tumor infiltration at the tumor margins and in defining the border between the neoplasm and its non-neoplastic vasogenic edema. Contrast-enhanced T1-weighted images can improve the sensitivity of MRI both preoperatively and postoperatively when looking for recurrence. This might be further improved by MRS either in single-voxel or in 2D multivoxel (chemical shift imaging) forms. On MRS, a significantly elevated amplitude of the choline peak relative to that of creatine or NAA suggests a recurrent neoplasm, whereas absence of elevated choline on the background of mobile lipids and lactate would favor radiation necrosis or postsurgical changes, or both (Fig. 17-27).^47–49 Additional new MRI techniques such as perfusion MRI and activation fMRI are now being used for preoperative evaluation of tumor and for surgical planning (Fig. 17-28).^50–53

FIGURE 17-27 A, Single-voxel magnetic resonance (MR) spectra (TR = 2000 msec, TE = 35 msec) from an experimental phantom showing the expected normal appearance of proton (¹H) MR spectroscopy, except for the double lactate peaks. The dominant peak is for N-acetyl aspartate (NAA). Peaks for creatine, choline, and myoinositol are also seen. Refer to the labeled normal spectra from the contralateral hemisphere in Figure 18-22A. B, Abnormal single-voxel MR spectra (TR = 2000 msec, TE = 35 msec) from a patient with a frontal lobe mass showing reduced NAA amplitude and an increased amplitude of choline, consistent with a neoplasm. C, Single-voxel MR spectra (TR = 2000 msec, TE = 135 msec) from a different patient with a neoplasm showing a similar abnormal pattern and double lactate peaks, indicative of the presence of anaerobic metabolism and production of lactate by the neoplasm. D, Color metabolite map for choline from multivoxel MR spectroscopy (TR = 2000 msec, TE = 144 msec) in a patient with a progressive high-grade astrocytoma demonstrating an abnormally increased concentration of choline (displayed in red) in the left inferior frontal lobe.

FIGURE 17-28 A, Fast spin echo T2-weighted axial image at the level of the roof of the lateral ventricle shows a slightly hyperintense mass (arrow) surrounded by prominently hyperintense vasogenic edema (arrowheads). B, Axial cerebral blood volume image from a perfusion magnetic resonance imaging study demonstrating increased blood volume (arrow) within the mass lesion and a halo of decreased perfusion in the surrounding area of edema in a patient with a central nervous system lymphoma.

Oligodendrogliomas, which arise from oligodendrocytes, are tumors of adults and have a peak incidence in the fourth to sixth decades. When made up predominantly or purely of oligodendrocytes, they behave in a benign manner. However, when the tumors are of mixed cell origin and contain both astrocytic and oligodendrocytic components, the astrocytic component often degenerates into a more anaplastic astrocytoma at recurrence. This occurs in about half of oligodendrogliomas. Calcifications are common in oligodendrogliomas, and histologically about 70% contain calcification.⁴⁰ On CT, about 30% to 40% of these tumors contain visible calcification. On MRI, calcifications are usually hypointense on T1- and T2-weighted images, but microcalcifications can demonstrate hyperintensity on T1-weighted images (Fig. 17-29). However, oligodendrogliomas, mixed oligoastrocytomas, and astrocytomas cannot be reliably differentiated by their MRI appearances.⁵⁴

FIGURE 17-29 A, Fast spin echo T2-weighted axial image showing a well-circumscribed, homogeneously hyperintense mass involving the right frontal lobe, insula, and basal ganglia, without significant surrounding vasogenic edema. B, Spin echo T1-weighted contrast-enhanced axial image showing a heterogeneously hypointense mass with no significant enhancement within this oligodendroglioma.

Ependymomas have protean MRI appearances with variable signal intensities on T1- and T2-weighted images, consistent with the variable cellularity and histologic composition found in these tumors. They are typically hypointense on T1- and hyperintense on T2-weighted images, but calcification or cystic or hemorrhagic components will often result in a variable heterogeneous MR signal and enhancement (Fig. 17-30).⁵⁵ Ependymomas can be found both in a supratentorial location typically within the lateral ventricles and in an infratentorial location within the fourth ventricle. They can demonstrate intraventricular or predominantly extraventricular appearances. A subtype called a subependymoma is typically manifested as a periventricular parenchymal mass.

FIGURE 17-30 A, Fast spin echo T2-weighted axial image showing a lobulated hyperintense mass in the fourth ventricle with extension into the region of the right foramen of Luschka. B, Spin echo T1-weighted contrast-enhanced axial image demonstrating heterogeneous enhancement of this ependymoma.

Choroid plexus papillomas and carcinomas can develop within the ventricular system. In children, they typically occur in the atrium of the lateral ventricles, whereas in adults, they are more common within the fourth ventricle. They are generally T1 hypointense and T2 isointense to hyperintense with prominent enhancement (Fig. 17-31). Papillomas are well circumscribed and lobulated in contour, which corresponds to the classic “cauliflower”-like appearance in gross pathology.⁵⁶ Choroid plexus papillomas are frequently associated with communicating hydrocephalus. There is debate about the etiology of the hydrocephalus in these patients, but it probably represents some combination of overproduction of CSF by the tumor and obstruction to normal CSF absorption by arachnoid granulations as a result of cellular debris shred from the papilloma.

FIGURE 17-31 A, Fluid-attenuated inversion recovery axial image showing a homogeneously hyperintense mass in the inferior aspect of the fourth ventricle with mild distortion of the medulla and vermis. B, Spin-echo T1-weighted contrast-enhanced sagittal image demonstrating prominent enhancement of this choroid plexus papilloma.

Meningiomas

Meningiomas are the second most common primary brain neoplasm. They occur at any age and in either sex but are usually found in middle-aged to older women. Meningiomas arise from arachnoid cap cells and are commonly found over the parasagittal cerebral convexity, sphenoid wing, parasellar region, tuberculum sella, olfactory groove, and cerebellopontine angle region. On MRI, meningiomas are usually isointense to gray matter on both T1- and T2-weighted images and can therefore be overlooked on a screening MRI study obtained without contrast enhancement.⁵⁷ Certain histologic subtypes such as syncytial and angioblastic meningiomas demonstrate hyperintensity on T2-weighted images.⁵⁸ When gadolinium contrast is administered, meningiomas are remarkably easy to detect because they exhibit reliably prominent and generally homogeneous enhancement (Fig. 17-32).⁵⁹ They are often associated with an enhanced thickened dura along the lateral margins of the tumor that is known as a “dural tail.”⁶⁰ In some cases these dural tails represent the margins of tumor extension, whereas in most cases they represent reactive dural enhancement without tumor infiltration.⁶¹ A dural tail is not unique to meningiomas and can be observed in other neoplastic and non-neoplastic processes, including exophytic gliomas, dural metastases, lymphoma, and granulomatous infections. Meningiomas sometimes have areas of necrosis, macroscopic calcification, hemorrhage, or cystic changes. Such findings lead to some heterogeneity of signal and enhancement (Fig. 17-33). Most meningiomas have relatively small amounts of vasogenic edema and are frequently found with no significant edema despite their large size. In some cases, however, they are associated with a large amount of vasogenic edema that resembles the edema observed with high-grade gliomas.⁵⁸

FIGURE 17-32 A, Fast spin echo T2-weighted axial image showing a well-circumscribed, homogeneously isointense anterior cranial fossa mass along the cribriform plate. B, Spin echo T1-weighted contrast-enhanced coronal image showing prominent enhancement of this olfactory groove–planum sphenoidale meningioma.

FIGURE 17-33 A, Fast spin echo T2-weighted axial image showing a heterogeneous signal from an extra-axial mass over the left cerebellar hemisphere and extending to the left cerebellopontine angle. The mass is centered posterior to the left internal auditory canal. B, Spin echo T1-weighted contrast-enhanced image showing prominent enhancement consistent with a meningioma.

Pituitary Adenomas

Pituitary adenoma is one of the more common primary neoplasms encountered in adult patients. The tumors may be secretory and “functioning” or nonsecretory, “nonfunctioning” adenomas. Functioning adenomas may arise from any of the secretory cell populations within the adenohypophysis, including cells that secrete prolactin, adrenocorticotropic hormone, growth hormone, or follicle-stimulating hormone/luteinizing hormone. Although histochemical evaluation reliably distinguishes these subtypes, MRI cannot. However, MRI can help identify and lateralize a functioning microadenoma (<1 cm) and thus direct a transsphenoidal surgical approach. In macroadenomas (>1 cm), MRI accurately identifies the sellar, suprasellar, and parasellar extension of the tumor and its relationship to surrounding structures such as the optic chiasm, optic nerves, hypothalamus, cavernous sinuses, and the cavernous and supraclinoid segments of the ICA.

MRI evaluation of a pituitary adenoma requires a tailored examination rather than a routine screening brain examination. Furthermore, the study should be tailored differently for a microadenoma and a macroadenoma. When a microadenoma is imaged, the most important goal of the study is to identify the presence of a microadenoma to correlate with the clinical and laboratory findings. On MRI for a macroadenoma, the primary goal is to define the tumor and its relationship to the surrounding parenchymal and vascular structures.

T1 weighting with gadolinium enhancement is the most important imaging sequence for evaluating pituitary adenomas. The images should include thin (2 to 3 mm) coronal images through the sella turcica with a small field of view. In microadenomas, dynamic imaging of the gland should be done with a series of thin three to four coronal images rapidly obtained over a period of 2 to 3 minutes during bolus intravenous injection of gadolinium.⁶² This technique is used to identify a relatively slowly enhancing microadenoma within a rapidly and homogeneously enhancing normal pituitary parenchyma. This pattern of enhancement results in the microadenoma being delineated from normal parenchyma as a hypoenhancing/hypointense lesion on contrast-enhanced T1-weighted images. Other imaging findings associated with a microadenoma include deviation of the infundibulum away from the side of the gland containing the adenoma, asymmetric convexity of the superior border of the gland, or abnormal contour of the sella turcica floor.⁶³

Additional imaging sequences, including T1-weighted sagittal images with contrast enhancement and T2-weighted coronal images, are valuable to define tumor extension and associated parenchymal changes, if present, of macroadenomas. 3D time-of-flight (TOF) MRA of the cavernous sinus and the circle of Willis can also better define the relationship between the macroadenoma and the ICAs and the M1 and A1 segments of the middle and anterior cerebral arteries, respectively (Fig. 17-34).

FIGURE 17-34 Spin-echo T1-weighted contrast-enhanced coronal image showing a moderately enhancing pituitary macroadenoma arising from the sella turcica with suprasellar extension, invasion of the left cavernous sinus, and a mass effect on the left anterior temporal lobe and inferior basal ganglia.

Metastatic Neoplasms

MRI with gadolinium enhancement is the most sensitive imaging technique for evaluating CNS metastasis. MRI is superior to CT in tissue contrast, ease of multiplanar evaluation, and absence of artifacts, especially in the posterior fossa.⁶⁴ Neoplasms that metastasize to the brain have a wide variety of MRI appearances that depend on their cells of origin, cellular density, and associated pigments (e.g., melanin or hemorrhage). Metastatic lesions are typically localized at the gray-white matter junction in the cerebral hemispheres and are more frequent in the distribution of the anterior than the posterior circulation; this probably reflects differences in blood flow (Fig. 17-35). Typically, the parenchymal lesions are isointense to hypointense on T1-weighted images and isointense to hyperintense on T2-weighted images, with prominent surrounding vasogenic edema relative to lesion size.⁶⁵ These lesions enhance moderately to prominently and are usually multiple. Some metastatic lesions, such as those from thyroid carcinoma, renal cell carcinoma, choriocarcinoma, and melanoma, are often hemorrhagic and demonstrate T1 and T2 signal changes corresponding to subacute and chronic blood breakdown products. Metastases from adenocarcinomas from lung, breast, or colon cancer are occasionally associated with neoplastic cysts (Fig. 17-36). In meningeal carcinomatosis, the metastases involve the meninges and appear as a diffusely thickened enhancement of the pia or pachymeninges (Fig. 17-37) or as multiple, dural-based nodular-enhancing lesions.⁶⁶

FIGURE 17-35 Spin-echo T1-weighted contrast-enhanced coronal image demonstrating multiple ring-enhancing lung carcinoma metastases, most of which are located at the gray matter–white matter junction.

FIGURE 17-36 A, Fast spin echo T2-weighted axial image showing a cystic mass in the left cerebellar hemisphere with distortion of the adjacent medulla. B, Spin-echo T1-weighted contrast-enhanced sagittal image showing peripheral enhancement of this cystic metastasis from breast carcinoma.

FIGURE 17-37 A, Fluid-attenuated inversion recovery axial image at the level of the centrum semiovale revealing multiple areas of abnormal hyperintensity within the sulci of both cerebral hemispheres. B, Spin-echo T1-weighted contrast-enhanced fat saturation image showing abnormal enhancement within these sulci in a patient with meningeal carcinomatosis from breast carcinoma.

Schwannomas

Intracranial schwannomas arise from the Schwann cells that envelop the cranial nerves as they exit the intracranial compartment through various canals and foramina. They are more often found within the posterior fossa. The most common intracranial schwannomas arise from the vestibular divisions of the eighth cranial nerve. They are typically located within the internal auditory canal (IAC) with extension through the porus acusticus into the cerebellopontine angle.⁶⁷ Depending on the size of the neoplasm, a variable degree of brainstem distortion or compression is observed together with minimal, if any, intra-axial brain edema. Tumor growth may cause bone erosion and widen the IAC and the porus acusticus. Vestibular schwannomas are usually isointense to hypointense to brain tissue on T1-weighted images and hyperintense to brain on T2-weighted images with prominent enhancement after intravenous gadolinium injection. When they are small (e.g., intracanalicular tumor), vestibular schwannomas typically enhance homogeneously (Fig. 17-38). However, with an increase in tumor size, cystic and necrotic foci can develop, which then contribute to more heterogeneous MRI signal characteristics and enhancement patterns (Fig. 17-39).

FIGURE 17-38 A, Spin echo T1-weighted non–contrast-enhanced axial image showing a small isointense mass with ill-defined borders in the right internal auditory canal (arrow). B, This high-resolution, fast spin echo T2-weighted image better delineates a small isointense intracanalicular mass. C, Spin echo T1-weighted contrast-enhanced axial image demonstrating prominent homogeneous enhancement of this small intracanalicular vestibular schwannoma.

FIGURE 17-39 A, Fast spin echo T2-weighted axial image showing heterogeneous signal from a mass in the left cerebellopontine angle at the level of the left internal auditory canal (IAC). There is a significant mass effect on the brainstem, left cerebellar hemisphere, and fourth ventricle. B, Spin echo T1-weighted contrast-enhanced axial image showing the heterogeneous enhancement pattern of this large vestibular schwannoma, with minimal enhancement in the left IAC.

Vestibular schwannomas are the most common mass lesions in the cerebellopontine angle and represent about 75% of all masses in this location. The next most common mass in the cerebellopontine angle is a meningioma. Meningiomas tend to exhibit more intermediate T2 signal characteristics than do vestibular schwannomas and infrequently extend into the porus acusticus. Although the epicenter of a vestibular schwannoma is at the porus acusticus, a meningioma in this location is located eccentric to the porus acousticus.⁶⁸

Other intracranial schwannomas include those from the trigeminal (cranial nerve V), facial (VII), glossopharyngeal (IX), vagus (X), spinal accessory (XI), and hypoglossal (XII) nerves. Of these, trigeminal nerve schwannomas are the most common. Facial nerve schwannomas often appear as a lytic, enhancing mass in the petrous temporal bone in the region of the geniculate ganglion (Fig. 17-40).⁶⁹ The origin of a schwannoma that arises from cranial nerve IX, X, XI, or XII in the caudal aspect of the posterior fossa is difficult to ascertain on preoperative MRI and may be obvious only at the time of surgery.

FIGURE 17-40 A, Fast spin echo T2-weighted axial image showing a hyperintense mass in the right petrous apex. B, Spin echo T1-weighted contrast-enhanced axial image demonstrating a homogeneously enhancing right facial nerve schwannoma at the level of the geniculate ganglion.

Primitive Neuroectodermal Tumors

Primitive neuroectodermal tumors (PNETs) were first defined to describe large, solid hemispheric neoplasms made up of undifferentiated cells, often found in infants and young children, that could not be designated, because of their locations, as a retinoblastoma, medulloblastoma, pineoblastoma, or ependymoblastoma. However, most pathologists are of the opinion that these various neoplasms are similar in their histopathology. Therefore, the designation PNET is used to describe all these undifferentiated, primitive neoplasms found in children and young adults.⁴⁰ In the posterior fossa, PNETs arise from the germinal matrices surrounding the fourth ventricle in children and along the anterolateral cerebellar hemisphere in young adults. They can also occur in the supratentorial compartment as intraventricular or intra-axial masses. PNETs are made up of small round cells and are hypercellular with a high nuclear-to-cytoplasm ratio. These characteristics contribute to their MRI signal characteristics, which are slightly hypointense to isointense to brain tissue on T1- and isointense on T2-weighted images. They demonstrate moderate to prominent enhancement on contrast-enhanced T1-weighted images. PNETs are typically homogeneous solid masses but can have cystic or necrotic components that result in more heterogeneous signal and enhancement characteristics (Fig. 17-41).⁷⁰

FIGURE 17-41 A, Fast spin echo T2-weighted axial image showing a heterogeneous posterior fossa mass with a cystic hyperintense anterior component and a solid isointense posterior component. There is nearly complete effacement of the fourth ventricle by this mass. B, Spin echo T1-weighted contrast-enhanced axial image showing a central focus of prominent enhancement with variable enhancement in the remainder of this primitive neuroectodermal tumor.

Infections

MRI evaluation for meningitis is limited by its marginal sensitivity. With the advent of the FLAIR pulse sequence and especially with gadolinium-enhanced FLAIR studies, increased sensitivity for meningeal diseases is reported, but further large-scale prospective studies are required. A normal brain MRI study does not exclude meningitis. Therefore, the diagnosis of meningitis should still be made by chemical and bacteriologic examination of CSF. However, in a patient with high suspicion or a proven meningeal infection, MRI is valuable and can help ascertain the degree of involvement and the presence of any parenchymal involvement (i.e., encephalitis, infarction, or abscess formation). MRI also can help monitor a patient’s response to therapy if an objective parameter, in addition to the clinical findings, is required.⁷¹

Viral meningitis, when visible on MRI, is typically manifested as diffuse meningeal enhancement on gadolinium-enhanced T1-weighted images (Fig. 17-42). When encephalitis is present (e.g., herpes encephalitis), the involved parenchyma reflects the presence of cytotoxic and vasogenic edema with hypointense T1 and hyperintense T2 signal characteristics involving both gray and white matter (Fig. 17-43). Effacement of local cortical sulci, hemorrhage, and prominent leptomeningeal enhancement may also be present.⁷²

FIGURE 17-42 Spin echo T1-weighted contrast-enhanced coronal image demonstrating abnormal, thick meningeal enhancement in this patient with viral meningitis.

FIGURE 17-43 A, Spin echo T1-weighted non–contrast-enhanced axial image illustrating a lesion in the left anterior temporal lobe with a hypointense medial component and a hyperintense lateral component. B, Fast spin echo T2-weighted axial image showing the abnormal hyperintensity of this lesion, which represents a focus of encephalitis from herpes infection. The T1 hyperintense area represents subacute parenchymal hemorrhage.

Bacterial meningitis is often associated with prominent diffuse meningeal enhancement. It is typically hematogenous, but it also results from direct extension of infections in the adjacent paranasal sinuses or mastoid air cells. Complications of bacterial meningitis include subdural empyema, infarction, and parenchymal abscess. An abscess is seen on T1-weighted images as a mass with a central nonenhancing or slightly enhancing T1 hypointense cavity with an enhancing wall of variable and irregular thickness that is surrounded by T1 hypointense and T2 hyperintense edema. The central cavity is typically hyperintense on T2-weighted images, but it can be isointense or hypointense, depending on its contents.⁷³ However, these MRI appearances are similar to those observed with a necrotic neoplasm. MRI findings can be crucial in distinguishing an abscess from a necrotic neoplasm when other clinical or laboratory data are lacking. The enhancing wall of an abscess cavity is typically thinner along its medial/deeper aspect and thicker along its lateral/superficial aspect. The extent of associated edema may be smaller with an abscess, whereas a necrotic brain neoplasm such as glioblastoma multiforme usually has a large area of surrounding vasogenic edema.⁷⁴ Newer MRI techniques such as MRS and diffusion-weighted imaging (DWI) can help differentiate abscesses from necrotic neoplasm. In MRS, spectroscopic evaluation of an abscess reveals several amino acid signatures and abundant lactate and mobile lipids. These, however, are also observed in necrotic neoplasms. On DWI, an abscess cavity typically has a restricted diffusion pattern with a low apparent diffusion coefficient (ADC) and hyperintense signal on diffusion-weighted images (Fig. 17-44).^75,76 By contrast, a cystic/necrotic neoplasm has a relatively increased ADC with hypointense signal on diffusion-weighted images.

FIGURE 17-44 A, Spin echo T1-weighted contrast-enhanced axial image showing a ring-enhancing left parietal mass with surrounding hypointense edema. B, Axial diffusion-weighted image revealing abnormal hyperintensity within the nonenhancing central cavity, which indicates a marked restriction of water diffusion in this abscess cavity.

MRI can be used to image other infections such as tuberculosis, fungal infections, and cysticercosis. Tubercular meningitis usually involves the basal meninges and typically extends into the Virchow-Robin perivascular spaces of the basal ganglia, which may cause lacunar infarctions (Fig. 17-45).^77–79 Fungal infections involve the brain by direct extension from contiguous structures such as the paranasal sinuses, orbits, and mastoid or by a hematogenous route. These infections are usually found in diabetic patients or immunocompromised individuals. Meningitis and encephalitis with parenchymal abscesses are possible. In Aspergillus infection, the fungus has a propensity for vascular invasion, which results in hemorrhagic transformation of encephalitic foci.^80,81 Cysticercosis is an infection that results from infestation with a pork tapeworm. When these parasites die within the brain, parenchymal reaction to the dying parasites contributes to the edema, enhancement, and calcification seen on imaging. Cysticercosis appears as small, parenchymal lesions with punctate calcifications, intraventricular masses, or multiple lobulated T1 hypointense masses that expand the subarachnoid spaces (racemose form) (Fig. 17-46).^82,83

FIGURE 17-45 A, Spin echo T1-weighted contrast-enhanced axial image demonstrating diffuse basal meningeal enhancement with edema in the right temporal lobe in this patient with tubercular meningitis. B, Spin echo T1-weighted contrast-enhanced images showing multiple enhancing parenchymal nodules within the brain and thoracic spinal cord (C). They represent multiple tuberculomas associated with miliary tuberculosis infection.

FIGURE 17-46 Fast spin echo T2-weighted axial image (A) and spin echo T1-weighted contrast-enhanced axial image (B) revealing multiple cystic parenchymal lesions that are T2 hyperintense and T1 hypointense at the gray matter–white matter junction and in the periventricular region. These lesions are associated with central nervous system cysticercosis.

Stroke and Vascular Diseases

Visualization of an acute infarction on MRI depends on several factors, including time after the ictus, stroke size, and location. Of these, time is the single most important factor. The high tissue contrast of MRI has decreased the time between the ictus and when an acute infarction can first be identified on imaging. Acute infarction appears as a hyperintense T2 lesion involving the cortical, white matter, or deep gray matter structures (or any combination). Larger cortical infarctions have associated effacement of adjacent sulci from cytotoxic edema. Abnormal hyperintensity in cortical vessels within these sulci may be the first signs of an acute infarction, even before any significant changes in T2-weighted images are apparent. Slow vascular transit time within an infarction results in loss of the normal flow void and increased signal within these vessels. Absence of a normal flow void is occasionally observed in the setting of a large infarction involving the internal carotid or anterior, middle, or posterior cerebral artery territory. Acute infarction changes on MRI are usually present within 24 hours of the ictus and are often evident between 6 and 12 hours after the ictus. Cytotoxic edema increases during the first 1 to 2 weeks and resolves within 1 month. The infarcted area, in the chronic stage, exhibits local volume loss with compensatory enlargement of adjacent sulci, cisterns, and ventricles along with a hyperintense signal on T2-weighted images. If hemorrhage was present during the acute or subacute stages, T1 and T2 hypointensity, representing hemosiderin and ferritin, are present in the chronic stages.

Echo planar DWI has even further decreased the time required for MRI visualization of an acute stroke.⁸⁴ In animal models, DWI can demonstrate large vascular territory infarctions (e.g., ICA or MCA) within minutes after vessel occlusion.⁸⁵ In humans, acute infarctions can be demonstrated on DWI within 30 minutes to 1 hour. This implies that most acute infarctions should be visible on DWI by the time that a patient arrives at the hospital after the onset of stroke (Fig. 17-47).⁸⁶ Once present on MRI, the hyperintense infarction on DWI persists for approximately 2 weeks, depending on size and location and the diffusion sensitivity factor (i.e., b value) that is used to obtain the diffusion-weighted images. After this period, subtle hyperintensity may remain and corresponds to the areas of abnormal hyperintensity on T2-weighted and FLAIR images. The degree of such hyperintensity on DWI is much less than that observed with an acute infarction. This phenomenon is known as the “T2 shine-through” effect and should not be confused with an acute infarction.⁸⁷

FIGURE 17-47 Fluid-attenuated inversion recovery axial image (A) showing abnormal hyperintensity in the left frontal lobe, temporal lobe, and insula, with corresponding abnormal hyperintensity on the diffusion-weighted image (B), in this patient with an acute middle cerebral artery infarction.

Perfusion MRI, in the form of dynamic susceptibility imaging, shows promise in evaluation of cerebral perfusion, expressed as relative CBV or MTT. Again, using an echo planar technique, several hundred images of the brain can be obtained in less than 2 minutes. These raw images are then postprocessed to obtain a “map” of relative CBV or MTT, which is then used to qualitatively and quantitatively assess cerebral perfusion (Figs. 17-48 and 17-49). Along with diffusion-weighted images, these perfusion images can be used to determine the extent of ischemic brain that is at risk for infarction. The difference between the area of abnormally low perfusion and the area of abnormally low diffusion coefficients represents the ischemic penumbra that is at risk for infarction but still may be salvageable if treated.^88,89

FIGURE 17-48 A, Images from a multiple-level relative cerebral blood volume (rCBV) map in a patient with left middle cerebral artery ischemic symptoms demonstrate decreased perfusion (dark blue) in the left basal ganglia and superior temporal lobe and within the deep white matter (black) of the left frontal and parietal lobes. B, Mean transit time (MTT) maps at the same levels demonstrate increased MTT (yellow and green) in the same areas.

FIGURE 17-49 A, Normal relative cerebral blood volume (rCBV) map from a perfusion magnetic resonance imaging study at the level of the basal ganglia. B, rCBV map from a patient with an infarction in the left parietal and occipital lobes demonstrating reduced perfusion (dark blue) within the left parietal and occipital lobes.

MRS can also play a role in acute stroke evaluation. Evidence of anaerobic metabolism (lactate production and decreased NAA) can be detected with ¹H-MRS when ischemia and infarction are present. Single-voxel or 2D chemical shift imaging (multivoxel) spectroscopy can therefore define the extent of ischemic tissue that is producing significant amounts of lactate. When 2D spectroscopy is performed, a map of areas with lactate production or areas with decreased NAA can be displayed and compared with maps of abnormal diffusion or perfusion to further delineate the cerebral tissue at risk for infarction.⁹⁰ MRS may also help predict clinical outcomes in young children with hypoxic ischemic CNS injury.⁹¹

Nonemergency MR evaluation of patients with symptoms of intracranial or extracranial ischemic vascular disease usually begins with a routine brain MRI study that includes T2-weighted and FLAIR images for maximal sensitivity to detect subacute and chronic evidence of cerebral ischemic disease. MRA of the brain to evaluate the major intracranial arteries and their proximal branches is now a well-established technique. MRA typically uses a 3D TOF technique with added magnetization transfer pulse to improve the visualization of distal branches (Fig. 17-50). An MR venogram to exclude major venous thrombosis can be performed with either 3D TOF or phase-contrast techniques, which can easily identify the major superficial and deep venous sinuses and veins (Fig. 17-51). MRA has acceptable accuracy in evaluation of the intracranial internal carotid, vertebral, and basilar arteries, as well as the proximal branches of the anterior, middle, and posterior cerebral arteries in the setting of significant atherosclerotic stenoses, vasospasm, or vasculitis (Fig. 17-52).⁹² However, MRA is limited in its evaluation of medium-sized and small arteries. When such a disease process is clinically suspected and suggested by a routine MRI study (i.e., abnormal lesions on T2-weighted or FLAIR images), conventional angiography is often warranted. Evaluation of the carotid and vertebral arteries in the neck can be performed with 2D or 3D TOF or phase-contrast MR techniques without the use of intravenous gadolinium contrast. The accuracy of these techniques in comparison to conventional contrast-enhanced angiography is well described.^93,94 A newer technique using a 3D spoiled gradient recalled (SPGR) pulse sequence obtained after rapid bolus intravenous injection of gadolinium can improve the accuracy of noninvasive MR vascular evaluation (Fig. 17-53).⁹⁵ Similar to brain CTA, MRA is increasingly being used as an initial noninvasive imaging study for intracranial aneurysms or as a follow-up imaging study for untreated or treated intracranial aneurysms.^95,96 MRA or CTA of a previously coiled or clipped aneurysm can be challenging because of inherent MR or CT artifacts generated by coils or clips. It is difficult to exclude a small recanalization or residual aneurysm on CTA because of a large amount of artifact from an aneurysm clip or coils. However, gadolinium bolus MRA of the brain may demonstrate recanalization of an intracranial aneurysm previously treated by endovascular coil embolization better than conventional TOF MRA without contrast enhancement (Fig. 17-54).^97,98

FIGURE 17-50 A, Postprocessed image from three-dimensional time-of-flight magnetic resonance angiography of the brain demonstrating major intracranial arteries, including the A2, M2, and P2 branches of the anterior, middle, and posterior cerebral arteries. B, Isolated “cutout” view of the right internal carotid artery and its intracranial branches. C, Isolated view of the posterior circulation.

FIGURE 17-51 A, Postprocessed sagittal image from a two-dimensional time-of-flight magnetic resonance venogram (MRV) showing a normal appearance of the sagittal sinus but absence of signal from the deep internal cerebral veins, vein of Galen, and straight sinus, indicative of thrombosis of these venous structures. B, Postprocessed axial image from the same MRV showing thrombosis of the left transverse and sigmoid sinuses.

FIGURE 17-52 A, Three-dimensional time-of-flight brain magnetic resonance angiogram (MRA) demonstrating reduced flow signal within the distal intracranial segment of the left internal carotid artery (ICA) and proximal left middle cerebral artery (MCA) (compare with the right side). B, Magnified MRA view of the left ICA and MCA demonstrating thrombus within the terminal left supraclinoid ICA, proximal A1 segment of the anterior cerebral artery, and M1 segment of the MCA, with significantly diminished flow in the left MCA distribution distal to the thrombus.

FIGURE 17-53 A, Three-dimensional gadolinium bolus neck magnetic resonance angiogram (MRA) demonstrating the major craniocervical arteries from the aortic arch to the proximal intracranial circulation. B, Isolated view of the left extracranial and intracranial carotid artery system. C, In a different patient, a gadolinium bolus neck MRA demonstrates a focal stenosis at the origin of the right internal carotid artery.

FIGURE 17-54 A, Axial image from a brain CTA at the level of the anterior clinoid process demonstrating prominent streak artifacts from previous endovascular coil embolization of a left parophthalmic internal carotid artery (ICA) aneurysm (arrow). B, A two-dimensional coronal reformatted image from the CTA data set is compromised by the artifacts, with poor visualization of the region of the aneurysm neck (arrow). C, Coronal image from a gadolinium bolus brain magnetic resonance angiogram (MRA) demonstrating small recurrent filling of the left ICA aneurysm (arrow). D, Three-dimensional reconstruction image from the MRA data set with a cutout view of the left ICA revealing definite small recanalization of the previously coiled aneurysm (arrow).

Trauma

In acute CNS trauma, evaluation for a fracture or hemorrhage is still best done with a non–contrast-enhanced CT scan. However, because of its superior tissue contrast, MRI may better define small parenchymal abnormalities on T2-weighed, FLAIR, and gradient-recalled echo (GRE) images that may not be visible on CT. GRE is the most sensitive pulse sequence for acute/subacute blood breakdown products.⁹⁹ MRI also appears to be superior to CT in evaluating posterior fossa pathology because the usual posterior fossa beam-hardening artifacts typically observed on CT are absent. MRI is also useful in the evaluation of small cortical contusions and subdural and epidural hematomas because of its higher tissue contrast and the ease of multiplanar acquisition with MRI. Small subacute SDHs, which are often isodense and therefore subtle on CT, are typically T1 and T2 hyperintense on MRI and thus more conspicuous. The MRI appearances of subdural and epidural hematomas and hygromas are variable and depend on the age and history (i.e., repeat hemorrhages). In the acute and subacute stages, T1 and T2 signal characteristics are similar to those observed with parenchymal hematomas. They are T1 isointense to hypointense and T2 hypointense in the acute stage because of intracellular deoxyhemoglobin and methemoglobin and become T1 hyperintense and T2 hyperintense in the subacute stage from extracellular methemoglobin. In the chronic stage, parenchymal hemorrhages are hypointense on T1- and markedly hypointense on T2-weighted images as a result of susceptibility effects from intracellular ferritin and hemosiderin, which are mostly present in the interstitium and within macrophages. However, chronic SDHs (hygromas) are T1 hypointense and T2 hyperintense in signal, similar to those of CSF. This reflects the low-density fluid collection that is typically observed on CT. Some authors suggest that MRI with a FLAIR pulse sequence has high sensitivity for acute SAH, but significant debate remains about its accuracy in the presence of frequently observed CSF flow–related artifacts associated with this pulse sequence, especially in the basal cisterns.^100–102

MRI is especially helpful in the evaluation of patients with diffuse axonal injury (DAI), which usually results from a high-speed acceleration and deceleration injury that causes mechanical stress on the brain and thus results in “shear” injuries.¹⁰³ The multiple, small and subtle shear injuries at the gray-white matter border, along the corpus callosum, and within the brainstem, typical of DAI, are much better evaluated with MRI than CT. The GRE pulse sequence, which is highly sensitive to areas of changes in susceptibility, is valuable when diffuse small parenchymal hemorrhages in the setting of trauma are evaluated. On T2-weighted and FLAIR images, these small superficial gray-white matter border lesions and deep pericallosal, deep gray matter, and upper brainstem lesions are hyperintense. If hemorrhagic, these lesions are also hypointense on GRE images because of changes in susceptibility associated with acute blood breakdown products such as deoxyhemoglobin and intracellular methemoglobin.¹⁰⁴ In the chronic stage, hemosiderin and ferritin result in hypointensity on T2-weighted and GRE images.¹⁰⁵

Brain MRI can be useful when a patient with an acute head injury has minimal CT abnormalities but has profound neurological deficits. In addition, MRI may be helpful when an objective measure of extensive brain injury, particularly brainstem injury, is required before a patient’s prognosis is discussed with the family. Newer MRI techniques such as DWI, perfusion MRI, and MRS may play additional roles in trauma patients to detect areas of ischemia and infarction.¹⁰⁶

Acute traumatic vascular injury at the skull base that involves the ICA or vertebral artery can be quickly and noninvasively evaluated with MRI. MRI is highly sensitive to high-grade vessel stenosis or occlusion from traumatic dissection, with demonstration of the absence of normal flow void in these vessels. Additional T1-weighted images with fat saturation in the neck and at the skull base readily identify the false lumen containing thrombus with subacute methemoglobin. 2D/3D MRA of the neck and the skull base can add additional information and display the relevant arterial tree from the aortic arch to the brain. Although conventional catheter angiography still remains the “gold standard,” MRA in a posttraumatic, possibly unstable patient can help guide clinical management and direct further conventional angiographic evaluation once the patient is stable.

Vascular Malformations

Intracerebral vascular malformations include four distinct types: AVM, cavernous malformation (cavernous angioma or hemangioma), developmental venous anomaly (DVA, previously known as venous malformation or venous angioma), and capillary telangiectasia. All except capillary telangiectasia are typically visible on MRI. An AVM is a vascular malformation that consists of a nidus that lacks the normal capillary network with ensuing direct communication between abnormally enlarged thick-walled feeding arterial channels and thin-walled draining venous channels. The vessels often lack normal smooth muscle and elastic lamina layers. AVMs may often be associated with aneurysms of feeding arteries or the draining veins (venous varix). There is no normal parenchyma intermixed with an AVM, although some gliosis may be observed. On MRI, an AVM appears as a complex tangle of abnormally enlarged vessels that demonstrate flow voids because of a high-flow state. Secondary findings such as subacute and chronic blood breakdown products (methemoglobin, ferritin, and hemosiderin) can be found with corresponding areas of T2 and GRE hypointensity.¹⁰⁷ Similar small T2 hypointensity is also observed in the setting of calcifications associated with an AVM. Surrounding areas of edema, ischemia, or encephalomalacia may also be present.¹⁰⁸ These areas are visible on T2-weighted and FLAIR images as hyperintensity with or without loss of volume in the case of chronic atrophy. Thin, linear cortical T1 hyperintensity may also be observed if cortical laminar necrosis from chronic ischemia as a result of a physiologic steal phenomenon occurs in the parenchyma adjacent to the AVM. AVMs can also be further visualized with 2D or 3D MRA using either phase-contrast or TOF techniques (Fig. 17-55).¹⁰⁹ These MRI techniques can be used to diagnose AVMs and help in preoperative localization with 3D intraoperative guidance software. MRI is also useful in monitoring smaller AVMs treated by stereotactic radiosurgery (linear accelerator or Gamma Knife techniques).^110,111

FIGURE 17-55 Spin echo T1-weighted contrast-enhanced coronal image (A), fast spin echo T2-weighted axial image (B), and postprocessed axial image from three-dimensional time-of-flight magnetic resonance angiography (C) illustrate a large left temporal lobe arteriovenous malformation with multiple flow voids within a large nidus (arrowheads) and prominently dilated feeding arteries (large arrow) and draining veins (small arrow).

Cavernous malformations are slow-flow vascular malformations made up of abnormally large venous channels without associated abnormally enlarged feeding arteries or draining veins. They become symptomatic and also visible on MRI because they frequently bleed.¹¹² Typically, cavernous malformations are discrete parenchymal masses with heterogeneous hypointensity and hyperintensity on T1- and T2-weighted images as a result of blood breakdown products of various ages. When recent hemorrhage is present, reactive T2 hyperintense surrounding edema may be seen with associated local mass effect.¹¹³ Multiple cavernous malformations are found in patients with vascular malformation syndromes. Because MRI, among all the imaging techniques, including conventional angiography, best visualizes these cavernous malformations, it is the study of choice in the initial evaluation and follow-up of patients with cavernomas (Fig. 17-56).

FIGURE 17-56 A, Fast spin echo T2-weighted axial image showing a mass with heterogeneous signal in the left frontoparietal lobes, central hyperintensity, and a peripheral rim of hypointensity with associated surrounding hyperintensity. This is a large cavernous malformation (arrows) with evidence of recurrent hemorrhages—blood breakdown products of different age surrounded by mild edema (arrowheads), suggestive of a recent episode of hemorrhage. B, Fast spin echo T2-weighted axial image at the level of the basal ganglia in the same patient illustrating a focus of abnormal hypointensity, representing hemosiderin, adjacent to the left frontal horn (arrow). This represents a focus of old hemorrhage from a smaller cavernous malformation.

A DVA (as mentioned before, previously known as venous malformation or venous angioma) is not a pathologic vascular malformation. Instead, a DVA is a normal variant in which several prominent deep parenchymal veins drain a normal part of the brain and then drain into an unusually large superficial or deep draining vein. The angiographic and MRI appearance of these malformations has been described in the literature as resembling the “head of Medusa,” a Greek mythologic figure with a head of “hair” made of multiple live serpents. Because this is a benign entity and requires no intervention, its recognition on imaging studies is essential. DVAs have a typical imaging appearance on MRI that consists of several slightly prominent vessels with flow voids on T2-weighted images that drain into a larger draining vein.¹¹⁴ They are best visualized on contrast-enhanced T1-weighted images (Fig. 17-57). DVAs may sometimes coexist with cavernous malformations. If a hemorrhage is present in the vicinity of a DVA, the cause of hemorrhage is most likely an occult cavernous malformation neighboring the DVA.

FIGURE 17-57 A, Fast spin echo T2-weighted axial image illustrating branching flow voids in the right brachium pontis region with extension into the adjacent cisternal space. B, Spin echo T1-weighted contrast-enhanced coronal image demonstrating a typical branching pattern of dilated veins in a venous malformation next to the fourth ventricle (arrows).

Capillary telangiectasia is a histopathologic diagnosis rather than a radiologic one. It is rarely diagnosed preoperatively. If capillary telangiectasia is visible on MRI, it appears as a nonspecific focus of T2 hyperintensity, which may be associated with subtle enhancement on contrast-enhanced T1-weighted images. Multiple capillary telangiectases are found in patients with ataxia-telangiectasia syndrome, which affects both the CNS and other viscera.¹¹⁵

Seizure and Epilepsy

Imaging of a patient with a first-time seizure usually begins with a head CT without contrast, followed by a contrast-enhanced study if necessary. Causes of seizures such as stroke, hemorrhage, neoplasm, abscess, or benign cysts can often be adequately evaluated with CT. Routine brain MRI has higher sensitivity than CT for subtle parenchymal lesions and meningeal diseases. However, in epilepsy, where a recurrent seizure disorder is present, a more tailored MRI study is required to exclude a structural substrate for epilepsy, including mesial temporal sclerosis, subtle cortical and subcortical neoplasms, or migration anomalies (heterotopias). These lesions may not be apparent on routine brain MRI examination because they are often quite small and have signal characteristics similar to adjacent normal brain tissue on both T1- and T2-weighted images.

Several different MRI protocols have been developed over the years to evaluate epilepsy patients. These protocols usually include a type of high-resolution, 3D, and heavily T1-weighted acquisition of the whole brain. An SPGR sequence is one such pulse sequence that can be obtained at 1-mm intervals in the axial, coronal, and sagittal planes. Small slice thickness along with its heavily T1-weighted parameters accentuates the signal difference between gray and white matter and thus improves the detection of small focal heterotopias and migration anomalies.^116,117 MRI acquisition with the use of phased-array surface coils over a part of the head instead of the usual head coil used for brain MRI may also improve resolution and sensitivity. These surface coils have an improved signal-to-noise ratio that is usually severalfold greater than that obtained with a routine head coil and have the flexibility to be placed directly over the part of the brain that is suspected of containing the epileptic focus. They are useful in the evaluation of small cortical or subcortical lesions, which may not otherwise be visible on routine brain MRI. The advantages of surface coils diminish for deeper structures because the signal decreases as the distance between the coil and the structure increases. Nonetheless, phased-array surface coils have been used successfully to evaluate the mesial temporal lobes in patients with complex partial seizures referable to the temporal lobes. In many of these patients, MRI correlates of mesial temporal sclerosis, such as hippocampal atrophy, abnormal T2 hippocampal hyperintensity, and disruption of the hippocampal internal architecture, are observed.^118–120 Such imaging of the mesial temporal lobes is best done in a coronal or modified coronal plane perpendicular to the anteroposterior axis of the hippocampus at 1- to 20-mm intervals using T1, T2, and FLAIR images (Fig. 17-58).

FIGURE 17-58 A, This high-resolution, fast spin echo T2-weighted coronal image obtained with phased-array surface coils demonstrates subtle abnormal hyperintensity in the right hippocampus (arrow) in a patient with temporal lobe epilepsy. B, A coronal image through the body of the right hippocampus demonstrates prominent asymmetric atrophy of the right hippocampus (arrow) in comparison to the contralateral left hippocampus.

Tailored MRI techniques are valuable in the preoperative imaging evaluation of epileptic patients. In particular, these techniques are useful in the preoperative evaluation and lateralization of the seizure focus in patients with temporal lobe epilepsy. Various MRI parameters can be used to diagnose the presence of mesial temporal sclerosis in patients with temporal lobe epilepsy, including qualitative techniques such as visual evaluation for asymmetrically and abnormally increased signal on T2 and FLAIR coronal images, decreased hippocampal volume, and larger than normal size of the temporal horn of the lateral ventricle within the ipsilateral temporal lobe. Quantitative measurements of these changes can be studied with 3D hippocampal volumetry to quantify the hippocampal atrophy and T2 relaxometry to quantify the change in T2 signal. Such MRI lateralization, when possible, has significant prognostic value for patients who undergo surgery to treat medically intractable temporal lobe epilepsy. Patients with a lateralizing MRI abnormality have a significantly greater seizure-free rate after hippocampal resection and partial temporal lobectomy than do those who do not demonstrate such a lateralizing abnormality on preoperative MRI.^121,122 MRS can also be used to study the mesial temporal lobes, including the hippocampus, in temporal lobe epilepsy. Single-voxel and 2D chemical shift imaging techniques are used to obtain proton MR spectra from the hippocampi and surrounding mesial temporal lobe structures. In MRS, neuronal concentration and integrity are ascertained by measuring the concentration of the neuronal marker metabolite NAA with an absolute quantitative technique or, more commonly, by a ratio that compares NAA with a reference metabolite such as creatine. A high accuracy rate in lateralization of the diseased hippocampus in patients with temporal lobe epilepsy is reported with the use of single-voxel or chemical shift imaging proton MRS.^47,123–125

More recently, activation fMRI and diffusion tensor imaging with 2D and 3D white matter tractography are also being used as a part of preoperative evaluation and surgical planning for patients with epilepsy, especially temporal lobe epilepsy. Molecular imaging using single-photon emission CT (SPECT)- and PET-visible receptor ligands can also help localize areas of cortical malformations and epileptogenic foci.^126–129