Plain Films

K. Sartor

Basic Principles and Applications

The term “plain films” refers primarily to plain radiographs, in this case radiographs of the skull that have been obtained using conventional or digital technology without the prior administration of contrast material.

Metallic foreign bodies have particularly high radiographic contrast, for example:

• Vascular clips

• Bioimplants

• Fragments of projectiles.

Air in the integumental soft tissues or inside the skull is relatively easy to identify because it appears darker than its surroundings. Today, however, the main purpose of plain skull films is to detect abnormalities in the cranial bones themselves. Their primary role is to provide a basic diagnostic impression.

If more detailed information is needed and especially if brain pathology is suspected, computed tomography (CT) and magnetic resonance imaging (MRI) are more rewarding than plain skull films. MRI can yield particularly high softtissue contrast, even in marrow-containing cancellous bone. Plain films have shown relatively low sensitivity to soft-tissue changes, including changes in paranasal sinuses and mastoid/middle ear air spaces, whose high natural contrast is decreased when inflammatory mucosal thickening is present.

Survey Films and Special Views

Survey Films

Fig. 1.1 a, b Plain skull films (schematic): normal PA and lateral views (after Piepgras).

a PA projection.

1 Outer table

2 Diploë

3 Inner table

4 Pacchionian granulations

5 Coronal suture

6 Lambdoid suture

7 Sagittal suture

8 Canal of diploic vein

9 Pineal gland (projected into the frontal sinus

10 Mastoid process

11 Mandibular condyle

12 Petrous part of temporal bone (superior margin)

13 Lateral border of anterior cranial fossa

14 Innominate line = anterior lateral wall of middle cranial fossa projected into orbit

15 Foramen rotundum

16 Superior orbital fissure

17 Lesser wing of sphenoid

18 Anterior clinoid process

19 Frontal sinus

20 Septum of frontal sinus

21 Crista galli

22 Sphenoid sinus (with superimposed ethmoid cells)

23 Ethmoid cells

24 Bony nasal septum

25 Orbit

26 Squamous suture

27 Zygomatic arch

28 Maxillary sinus

29 Inferior border of floor of posterior cranial fossa

30 Occipital condyle

31 Atlanto-occipital joint

32 Inferior turbinate

33 Nasal cavities

34 Dens axis

35 Anterior nasal spine

36 Posterior arch of atlas

37 Transverse process of atlas

38 Lateral mass of atlas

39 Atlantoaxial joint

40 Spinous process of axis

41 Body of axis (projected over the maxilla)

42 Mandibular angle

43 Mandibular canal

44 Mental foramen

45 Internal acoustic meatus

b Lateral projection.

1 Sella turcica

2 Dorsum sellae

3 Anterior clinoid process

4 Tuberculum sellae

5 Planum sphenoidale

6 Clivus

7 Petrous part of temporal bone (petrous pyramid)

8 Internal acoustic meatus

9 Squamous suture

10 Mastoid process (both sides superimposed)

11 Occipital condyle

12 Occipital squama

13 Internal occipital crest

14 Internal occipital protuberance

15 External occipital protuberance

16 Lambdoid suture

17 Parietal bone

18 Frontal bone

19 Lambda

20 Bregma

21 Outer table

22 Diploë

23 Inner table

24 Coronal suture

25 Pineal gland

26 Groove for middle meningeal artery

27 Diploic vein (sphenoparietal sinus)

28 Greater and lesser wings of sphenoid

29 Anterior border of middle cranial fossa

30 Frontonasal suture

31 Nasal bone

32 Infraorbital margin

33 Ethmoid bone

34 Sphenoid sinuses

35 Maxillary sinuses

36 Frontal sinuses

37 Zygomatic process of maxilla (projected onto body of zygoma)

38 Frontal process of zygoma

39 Coronoid process

40 Mandibular condyles (both sides)

41 Mandibular rami (both sides)

42 Air in epipharynx

43 Soft palate

44 Hard palate

45 Anterior nasal spine

46 Dens axis

47 Opisthion (posterior rim of foramen magnum)

Fig. 1.2 a, b Plain skull films (schematic) with normal variants (after Piepgras).

a AP projection.

b Lateral projection.

1 Hyperostosi s frontalis interna

2 Calcified choroid plexus (glomus) in trigone of lateral ventricles

3 Metopic suture

4 Sutural bone (incarial bone)

5 Frontal emissary vein

6 Diploic veins (Breschet confluence)

7 Pacchionian granulations

8 Ossification of the falx

9 Dural calcifications in the wall of the superior sagittal sinus

10 Bony groove for the transverse sinus

11 Diploic vein (sphenoparietal sinus)

12 Calcified interclinoid ligament

13 Calcified petrosellar ligament

14 Occipital spur

15 Pineal gland

Views of the facial bones primarily serve a screening function. They should be held to a minimum and are unnecessary if HR CT is planned. Given the comprehensive information that is furnished by HR CT, including detailed images of the temporal bone, there is no longer a sound rationale for traditional special roentgen views such as the Mayer, Schiiller, or Stenvers views.

Conventional Tomography

The goal of conventional tomography is to provide nonsuperimposed views of complex bony structures. Unlike CT, however, it yields images in which a particular tissue layer is displayed preferentially but not in isolation. The spatial resolution depends mainly on the tomographic slice thickness and the degree to which the image blurs areas that are of no diagnostic interest. High resolution requires a high tomographic angle and a multidimensional (nonlinear) pattern of tomographic motion. Conventional tomography may still have a role in the radiologic evaluation of the facial skeleton, but HR CT is usually more rewarding and also involves less radiation exposure.

Computed Tomography

O. Jansen and K. Sartor

In this digital, computer-based sectional imaging modality, the body region of interest—in this case the skull—is scanned with a narrow bundle of roentgen rays that is passed across the region. As successive projections are obtained around the region of interest, a curved or circular array of highly sensitive detectors registers the resulting roentgen ray attenuation along each projection. From the large volume of data that is generated, a computer calculates the spatial distribution of the attenuation values, or “tissue densities,” and uses a Fourier transform to reconstruct an image of the scanned body section. The resulting matrix of 256×256, or 512×512, or more numerical values undergoes a digital-to-analog conversion to produce an analog image, which is displayed on the computer monitor. This two-dimensional (2-D) image is composed of picture elements, or pixels, each of which represents a volume element (voxel) since the scanned tissue section has a definite thickness. The size of the individual volume element depends on the slice thickness, the size of the matrix, and the diameter of the image field. In the visible image, which can be documented on sheet film, each matrix or density value is assigned a corresponding shade of gray that is described quantitatively in terms of Hounsfield units (HU). The Hounsfield scale ranges from –1000 H U to +3000 HU, in which water is assigned an arbitrary value of 0. In a standard CT examination, additional sectional images are generated by advancing the examination table a given distance, usually equal to the slice thickness, and repeating the scanning process. The term “scan” usually refers to the sum of the sectional images from one body region, or it may refer to the sectional image itself.

Visual interpretation of the sectional images is facilitated by selecting portions of the density scale (windows) for modulating the image contrast. Since the examiner can vary these windows electronically over a wide range, controls are needed for image manipulation. Thus, standard window settings have been devised for different body regions and for specific diagnostic situations. The correct window setting is particularly important for cranial CT, because the density differences in the brain parenchyma cover a narrow range of values. The evaluation of two tissue types with very different densities, such as the brain parenchyma and cranial bone, requires two different window settings for image documentation:

• Soft-tissue window

Fig. 1.3 a, b Axial cranial CT scan at the level of the external auditory canals. Slice thickness: 4 mm; normal findings.

a Soft-tissue window.

b Bone window.

Hyperdense and hypodense streak artifacts are particularly common in slices that contain the petrous pyramids and brainstem or the skull base and portions of the paranasal sinuses or mastoid air cells. This type of image artifact, i.e., density abnormality with no anatomic counterpart, can also occur in the basal portions of the posterior fossa, which is why structures such as the lower brainstem or basal components of the cerebellar hemispheres can seldom be reliably evaluated on CT scans. Even below the calvaria, beam hardening can increase the apparent tissue density to a degree that can mimic a subdural hematoma. Scans occasionally show hypodense areas on the floor of the middle fossa that are difficult to distinguish from encephalitis or an early stage of ischemic brain infarction. Motion artifacts do not affect an isolated part of the image, and generally they can be readily identified by their characteristic pattern and their proximity to bone.

Standard Techniques

While CT is inferior to MRI in terms tracranial soft-tissue contrast, it still offers several advantages:

For these reasons, CT has become the imaging modality of choice in standard care settings and for emergency diagnosis.

Fig. 1.4a, b Axial cranial CT scan at the level of the basal ganglia. Slice thickness: 8 mm; normal findings.

a Before i.v. contrast administration.

b After i.v. contrasta dministration (100 mL, 300 mg/mL).

Fig. 1.5a, b Coronal cranial CT scan through the orbits and paranasal sinuses. Slice thickness: 4 mm; hybrid window for simultaneous bone and soft-tissue imaging; normal findings.

a Scan plane at the center of the globes.

b Scan plane 2 cm anterior to the orbital apex.

Fig. 1.6 CT angiography. 3-D reconstruction of the circle of Willis (viewed from above) shows an acute occlusion of the right middle cerebral artery. Spiral acquisition of image data following an i.v. contrast bolus (100 mL/20 s, 300mg/mL).

The blood flow to an organ, in this case the brain or portions of it, can be investigated by dynamic CT (perfusion CT). As in CTA, sectional images are acquired very rapidly after the injection of contrast medium at a rate of approximately one image per second, but the table remains stationary so that the image plane is unchanged. The inflow and washout of the contrast medium in the tissue can be precisely observed and displayed in parametric images that describe some aspect of cerebrovascular perfusion. For example, there are parametric images for cerebral blood flow, cerebral blood volume (CBV), and the “peak arrival time” of the contrast medium. A disadvantage of this method when used wit h a traditional CT scanner is that only one brain section can be examined per bolus injection. The latest generation of CT scanners, known as multislice scanners, can map the perfusion of large brain areas by using multiple detector rings to record roentgen ray absorption. One question of clinical interest is whether this new technology can supply cerebrovascular data that are useful in selecting acute stroke patients for a recanalization procedure.

Image Data Postprocessing

Even single CT slices can be used as a basis for making precise measurements of length and area. Reliable volume measurements, however, require a more sophisticated approach that ideally includes spiral scanning and a computer workstation. Density measurements, often performed directly at the CT console, are helpful for tissue discrimination. “Secondary” reconstruction, or image reformatting, to produce CT images in planes other than the primary scan plane requires an uninterrupted series of individual slices.

Reformatted images are used mainly to investigate topographic issues and questions regarding the extent of space-occupying lesions.

“Segmentation” is a technique that can be used to delete precisely defined normal or pathologic structures from the image data set.

Magnetic Resonance Imaging

S. Heiland and M. Skalej

Basic Principles and Applications

The time delay between two successive excitation cycles is called the repetition time (TR), and the delay between excitation and signal reception is called the echo time (TE). The frequency and phase encoding are deciphered using a 2-D Fourier transform (2-D FT). As in CT, this process yields sectional images in which the gray level of each picture element (pixel) is proportional to the amplitude of the RF signal received from the corresponding volume element (voxel).

Standard Techniques

With its high contrast resolution and multiplanar capabilities, MRI has joined CT as one of the premier imaging modalities in neuroradiology. The reasons why it is not generally preferred over CT relate to cost and capacity: the examinations are relatively expensive, and CT scanners far outnumber magnetic resonance (MR) imagers at the present time. On the other hand, the primary use of MRI can often shorten the diagnostic algorithm. A significant advantage of MRI over CT is that the acquisition technique can be modified to display a variety of parameters (e.g., relaxation times, PD, diffusion, perfusion, flow), whereas image contrast in CT is based solely on differences in electron density. MRI is also the first imaging modality that provides equally detailed information on both structure and function.

Intravenously administered paramagnetic gadolinium-containing contrast agents have a major role in current MRI practice. These agents amplify the differences in signal intensity between tissues that differ in their blood flow or BBB integrity. Paramagnetic contrast agents are highly effective and extremely safe. The frequency of their use depends basically on the clinical population. In Germany, gandoliniumcontaining contrast agents are administered in 50–80% of cranial MRI examinations. The usual dose is 0.1 mMol/kg body weight, but occasionally it is prudent to increase the dose (to as much as 0.3 mMol/kg) when, say, a low-field imager is used (field strength < 0.5 T) or when looking for faintly enhancing lesions.

The effective, economical utilization of MRI requires a more comprehensive knowledge of the physical principles of the modality and the examination technique than this book can convey, and so the reader is referred to other sources for details. The discussions below are intended as an introduction to help in understanding the clinical chapters.

The MR experiment described above can be modified further to achieve specific types of image contrast. The basic pattern in which RF pulses and gradients are switched on and off is called the sequence.

Spin-Echo Technique

Fig. 1.10 Pulse sequence in the spin-echo imaging method. Tr = 2 s, TE = 20 ms. Accordingly, a 180° rephasing pulse is applied 10 ms after the initial RF pulse. The phase encoding gradient is incremented a total of 256 times. The phase encoding gradient is incremented a total of 256 times.

G_ph Phase encoding gradient

G_R Readout gradient

G_S Slice selection gradient

H_F RF pulse

Gradient-Echo Technique

Because a 180° pulse is not used, transverse relaxation in GRE sequences occurs faster than in SE sequences and is described by the time constant T2*. Also, a flip angle < 90° is commonly used in GRE sequences. This allows the longitudinal magnetization to recover more quickly, making it possible to shorten the repetition time and total acquisition time. By selecting various combinations of flip angles, TR, and TE, and by using additional RF pulses or special gradients, the operator can obtain images with different weighting of the parameters PD, T1, and T2*.

Fig. 1.11 a—c Axial MR images at the level of the lateral ventricles. Image contrast can be manipulate by varying the acquisition parameters.

a T1-weighted image: Tr = 600 ms, TE = 12 ms.

b T2-weighted image: Tr = 3300 ms, TE = 85ms.

c PD-weighted image: Tr = 3300 ms, TE = 14 ms.

Thus, despite their shorter acquisition time, GRE sequences have not replaced SE sequences for clinical imaging. Currently they are used as a supplement to SE sequences for applications such as the high-resolution 3-D imaging of the brain or inner ear.

Inversion-Recovery Technique

The initial 180° pulse “inverts” the longitudinal magnetization, which begins to revert to its initial state through longitudinal relaxation. If a 90° excitation pulse is applied at time T1 (T1 = inversion time) during this relaxation process, it will rotate the recovering longitudinal magnetization into the transverse plane. Thereafter, the process is similar to an SE sequence, using a 180° rephasing pulse to generate an echo. The strength of the measured signal yields information on the longitudinal magnetization at time T1.

Another advantage of IR sequences is that the signal from certain tissue types or substances can be selectively suppressed by the proper selection of T1. This forms the basis for techniques such as short-TI inversion recovery (STIR) to suppress fat signals and fluid-attenuated inversion recovery (FLAIR) to suppress signals from the cerebrospinal fluid (CSF).

Special Techniques

Rapid Acquisition with Relaxation Enhancement

In rapid acquisition with relaxation enhancement (RARE), known also as turbo SE or fast SE, several echoes (each produced by a 180° rephasing pulse) are differently encoded and read in rapid succession following an initial 90° pulse. This shortens the scan time by a factor equal to the number of echoes that are received in each excitation.

Because only the data readout is modified in the RARE technique, it can be combined with any pulse sequencing method (SE, GRE, or IR). Numerous acronyms have been employed in the literature for these RARE variations. Wit h very fast gradient systems, all the echoes necessary for image encoding can be read after a single RF excitation. These single-shot RARE techniques have also been called HASTE and EXPRESS in the literature.

Echo-Planar Imaging

Echo-planar imaging (EPI), like the RARE technique, involves the readout of multiple echoes after each excitation, but EPI differs from RARE in that a 180° refocusing pulse is not used between the echoes. If only part of the echoes necessary for image generation are read in a single excitation, the technique is called multishot EPI. If a complete readout is obtained, it is called singleshot EPI.

Because the fast gradient variations call for fast gradient switching times and high maximal gradients, the hardware of the MRI unit and especially the gradient systems must meet stringent requirements. As a result, it is only since the mid-1990s that EPI-compatible systems have become available for clinical use.

Diffusion-Weighted MRI

Calculation of the rCBV and rMTT requires sophisticated postprocessing of the MR data. Both parameters yield important information, especially in the diagnosis of cerebrovascular diseases that involve microcirculatory changes, such as ischemia and hemodynamically significant stenoses of arteries supplying the brain.

Functional MRI

Functional MRI capitalizes on the fact that oxygen-rich blood has different magnetic properties than oxygen-poor blood, and that this difference can be visualized with T2*-weighted sequences. Thus, the paramagnetic deoxyhemoglobin leads to a fall of signal intensity in T2*-weighted images while the diamagnetic oxyhemoglobin has no effect on signal intensity. The activation of a particular brain area is associated with a rise in oxygen consumption accompanied by an increase in local blood volume and perfusion that overcompensates for the higher consumption. During stimulation, then, increased signal intensity is observed in activated brain tissue on T2*-weighted images. Because this signal increase is small (1–15%, depending on the field strength and type of stimulus), sophisticated postprocessing is required. This might include motion corrections, for example, as well as statistical techniques to distinguish true cerebral activation from pseudoactivation.

Magnetization Transfer Contrast Imaging

In all MR sequences described thus far, the signal arises entirely from “free” protons that are bound to very small molecules such as water or lipids. Protons that are bound to immobile macromolecules, such as proteins, cannot be detected. Because of their very short relaxation time (less than 1 ms), they no longer emit a signal by the time the signal is measured. However, the free protons and macromolecule-bound protons still interact through chemical exchanges and mutual energy transfer. In this way, the macromolecule-bound protons influence the relaxation times of the free protons even though they remain “invisible” themselves. Because the bound protons can be excited over a much larger frequency range, these proton species can be selectively excited and “saturated,” as in the magnetization transfer contrast imaging (MTC) experiment. But the exchange processes also transfer magnetization to the free protons. The greater the proportion of macromolecule-bound protons in the sample, the faster this transfer takes place. If the RF saturation pulse is followed immediately by a standard MR sequence (SE, GRE, IR), the presaturation in the more proteinrich tissue leads to a greater signal reduction than in protein-poor structures. The new type of contrast that results from this process is used in MRA to suppress signals from stationary tissues; it also facilitates the diagnosis of white-matter diseases and of contrast-enhancing primary tumors and metastases.

Fig. 1.13 Perfusion MRI. A transient, logarithmic fall in signal intensity is measured in the gray and white matter of the brain. Because of the higher cerebral blood volume, the signal decrease is greater in the gray matter (GM) than in the white matter (WM).

Fig. 1.14a, b MR angiography. MIP from data sets acquired with a TOF sequence.

a Coronal display of the large basal arteries, including the terminal segments of both carotid arteries.

b Lateral display.

These artifacts can be minimized by contrastenhanced MRA. This technique is based on the principle that paramagnetic contrast material, when administered as an i.v. bolus, causes a shortening of T1 almost exclusively within the vessels during the first pass. This provides contrast between the vessels and stationary tissue in fast, heavily T1-weighted GRE images. One disadvantage of this technique is that contrast injection and sequence initiation must be precisely timed to ensure that the vessels of interest are imaged during maximum bolus transit. Another problem with this method is that it is difficult to image arteries and veins separately.

A third commonly used MRA technique is phase-contrast MRA (PC-MRA). As in diffusionweighted imaging, a velocity encoding gradient is used to selectively image flowing protons. A technique based on PC-MRA has been developed for quantifying the flow velocity in blood vessels.

A special technique for postprocessing digital MRA data is MIP. In this technique the acquired 3-D data set is reduced to 2-D projections from arbitrarily selected viewing angles. This is done by projecting a ray electronically from each observation point through the 3-D data set. The maximum gray level encountered by each of the rays within the 3-D data set is then displayed on the projection plane orthogonal to the ray.

Magnetic Resonance Spectroscopy

Fig. 1.15a, b Volume-selective MRS (proton spectroscopy) in the cerebral white matter.

a Localizing image. the signal (spectrum) is acquired from a cube-shaped volume with a 2-cm edge length.

b Corresponding spectrum. The area under the peaks correlates with the concentration of the particular substance.

Cho Choline

NAA N-acetyl aspartate

Angiography

M. Forsting

Basic Principles and Applications

Selective transfemoral angiography to depict the cerebral arteries and veins can almost always be done in patients under 50 years of age using a 5F or 4F universal catheter with an approximately 45° angled tip. The catheterization time in older patients can often be reduced by using a sidewinder catheter with a doubly curved end. Occasionally a twist-resistant 6F catheter can also be used. A n introducer sheath helps to facilitate catheter changes and reduce trauma to the femoral artery during tedious vascular catheterizations. The former practice of injecting contrast medium into catheters that are difficult to see with fluoroscopy is not recommended with nonionic media, because they do not have the anticoagulant effect of ionic media. To avoid thromboembolic complications, the catheter should be periodically flushed with saline, especially after contrast injections and each time a guidewire has been used. This can be done manually or with an infusion pump, with 500IU of heparin added to the irrigating fluid per 500 mL of NaCl.

The complications of cerebral angiography are almost entirely thromboembolic, although no reliable figures are available on their incidence. When discussing the procedure with the patient, the angiographer should state the complication rate in his/her own department. Publications report about a 3% rate of transient neurologic deficits, a 0.2–1 % rate of persistent neurologic deficits, and a mortality rate of less than 0.1%. The angiographic risk is considerably higher in patients with underlying atherosclerotic disease than in the mostly younger patients who undergo angiography for a subarachnoid hemorrhage or cerebral vascular malformation.

Arch Aortography and Subclavian Angiography

Survey View of the Aortic Arch

Survey angiography of the aortic arch is performed with a 4F or 5F pigtail side-hole catheter that is advanced to the ascending aorta by the transfemoral route. Forty to sixty mL of contrast medium injected at a rate of 25 mL/s (cut-film technique) or 30 mL injected at a rate of 15 mL/s (DSA technique) will adequately opacify the aortic arch and supra-aortic vessels as far as the skull base. The optimum visualization of all branch vessels often requires two injections—one in a left oblique projection and one in a right oblique projection. A global contrast injection into the aortic arch is rarely sufficient to define the intracranial vessels, however. It should also be noted that the preoperative evaluation of atherosclerotic wall changes in the cervical carotid bifurcation requires a sound knowledge of intracranial vascular anatomy.

Subclavian Angiography

Subclavian angiography employs the same endhole catheters used for selective cerebral angiography. The catheter is placed in the initial portion of the subclavian artery, and 10–12 mL of contrast medium is injected at a rate of 10 mL/s. This study is recommended for preliminary visualization of the branch vessels prior to selective vertebral artery catheterization.

Carotid Angiography

Fig. 1.17a—d Carotid angiography. Diagram of vascular anatomy following injection of the common carotid artery.

a Arteries: AP projection.

1 Common carotid artery

2 Internal carotid artery

3 Carotid siphon

4 Ophthalmic artery

5 Posterior communicating artery

6 Posterior cerebral artery

7 Anterior choroidal artery

8 Anterior cerebral artery (A1 segment)

9 Middle cerebral artery (M1 segment)

10 Lenticulostriate arteries

11 Anterior communicating artery

12 Anterior cerebral artery (A2 segment)

13 Frontobasal artery

14 Frontopolar artery

15 Callosomarginal artery

16 Pericallosal artery

17 Insular arteries

18 Prerolandic artery

19 Rolandic artery

20 Parietal arteries

21 Artery of angular gyrus

22 Deep temporal artery

23 External carotid artery

24 Superior thyroid artery

25 Maxillary artery

26 Middle meningeal artery

27 Superficial temporal artery

28 Occipital artery

b Arteries: lateral projection.

1 Common carotid artery

2 Internal carotid artery

3 Carotid siphon

4 Ophthalmic artery

5 Posterior communicating artery

6 Posterior cerebral artery

7 Anterior choroidal artery

8 Anterior cerebral artery (A1 segment)

9 Middle cerebral artery (M1 segment)

10 Lenticulostriate arteries

11 Anterior communicating artery

12 Anterior cerebral artery (A2 segment)

13 Frontobasal artery

14 Frontopolar artery

15 Callosomarginal artery

16 Pericallosal artery

17 Insular arteries

18 Prerolandic artery

19 Rolandic artery

20 Parietal arteries

21 Artery of angular gyrus

22 Deep temporal artery

23 External carotid artery

24 Superior thyroid artery

25 Maxillary artery

26 Middle meningeal artery

27 Superficial temporal artery

28 Occipital artery

Fig. 1.17 c Veins: AP projection.

1 Superior sagittal sinus

2 Ascending cerebral veins

3 Inferior sagittal sinus

4 Vein of septum pellucidum (septal vein)

5 Thalamostriate vein

6 Internal cerebral vein

7 Basal vein (of Rosenthal)

8 Great cerebral vein (of Galen)

9 Straight sinus

10 Confluence of sinuses

11 Transverse sinus

12 Temporo-occipital vein (of Labbe)

13 Occipital sinus

14 Middle cerebral veins and junction with sphenoparietal sinus

15 Sigmoid sinus

16 Internal jugular vein

Fig. 1.17 d Veins: lateral projection.

1 Superior sagittal sinus

2 Ascending cerebral veins

3 Inferior sagittal sinus

4 Vena septi pellucidi (septal vein)

5 Thalamostriate vein

6 Internal cerebral vein

7 Basal vein (of Rosenthal)

8 Great cerebral vein (of Galen)

9 Straight sinus

10 Confluence of sinuses

11 Transverse sinus

12 Temporo-occipital vein (of Labbe)

13 Occipital sinus

14 Middle cerebral veins and junction with sphenoparietal sinus

15 Sigmoid sinus

16 Internal jugular vein

Fig. 1.18a—d Vertebral angiography. Diagram of vascular anatomy (after Piepgras).

a Arteries: AP projection (Towne projection).

b Arteries: lateral projection.

1 Vertebral artery

2 Muscular branches of vertebral artery

3 Anterior spinal artery

4 Left posterior inferior cerebellar artery

5 Inferior vermian artery

6 Tonsillohemispheric branch

7 Basilar artery

8 Anterior inferior cerebellar artery

9 Right and left superior cerebellar artery

10 Marginal branch

11 Superior vermian artery

12 Posterior cerebral artery (circular part)

13 Temporo-occipital artery

14 Internal occipital artery

15 Posterior callosal artery

16 Posterior choroidal arteries

17 Thalamic arteries

18 Posterior communicating artery

c Veins: Towne projection.

d Veins: lateral projection.

1 Superior sagittal sinus

2 Inferior sagittal sinus

3 Vena septi pellucidi (septal vein)

4 Thalamostriate vein

5 Internal cerebral vein

6 Great cerebral vein (of Galen)

7 Confluence of sinuses and falcotentorial veins

8 Straight sinus

9 Confluence of sinuses

10 Transverse sinus

11 Sigmoid sinus

12 Inferior petrosal sinus

13 Occipital sinus

14 Basal vein (of Rosenthal)

15 Precentral vein of cerebellum

16 Superior vermian vein

17 Inferior vermian vein

18 Hemispheric veins of cerebellum

19 Petrosal vein

20 Anterior pontomesencephalic vein

Fig. 1.19 Diagram of the circle of Willis (after Piepgras).

1 Anterior cerebral artery

2 Internal carotid artery

3 Posterior cerebral artery

4 Basilar artery

5 Circular part of posterior cerebral artery

6 Posterior communicating artery

7 Middle cerebral artery

8 Anterior communicating artery

Sonography

K. Atzor

Basic Principles and Applications

The following sonographic techniques are currently available for the diagnosis of craniocerebral diseases:

• Extracranial and transcranial Doppler ultrasound

• Color duplex sonography

• Cranial sonography in newborns and infants.

Ultrasound examinations are easy to perform, require no complex preparations such as sedation or anesthesia, and permit low-cost, risk-free follow-ups with no exposure to ionizing radiation. In infants, cranial sonography can generally be performed until the sixth to ninth month; after that the fontanelles are too small to provide a useful acoustic window for scanning with a 5 to 7.5-MHz sector transducer. The main difficulty in Doppler and duplex examinations of the cervical vessels is transducer orientation, i.e., selecting the proper insonation angle. In transcranial scanning, even an experienced examiner may have difficulty locating the acoustic window, which is in the temporal region above the zygomatic arch. Even with optimum transducer placement, much experience is needed to assign the recorded Doppler signals to specific circle of Willis vessels. Details on the physical and instrumental requirements of ultrasonography can be found in the specialized literature.

Sonography in Newborns and Infants

In recent years, cranial sonography has become a routine examination that is particularly useful in neonatology.

Newborns are screened with ultrasound to investigate the cause of developmental retardation, respiratory problems, or seizure disorders (malformations, hydrocephalus, cerebral atrophy), and to follow the progression of known diseases.

Fig. 1.20 Cranial ultrasound scan of a newborn. Coronal scan through the thalamus, normal findings (with kind permission of Dr. Dittrich, Pediatric Hospital, University of Mainz).

1 Lateral ventricle

2 Thalamus

Fig. 1.21 Cranial ultrasound scan of a newborn. Parasagittal scan, normal findings (with kind permission of Dr. Dittrich, Pediatric Hospital, University of Mainz).

1 Corpus callosum

2 Lateral ventricle

2 Thalamus

Transcranial and Perioperative Sonography

Doppler and duplex sonography are among the studies currently used in the hemodynamic evaluation of neurovascular diseases. Transcranial Doppler ultrasound (TCD) is used mainly to examine the large basal cerebral arteries for spasms and thromboembolic stenoses, particularly the basilar artery and middle cerebral arteries. It can also be used in conjunction with the balloon occlusion test before interventional procedures or radical tumor surgery to confirm that the remaining arteries can provide sufficient flow to the brain following the therapeutic occlusion of a major supply vessel. Transcranial Doppler ultrasound is also useful in the intraoperative assessment of collaterals before temporary or permanent vascular clipping and to detect arterial spasms. This can help determine the cause of postoperative parenchymal hypodensities on CT scans and assess the potential for ischemic infarction. Another application of TCD is for continuous hemodynamic monitoring in endarterectomies performed at the carotid bifurcation. In tumor resections, ultrasound can be used to locate and delineate more deeply situated tumor components, and it is also effective in the precise localization of abscesses and cysts.

As in other sonographic techniques, the quality of the study depends on the skills and experience of the examiner. In experienced hands, the results are comparable to those of intraarterial DSA. Intravenous DSA of the cervical vessels has been completely superseded by CDS. A major advantage of duplex sonography over invasive angiography is the ability to obtain precise crosssectional vascular images. The degree of stenosis revealed by sonography is often greater than the stenosis visible on angiograms. If the ultrasound findings are equivocal, isolated stenoses can also be reliably detected by CTA and MRA (especially when contrast agents are used).

Nuclear Medicine Imaging

M. Henze

Basic Principles and Applications

Functional imaging in nuclear medicine is based on the tracer principle described by Hevesy. It states that the biochemical properties of organic compounds remain unchanged when stable atoms are replaced by corresponding radioactive isotopes. The extremely low concentrations (in the micromolar to picomolar range) ensure that metabolic processes in the examined system are not altered and that pharmacologic actions do not occur. As a result, physiologic processes such as metabolism, neurotransmission, gene expression, and signal transduction can be studied noninvasively in living patients. Functional abnormalities develop well in advance of morphologic changes. Nuclear medicine procedures are sensitive enough to detect extremely low molar concentrations in the range of 10⁻¹⁰ to 10⁻¹²; in this respect they are 10⁴ to 10⁶ times more sensitive than MRI.

Planar Imaging

Fig. 1.24 ^99mTc-ECD SPECT. Axial scan on the right shows normal cerebral perfusion of the cortex, basal ganglia, thalamus, and cerebellum. Scan on the left shows decreased perfusion of the thalamus and occipital cortex (arrows).

Perfusion SPECT

Fig. 1.25 ¹²³I-FP-CIT SPECT. Axial scan shows normal, symmetrical distribution of presynaptic dopamine transporters.

Cerebrospinal Fluid SPECT

¹¹¹In-Ca²⁺-diethylenetriaminepentaacetic acid (DTPA) is administered intrathecally (usually in the lumbar region, sometimes suboccipitally) in suspected cases of occult or intermittent CSF rhinorrhea. Multiple nasal swabs are then taken and checked for activity, which is checked against the activity in the blood. CSF scintigraphy is also used in rare cases to test the function of CSF shunts and evaluate the direction and velocity of the CSF flow (e.g., in normal pressure hydrocephalus).

Positron Emission Tomography

With its increasing availability, PET is quickly gaining importance in the clinical diagnosis of cerebral diseases. Positron emitters emit one electron neutrino and one positron. The neutrino passes through tissue with no interactions, while the positron travels approximately 1–2m m before colliding with an electron and forming a positronium. After a life time of 1.2×10⁻¹⁰s, the resting mass of the positronium is transformed into pure energy. This annihilation process yields two gamma quanta, each with an energy of 511 keV, that diverge at an angle of almost exactly 180°. A dedicated PET scanner consists of multiple detector rings arranged axially around the patient. Most detectors in current use are made of bismuth germanate (BGO). For emission measurement during PET acquisition, pairs of opposing detectors respond simultaneously to the arrival of the two gamma quanta produced by the annihilation. This two-photon detection process, called coincidence detection (Δt ≈ 10–15 ns), yields considerably more information than the detection of a single photon by SPECT. This “electronic” collimation eliminates the need for lead collimators, resulting in a marked increase in counting rate yields. A transmission measurement with a rotating external source, such as Ge (half-life: 271 days), is obtained to correct for the attenuation of photons by the tissue. Modern full-ring PET systems have a transaxial resolution in the range of 3–6 mm. The newest hybrid systems combine PET and CT scanners; after correlation and fusion of the data sets, they permit the direct visualization and analysis of both functional and morphologic parameters.

The data can be analyzed semiquantitatively by calculating standard uptake values (SUV = tissue concentration adjusted for injected activity and body weight). Another approach is the absolute quantification of metabolic parameters, such as glucose turnover (in umol/min/lOOg) and tissue perfusion (in mL/min/100 g), based on the use of a pharmacokinetic compartment model and a knowledge of the arterial input function of the administered tracer.

Positron-emitting radionuclides are produced in a cyclotron—a circular accelerator in which targets of stable isotopes are bombarded with heavy charged particles (protons, deuterons, alpha particles) in a vacuum. Many positron emitters have a very short half-life (HL), and so they must be produced close to the facility where they will be used. Examples are: ¹⁵O (HL: 2 min), ¹³N (HL: 10 min), and ¹¹C (HL: 20 min). Exceptions are generator nuclides such as ⁶⁸Ga (HL: 68 min), which can be separated from the “mother-daughter” nuclide system without a cyclotron. One of these radiopharmaceuticals, ⁶⁸Ga-DOTATOC, is a labeled peptide that can be used to quantify the somatostatin receptor density of meningiomas, thereby differentiating them from other tumors such as neurinomas.

Fig. 1.26 ¹⁸F-FDG PET. Axial scan demonstrates normal, symmetrical cerebral glucose metabolism.

¹⁸F has assumed special importance, since its relatively long HL of 110 min allows it to be transported commercially over moderate distances. With its long HL, the agent can also be used to label hormones, proteins, and antibodies with slow kinetics. Almost any molecule can be labeled by substituting ¹⁸F for hydrogen or an OH group without significantly altering its properties. Also, F is distinguished by a low positron energy (0.64 MeV) and a correspondingly small tissue range.

The advantages of PET over SPECT are the possibility of absolute quantification, the ability to directly label a number of physiologically occurring molecules (e.g., glucose, water), accurate attenuation correction, better spatial resolution, and a better count-rate yield. The disadvantages of PET are its higher costs, the greater logistical demands due to the short HL of positron emitters, and the still limited availability of PET scanners.

Meanwhile, techniques have been devised for labeling numerous biomolecules with positron emitters. PET has found broad applications in clinical neurology and oncology in recent years. The most important PET tracers and their applications in nuclear medicine neuroimaging are described below.

Glucose Metabolism PET

Fig. 1.27 ¹⁸F-fluoro-dopa PET. Axial scan shows normal, symmetrical distribution of presynaptic nigrostriatal neurons

Perfusion PET

The measurement of cerebral blood flow is based on the indicator dilution method. The models most often used for quantitation are based on the unicompartment model developed by Kety during the 1950s. The tracers used in experimental settings consist of highly diffusible, usually ¹⁵O-labeled molecules such as H₂¹⁵O or ¹⁵O-butanol, which allow for repeated measurements owing to their short HL of 2 min (e.g., before and after visual or auditory stimulation, before and after cognitive or motor tasks to locate risk structures prior to tumor resection). Under physiologic conditions, glucose metabolism and blood flow are closely linked over a wide range of values. Gray matter perfusion measured by the ¹⁵O method in the awake, resting state is approximately 60–80 mL/min/100g. This contrasts with an average white-matter perfusion of only about 20–30 mL/min/100g.

Amino Acid Metabolism PET

The physiologic extraction of amino acids from the blood into the brain occurs rapidly, even across the intact BBB. But because it is very low (only about 1 %), physiologic intracerebral tracer uptake is also low, and even low-grade gliomas contrast sharply with the low surrounding activity. Some tracers label this carrier-mediated transport exclusively (e.g., ¹⁸F-fluoroethyl-tyrosine, FET), while others are additionally incorporated into proteins (e.g., L-methyl-¹¹C-methionine, MET). Labeled amino acids provide a means of distinguishing recurrent glioma from radiation necrosis. Because very little of these tracers accumulates in healthy brain parenchyma (unlike FDG), the histologic extent of even lowgrade gliomas can be determined.

Neurotransmission PET

Numerous agents have become available for the PET imaging of presynaptic or postsynaptic parameters, although most have been used for research purposes. Our present discussion is therefore limited to dopaminergic and gammaaminobutyric acid (GABA)-ergic neurotransmission.

I. Mader and W. Grodd

To evaluate many diseases in pediatric neurology, it is necessary to know the appearance of normal brain myelination on MR images. In this section we shall review normal myelination patterns in the developing brain and show examples of their imaging appearance.

The maturation of the various tract systems appears different on T1-weighted images than on T2-weighted images. Myelinated brain areas appear bright on T1-weighted images and dark on T2-weighted images. Myelination tends to occur in predictable patterns, as outlined below:

• Myelination begins at the neuron soma and proceeds distally along the axons,

• The brainstem and diencephalon myelinate from caudal to cephalad and from dorsal to ventral,

• Sensory tracts myelinate earlier than motor tracts,

• Myelination in the telencephalon begins at the central sulcus and progresses toward the periphery, spreading first in the occipital direction, then temporally, and finally in the frontal direction,

• The last structures to myelinate are the corticocortical association tracts (U fibers).

The imaging parameters should be selected to provide clear differentiation of the gray and white matter.

Conventional SE and turbo-SE (TSE) sequences (fast-SE, RARE sequences) are largely equivalent for detecting relevant changes in T2-weighted signal intensity when comparable TR (at least 3000 ms) and TE_eff values (at least 80 ms) are used.

Fig. 1.28a—h Normal brain maturation in the first postnatal month.

a—d T1-weighted axial images:

a High signal: pontine tegmentum with medial lemniscus.

b High signal: quadrigeminal plate and decussation of superior cerebellar peduncles.

c High signal: posterior limb of internal capsule (incomplete) and ventrolateral thalamic nuclei.

d High signal: centrum semiovale and cortex of central region.

e—h T2-weighted axial images:

e, f Low signal (infratentorial): structures that appear brighton T1-weighted images.

g, h Low signal (supratentorial): posterior limb of internal capsule (less extensive than on T1-weighted images), ventrolateral thalamic nuclei, and cortex of central region.

Fig. 1.29a—h Normal brain maturation at 4 months.

a—d T1-weighted axial images:

a High signal: middle cerebellar peduncles.

b, c High signal: cerebral peduncles, optic radiation as far as primary visual cortex, anterior limb and entire posterior lim b of internal capsule.

d High signal: white matter (centrum semiovale), precentral and postcentral gyri.

e—h T2-weighted axial images:

e Low signal: middle cerebellar peduncles.

f Low signal: quadrigeminal plate and decussation of superior cerebellar peduncles.

g, h Low signal: posterior limb (incomplete) of internal capsule(g) and slight hypointensity of centrum semiovale (h).

Fig. 1.30a—h Normal brain maturation at 8 months.

a—d T1-weighted axial images:

a High signal: middle cerebellar peduncles.

b High signal: cerebral peduncles.

c, d High signal: genu of corpus callosum, internal and external capsule(c), supratentorial white matter, and subcortical U fibers(c, d) except temporal and frontopolar areas.

e—h T2-weighted axial images:

e Low signal: middle cerebellar peduncles.

f Low signal: cerebral peduncles.

g, h Low signal: genu of corpus callosum, internal and external capsule (c), optic radiation and white matter of pericentral region (g, h).

Fig. 1.31 a—h Normal brain maturation at 16 months.

a—d T1-weighted axial images:

a High signal: middle cerebellar peduncles.

b High signal: cerebral peduncles.

c High signal: corpus callosum, internal and external capsule, lateral portions of thalamus.

d High signal: cerebellar and cerebral white matter (a—d).

e—h T2-weighted axial images:

e Low signal: middle cerebellar peduncles and folia.

f—h Low signal: cerebral peduncles (f), corpus callosum, internal and external capsule (g), supratentorial white matter (patchy appearance, f—h) except for bright areas bordering the occipital horns of the lateral ventricles(h).

Fig. 1.32 a—h Normal brain maturation at 24 months.

a—d T1-weighted axial images: myelination is largely complete

e—h T2-weighted axial images:

e Low signal: middle cerebellar peduncle is darker than cerebellar white matter.

f—h Low signal: optic radiationto occipital pole (f), putamen and pallidum separated by narrow dark zone (g), all of supratentorial white matter is dark (h) exceptfor frontobasal and temporopolar areas (f).

Fig. 1.33 a—h Normal brain maturation at 24 months.

a—d T1-weighted axial images.

e—h T2-weighted axial images.

Myelination is largely complete. Physiologic high-signal areas border the occipital horns of the lateral ventricles (h) and may persist into adolescence.

Bird, R. C., M. Hedberg, B. P. Drayer, P.J. Keller, R. A. Flom, J. A. Hodak: MR assessment of myelination in infants and children. Amer. J. Neuroradiol. 10 (1989) 731–740

Camargo, E. E.: Brain SPECT in Neurology and Psychiatry. J. Nucl. Med. 42 (2001) 611–623

Croft, B.Y.: Single-Photon Emission Computed Tomography. Year Book, Chicago 1986

Fareed, J., J.M. Walenga, G. E. Saravia, R.M. Moncada: Thrombogenic potential of non-ionic contrast media? Radiology 174 (1990) 321–325

Forsting, M., H. Sahl, K. Sartor: Selektive Darstellung der supraaortalen Arterien. In Kauffmann, G. W., W. S. Rau, T. Roeren, K. Sartor: Rontgenfibel. Springer, Berlin 1995

Friebolin, H.: Ein- und zweidimensionale NMR-Spektroskopie: Eine Einfuhrung. VCH, Viernheim 1988

George, M. S.,H. A. Ring, D. C. Costa, P. J. Ell, K. Kouris, P. H. Jarritt: Neuroactivation and Neuroimaging with SPET. Springer, Berlin 1991

Grodd, W.: Kernspintomographie neuropädiatrischer Erkrankungen. Normale Reifung des kindlichen Gehirns. Klin. Neuroradiol. 3 (1993a)13–27

Grodd, W.: Normal and abnormal patterns of myelin and development of the fetal and infantile human brain using magnetic resonance imaging. Curr. Opin. Neurol. Neurosurg. 6 (1993b) 393–397

Grosu, A. L. et al.: First experience with I-123-Alphamethyl-tyrosine SPECT in the 3-D radiation treatment planning of brain gliomas. Int. J. Radiation Oncology Biol. Phys. 47 (2000) 517–526

Harders, A. G.: Möglichkeiten und Ergebnisse der transkraniellen Doppler-Sonographie in der prä- und postoperativen Anwendung. Jahrbuch der Neurochirurgie. Regensberg & Biermann, Münster 1988

Heiland, S., K. Sartor: Magnetresonanztomographie beim Schlaganfall – Methodische Grundlagen und klinische Anwendungen. Fortschr. Röntgenstr. 171 (1999)3–14

Henze, M., J. Schuhmacher, P. Hipp, J. Kowalski, D. Becker, J. Doll, M. Mäcke, M. Hofmann, J. Debus, U. Haberkorn: PET Imaging of somatostatin receptors using [⁶⁸Ga]DOTA-D Phe¹-Tyr³-Octreotide (DOTATOC): First results in meningioma-patients. J. Nucl. Med. 42 (2001) 1053–1056

Hoffman, J. M., K. A. Welsh-Bohmer, H. Hanson et al.: FDG PET in patients with pathologically verified dementia. J. Nucl. Med. 41 (2000) 1920–1928

Hör, G., S. Adams, R. P. Baum, A. Hertel, I. A. Adamietz, H. D. Böttcher et al.: Impact of single photon emission computed tomography and positron emission tomography on diagnostic oncology. Diagn. Oncol. 4 (1997) 297–321

Kadir, S.: Diagnostische Angiographic Thieme, Stuttgart 1991

Kalender, W. A., K. Wedding, A. Polacin, M. Prokop, C. Schaefer-Prokop, M. Galanski: Grundlagen der Gefäßdarstellung mit Spiral-CT. Akt. Radiol. 4 (1994) 287–297

Kalender, W. A., W. Seissler, E. Klotz, P. Vock: Spiral volumetric CT with single breath hold technique, continuous transport and continous scanner rotation. Radiology 176 (1990) 181–183

Koenig, M., E. Klotz, B. Luka, D.J. Venderink, J. F. Spittler, L. Heuser: Perfusion CT of the brain: diagnostic approach for early detection of ischemic stroke. Radiology 209 (1998) 85–93

Kremer, H., W. Dobrinski: Sonographische Diagnostik. Urban & Schwarzenberg, München 1987

Kuwert, T. et al.: Iodine-123-α-methyl tyrosine in gliomas: Correlation with cellular density and proliferative activity. J. Nucl. Med. 38 (1997) 1551–1555

Kuwert, T: Hirntumoren. In: PET in der klinischen Onkologie (Hrsg.: Wieler H.J.). Steinkopff 1999

Kwong, K. K., J. W. Belliveau, D. A. Chesler et al.: Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc. natl. Acad. Sci. 89 (1992) 5675–5679

Langleben, D. D., Segall G. M.: PET in differentiation of recurrent brain tumor from radiation injury. J. Nucl. Med. 41 (2000) 1861–1867

Le Bihan, D.: Diffusion and Perfusion Magnetic Resonance Imaging. Raven, New York 1995

Lee, B. C. P., A. E. Grassi, S. Scheduer, P. Auld: Neonatal intraventricular hemorrhage: a serial computed tomography study. J. Comput. assist. Tomogr. 3 (1979) 483–490

Malms, J.: Angiographic von Kopf, Hals und oberen Extremitaten. In Schild, H.: Angiographie – angiographische Interventionen. Thieme, Stuttgart 1994

Mayo, J. R.: High resolution computed tomography technique. Semin. Roentgenol. 26 (1991) 104–112

Neidl, K. F. W., M. Hermes, M. Voges, C. Kujat, K. D. Hagspiel: Single Photon Emission Tomography Hirnfunktionsdiagnostik für die klinische Routine. Radiologe 33 (1993) 603–611

Ogawa, S., D. W. Tank, R. Menon et al.: Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc. natl. Acad. Sci. 89 (1992) 5951–5955

Partridge, J.C, D. S. Babcock, J. J. Steiden, B. K. Han: Optimal timing for diagnostic ultrasound in low birthweight infants. J. Pediat. 102 (1983) 281–287

Phelps, M.E., J.C. Mazziotta, S.C. Huang: Study of cerebral function with positron computed tomography. J. Cereb. Blood Flow Metab. 2 (1982) 113–162

Piepgras, U.: Neuroradiologie. Thieme, Stuttgart 1977

Reiche, W., M. Grundmann, G. Huber: Die Dopamin (D₂-)Rezeptor-SPECT mit 123-Jodbenzamid (IBZM) in der Diagnostik des Parkinson-Syndroms. Radiologe 35 (1995) 838–843

Rüttinger, R., Kujat, C., Huismans, R.: Toxocariasis. In Henkes, H., H. W. Kölmel: Die entzündlichen Erkrankungen des Zentralnervensystems. Handbuch und Atlas. Ecomed, Landsberg 1993.

Sartor, K.: MR Imaging of the Skull and Brain. Springer, Berlin 1992

Schmitt, F., Stehling, M.K., Turner, R.: Echo-Planar-Imaging. Theory, Technique and Application. Springer, Berlin 1998

Schwarz, A., T. Kuwert: Nuklearmedizinische Diagnostik bei Erkrankungen des zentralen Nervensystems. Der Radiologe 40 (2000) 858–862

Stark, D. D., W. G. Bradley: Magnetic Resonance Imaging, 3^rd ed. Mosby, St. Louis 1999

Staudt, M., A. Breit: Myelination of the brain in MRI: a staging system. Pediat. Radiol. 23 (1993) 169–176

Tetickovic, E., K., Miksic, S. Tetickovic: Transkranielle Doppler-Sonographie während der Karotisendarteriektomie. Nervenarzt 63 (1992) 347–351

Van der Knaap, M. S., J. Valk: Magnetic Resonance of Myelin, Myelination and Mylelin Disorders, 2^nd ed. Springer, Berlin 1995

Wienhard, K., R. Wagner, W.-D. Heiss, 1989. PET. Klinische Anwendungen der PET-Untersuchungen des Gehirns. Springer-Verlag Berlin (Germany). 120–153

Wolf, K.-J., F. Fobbe: Farbkodierte Duplexsonographie. Thieme, Stuttgart 2000

   Malformations and Developmental Abnormalities

General Pathology and Neurology

H. Traupe and S. Trittmacher

• Holoprosencephaly

• Septo-optic dysplasia

• Midline cysts

• Cystic malformations of the posterior fossa.

Holoprosencephaly-type malformations of the forebrain are rarely observed clinically because the affected infants die shortly after birth. When the insult occurs at a later stage of embryonic development, the malformation is less severe and there are more opportunities for clinical observation.

For example, anomalies of the corpus callosum are often clinically silent even if the entire corpus callosum is absent. Detecting such an anomaly with computed tomography (CT) or magnetic resonance imaging (MRI) should prompt a search for other, more severe malformations. The clinical neurologic findings rarely point to a specific diagnosis in children with a suspected brain malformation. The dominant features are psychomotor retardation and seizures.

Imaging Principles

H. Traupe and S. Trittmacher

The best imaging modality for detecting or excluding cerebral malformations and developmental abnormalities is MRI, whose arbitrary plane selection, good spatial resolution, and high sensitivity to the varying water and lipid contents of different brain tissues are ideal for detecting even minor structural anomalies, including changes in the cortex and white matter. MRI is superior to ultrasound and CT for detecting abnormalities in the formation and migration of neurons and glial cells during the course of brain development.

T1-weighted sagittal images are a good initial study for evaluating the midline structures of the (pediatric) brain but should be supplemented by proton density (PD)— and T2-weighted imaging, with axial scans often giving the best results. The need to add axial or coronal T1-weighted images depends on the situation. The use of heavily T1-weighted inversion recovery (IR) sequences also depends on individual requirements and may be needed, for example, to investigate the distribution of gray and white matter. Paramagnetic contrast agents are particularly useful if there is suspicion of neoplastic degeneration, as in the neurocutaneous syndromes.

Abnormalities of Ventral Induction

H. Traupe and S. Trittmacher

Holoprosencephaly

Fig. 1.35 Alobar holoprosencephaly. Axial Ctshows an undifferentiated, horseshoe-shaped soft-tissue mass surrounding a monoventricle and “floating” in a greatly expanded CSF space.

Lobar holoprosencephaly, septo-optic dysplasia. Lobar holoprosencephaly is the least severe form. The septum pellucidum is always absent, and there may be fusion of the frontal lobes with dysplasia of the frontal horns and rostral falx. Often this condition is indistinguishable from septooptic dysplasia (absence of septum pellucidum, deformity of frontal horns, hypoplasia of optic nerves and chiasm), which some consider to be the mildest form of holoprosencephaly.

Neuroepithelial Cysts

These rare cysts require differentiation from arachnoid cysts. They are usually supratentorial and rarely infratentorial. Supratentorial cysts are typically located in the wall of the lateral ventricles but may also occur deeper in the brain tissue. Infratentorial lesions are usually located in the fourth ventricle. Because the cysts contain cerebrospinal fluid (CSF)-like fluid, intraventricular neuroepithelial cysts are usually detected only if they are large enough to cause gross distortion of ventricular shape and obstructive hydrocephalus. On T1-weighted MRI, they may appear slightly hyperintense to CSF because of their high protein content, and this will facilitate the diagnosis. Paraventricular cysts are always easy to identify, even with CT.

These rare neuroepithelial cysts should not be confused with two normal variants that are more common:

Both of these variants are congenital cavities that contain CSF and extend posteriorly below the corpus callosum and between the two layers of the septum pellucidum. Very rarely, a cavum septi pellucidi may develop into a septum pellucidum cyst that produces a mass effect and clinical symptoms. A persistent cavum veli interpositi is occasionally seen above the roof of the third ventricle and below the splenium of the corpus callosum, especially in children. Again, this feature represents a clinically insignificant normal variant.

Dandy—Walker Malformation

Fig. 1.36 Dandy—Walker malformation. the large cyst, which communicates broadly with the fourth ventricle, has caused enlargement of the posterior fossa with upward displacement of the tentorium. the cerebellar vermis is hypoplastic, and the corpus callosum is completely visualized. Mild hydrocephalus is present. T1-weighted sagittal MRI.

Fig. 1.37 Infratentorial (retrocerebellar) arachnoid cyst. Midline cystic mass with compression and anterior displacement of the cerebellum and brainstem. T1-weighted sagittal MRI.

• Syringohydromyelia

• Obstructive hydrocephalus

The farther the cerebellar tonsils extend into the spinal canal along the posterior aspect of the medulla oblongata and upper cervical cord, the more elongated and tongue-like they appear. They rarely extend past the atlas.

Fig. 1. 39a, b Chiari II malformation.

a Elongated cerebellar tonsils and biventral lobule of the cerebral hemispheres extend deep into the cervical canal (well below a reference line from the posterior rim of the expanded foramen magnum to the hard palate). Note the beak-shaped tectum, callosal dysgenesis, and dysmorphic widening of the medial parieto-occipital cerebral fissures. T1-weighted sagittal MRI.

b Aplasia of the falx cerebri with interdigitation of the hemispheres across the midline. T1-weighted axial MRI.

The diagnostic modality of choice is MRI. The typical imaging appearance of syringobulbia is that of an eccentric, slit-like cavity of CSF signal intensity located within the medulla oblongata. With syringomyelia, the cord is often expanded and shows incomplete, haustra-like septations. Glial reactions are frequently observed at the upper and lower end of the cavity (high signal intensity in more T2-weighted images). Whenever a syrinx is found, a spinal tumor or a tumor at the foramen magnum should be excluded as the cause (using contrast administration in doubtful cases). Other important findings are:

• Chiari I or II malformation

• Tethered cord

• Disturbances of spinal CSF pulsation indicative of arachnoiditis.

Developmental Abnormalities of the Cerebral Cortex, Corpus Callosum, and White Matter

W. Grodd

Abnormalities of Proliferation and Migration

Fig. 1.40a, b Lissencephaly type I.

a T1-weighted axial MRI.

b T2-weighted axial MRI.

Note the abnormally smooth occipital surface and the thickened cortical ribbon with an interposed layer of white matter (arrow).

Polymicrogyria

Because the outer cortical neurons are the last to migrate to the cortex, it is assumed that polymicrogyria is based on postmigratory cellular damage caused by hypoxic-ischemic events or a maternal cytomegalovirus infection. This leads to the formation of multiple small gyri in cortical areas of variable size, which often show anomalous venous drainage. Polymicrogyria can affect any lobe of the brain but is often seen around the sylvian fissure. Symptoms depend on the location and extent of the cortical dysplasia; seizures and motor deficits are common. The small gyri are easily missed on magnetic resonance (MR) images, and in rare cases they are calcified.

Fig. 1.44a—d Dysgenesis of the corpus callosum.

a T1-weighted sagittal MRI.

b T1-weighted axial MRI.

c T1-weighted coronal MRI.

d T2-weighted axial MRI.

Note the absence of the cingulate gyrus in a, the steerhorn shape of the frontal horns in c, and the parallel configuration of the lateral ventricles in b and d.

Callosal dysgenesis is clearly visualized on axial, sagittal, and coronal MR images.

Besides the callosal anomaly itself, other characteristic features are present:

• Steerhorn shape of the frontal horns on axial or coronal images,

• Parallel arrangement of the lateral ventricles on axial images,

• Upward extension of the third ventricle into the interhemispheric fissure.

It is important to distinguish callosal dysgenesis from callosal hypoplasia, which is mainly seen in myelination disorders. Hypoplasia is associated with a thinned but complete corpus callosum that is most clearly visualized on sagittal images.

Developmental Abnormalities of the White Matter

Delayed Myelination

Delayed myelination occurs predominantly in children with hypoxic-ischemic brain injury and in hydrocephalus. Vitamin B₁₂ deficiency and persistent seizures can also lead to a delay in myelination. During the firstyear of life, this delay is best demonstrated by T1-weighted MRI and should affect all of the supratentorial white matter equally.

Periventricular Leukomalacia

Periventricular leukomalacia (PVL) is the stereotypical brain injury occurring in premature infants in response to hypoxia, ischemia, or germinal matrix hemorrhage. Various underlying causes have been cited:

• Less oxygen saturation of the blood of the immature lung,

• High metabolic activity during myelination,

• Lack of autoregulation with insufficient vasodilative capacity of the cerebral vessels, making the periventricular white matter more susceptible to ischemia.

Fig 1.45a—d Periventricular leukomalacia.

a T1-weighted sagittal MRI.

b T1-weighted axial MRI.

c T2-weighted coronal MRI.

d T2-weighted axial MRI.

Note the thinned corpus callosum in a and the expanded occipital horns in b, with increased whitematter signal intensity extendin gto the ventricular walls in c and d.

Fig. 1. 46a, b Parasagittal whitematter injury.

a T1-weighted sagittal MRI.

b T2-weighted axial MRI.

Note the parasagittal distribution pattern of the lesions, which spares a thin subependymal rim (arrow).

M. Hermes

The neurocutaneous syndromes, or phakomatoses, are a heterogeneous group of congenital systemic disorders that are associated with dysplasias and neoplasias of tissues of neuroectodermal and mesenchymal origin. The modes of inheritance have been established for the most important neurocutaneous syndromes, and these diseases will be discussed below.

Neurofibromatosis

Neurofibromatosis (NF) is the most common neurocutaneous syndrome. It is classified genetically and clinically into two subtypes:

NF-1 (classic von Recklinghausen disease). NF-1 predisposes to intracranial tumors such as gliomas, especially of the anterior visual tract. Other common brain tumors are diencephalic and brainstem gliomas, consisting mainly of pilocytic astrocytomas, and also gangliogliomas. Additional features of NF-1 are:

• Changes in skin pigmentation,

• Cutaneous fibromas,

• Cranial dysplasias such as absence or dysplasia of the greater sphenoid wing with orbital herniation of temporal lobe tissue, characteristic defects in the calvaria,

• Vertebral dysplasias such as kyphoscoliosis and posterior scalloping of vertebral bodies,

• Dysplasias of other structures of mesenchymal origin.

Fig. 1. 47 NF–1. Plexiform neurofibroma of the right cheek and sphenoid wing dysplasia on the right side with exophthalmos.

a T1-weighted axial MRI after paramagnetic contrast administration.

b 3-D reconstruction of the skull base from CT data (same patient, superior view).

Optic gliomas lead to elongation and thickening of the optic nerve and expansion of the optic chiasm. They frequently enhance on postcontrast MRI and CT scans. Usually only MRI can reliably detect postchiasmatic involvement. MRI can also demonstrate hamartomas, which appear as areas of increased signal intensity on T2-weighted images located in the visual tract, the periventricular cerebral or cerebellar white matter, basal ganglia, and dentate nucleus. The morphologic substrate of these changes is still uncertain. Apparently they are not tumor precursors, because they have not been found to undergo neoplastic transformation and they frequently regress.

The neurofibromas that characterize NF-1 are tumors of fibroblasts, myelin sheaths, and nerve cells.

Fig. 1.48a—d Tuberous sclerosis. MRI reveals multiple bilateral cortical tubers, especially in the frontoparietal areas, and numerous calcified subependymal nodules on the walls of the lateral ventricles.

a Noncontrast axial CT.

b T1-weighted axial MRI.

c T2-weighted axial MRI.

d T1-weighted axial MRI.

Fig. 1.49 Hippel—Lindau disease. Nodular tumor (hemangioblastoma) in the upper partof the cerebellum shows intense, homogeneous enhancement. Vertebral angiogram, lateral projection.

Noncommunicating Hydrocephalus

In approximately 70% of congenital forms of hydrocephalus, the obstruction is located within the ventricular system.

Arachnoid cysts are fluid-filled cavities that result from duplication or splitting of the arachnoid membrane. It is common to find dysplasia of adjacent brain areas, such as partial aplasia of the temporal lobe associated with arachnoid cysts in the middle fossa, but it is unclear whether this is a cause or effect of the cysts.

Many syndromes are characterized by ocular and orbital involvement, including:

• Hemifacial microsomia

• Oculoauricular dysplasia (Goldenhar syndrome)

• Mandibulofacial dysostosis (Treacher-Collins-Franceschetti syndrome)

• Pierre-Robin syndrome.

Details on these syndromes can be found in textbooks on pediatric neurology and pediatric neuroradiology.

Cranial and Craniospinal Malformations

Microcephaly is usually an expression of severe brain damage, while macrocephaly reflects an abnormally raised intracranial pressure. When a cranial suture closes prematurely, cranial growth at right angles to the suture is blocked or suppressed, resulting in a bony stenosis with compensatory growth. Premature synostosis of the coronal suture leads to decreased longitudinal skull growth with flattening of the frontal calvarial convexity and compensatory growth along the sagittal and lambdoid sutures (turricephaly). Premature closure of the sagittal suture leads to increased longitudinal growth at the coronal and lambdoid sutures, resulting in a long, narrow skull (scaphocephaly). Premature closure of the coronal and lambdoid sutures results in a short skull with compensatory growth toward the vertex (acrocephaly). In principle, premature synostosis can affect any cranial suture or a portion of it. Because the coronal suture is most often affected, turricephaly is the most common form of craniosynostosis. CT is the preferred imaging modality, and three-dimensional (3-D) surface reconstructions supplement the standard workup by showing the cranial deformity in its entirety, while permitting an analysis of the synostosis itself and the accompanying “crest. ”

Fig. 1.54 Ectopic neurohypophysis in a 12-year-old boy of shortstature. the pituitary stalk is extremely narrow, and a nodular soft-tissue mass of high signal intensity is visible on the floor of the third ventricle. The sella is relatively small and contains only the adenohypophysis; the normally hyperintense neurohypophysis is absent. T1-weighted sagittal MRI.

According to Tessier, craniofacial malformations result from the incomplete fusion of clefts combined with the premature fusion of embryonic growth zones. They lead to various deformities of the facial skeleton. The best known condition is Crouzon disease, a syndrome characterized by craniosynostosis with brachycephaly, maxillary hypoplasia, and exophthalmos.

Bony deformities of the craniovertebral junction consist predominantly of:

• Dysplasias

• Dysraphias

• Segmentation anomalies.

Fig. 1.56a, b Cortical dysplasia. Cortical thickening (arrows) with circumscribed high signal (arrow head) is noted in the right frontal area. the high signal is clearly distinguishable from CSF only on the PD-weighted image.

a PD-weighted axial MRI.

b T2-weighted axial MRI.

In patients with temporal lobe epilepsy, which accounts for approximately 80% of all focal epilepsies, axial slices should be angled parallel to the long axis of the temporal lobes or hippocampi, and coronal slices should be perpendicular to that axis. Generally it is best to have the highest spatial and contrast resolution possible. If Mr images are negative, CT should be performed since small calcified lesions may be missed on MRI.

• Astrocytoma

• Oligodendroglioma

Cerebral angiography has no diagnostic role except in patients with vascular malformations. It does form the basis for the Wada test, however, in which a barbiturate is injected into the internal carotid artery to produce short-term anesthesia of one hemisphere in order to identify the dominant speech hemisphere and suppress a presumed epileptogenic focus. Some cases also require selective Wada testing in which a microcatheter is used to inject the drug into peripheral cerebral vessels.

References

Ashwal, S.: Congenital structural defects. In Swaiman, K. F.: Pediatric Neurology, Vol. I, 2^nd ed. Mosby, St. Louis 1994 (pp. 421–470)

Barkovich, A. J., B. O. Kjos, D. Norman, M. S. Edwards: Revised classification of posterior fossa cysts and cystlike malformations based on the results of multiplanar Mr imaging. Amer. J. Neuroradiol. 10 (1989) 977–988

Barkovich, A. J., T. K. Koch, C. L. Carrol: The spectrum of lissencephaly: report of ten patients analyzed by magnetic resonance imaging. Ann. Neurol. 1991 (1991)139–146

Barkovich, A. J., B. O. Kjos: Gray matter heterotopias: Mr characteristics and correlation with developmental and neurologic manifestations. Radiology 182 (1992a) 493–499

Barkovich, A. J., B. O. Kjos: Nonlissencephalic cortical dysplasias: correlation of imaging findings with clinical deficits. Amer. J. Neuroradiol. 13 (1992b) 95–103

Barkovich, A. J., B. O. Kjos: Schizencephaly: correlation of clinical findings with Mr characteristics. Amer. J. Neuroradiol. 13 (1992c) 85–94

Barkovich, A. J., P. Gressens, P. Evrard: Formation, maturation, and disorders of brain neocortex. Amer. J. Neuroradiol 13 (1992d) 423–446

Barkovich, A. J.: Pediatric Neuroimaging, 2^nd ed. Raven, New York 1995

Barth, P. G.: Migration disorders of the brain. Curr. Opin. Neurol. Neurosurg. (1992) 339–343

Brant-Zawadzki, M., D. Norman: Magnetic Resonance Imaging of the CNS. Raven, New York 1986

Dietrich, R. B., S. El-Saden, H. T. Chugani, J. Bentson, W.J. Peacock: Resective surgery for intractable epilepsy in children: radiologic evaluation. Amer. J. Neuroradiol. 12 (1991)1149–1158

Fahrendorf, G., J. Brämswig, M. Bals-Pratsch: MR-Befunde bei Patienten mit „idiopathischem” Panhypopituitarismus. Fortschr. Röntgenstr. 152 (1990) 550–555

Friede, R. L: Developmental Neuropathology. Springer, Berlin 1975

Galassi, E., G. Giast, G. Giuliani, E. Pozzati: Arachnoid cysts of the middle cranial fossa: experience with 77 cases treated surgically. Act a neurochir. Suppl. 42 (1988) 201–204

Grodd, W.: Kernspintomographie neuropädiatrischer Erkrankungen. Normale Reifung des kindlichen Gehirns. Klin. Neuroradiol. 3 (1993)13–27

Hermes, M., G. Huber, C. Kujat, K. Neidl: Das juvenile pilozytische Astrozytom bei Neurofibromatose 1. Klin. Neuroradiol. 3 (1993) 20–22

Huber, G., M. Voges, E. Donauer, G. Vince, K. Neidl: Intrakranielle Arachnoidalzysten. Neurologische Klassifizierung und operative Behandlungsergebnisse. Klin. Neuroradiol. 3 (1993) 23–28

Kiefer, M., R. Eymann, S. von Tiling, A. Muller, W. I. Steudel, K. h. Booz: The ependyma in chronic hydrocephalus. Child’s nerv. Syst. 6 (1998) 263–270

Kiefer, M., R. Eymann, M. Voges, W. I. Steudel: the meaning of hydrostatic valves in hydrocephalus therapy. Zbl. Neurochir. Suppl. 1 (1999) 10–11

Kuzniecky, R., F. Andermann, R. Guerrini et al.: Congenital bilateral perisylvian syndrome: study of 31 patients. Lancet 341 (1993) 608–612

McConnell, S. K.: Development and decision making in the mammalian cerebral cortex. Brain Res. Rev. 1988 (1988)1–23

Mori, K.: Anomalies of the Central Nervous System: Neuroradiology and Neurosurgery. Thieme, Stuttgart 1985

Moss, M. L, L. Salentijn: the primary role of functional matrices in facial growth. Amer. J. Orthodont. 55 (1969) 566–577

Naidich, T. P., R. A. Zimmermann: Common congenital malformations of the brain. In Brant-Zawadzki, M., D. Norman: Magnetic Resonance Imaging of the Central Nervous System. Raven, New York 1987

Osborn, A. G.: Diagnostic Neuroradiology. Mosby, St. Louis 1994

Ostertun, B., L. Solymosi, M. Kurthen, C.E. Elger, J. Schramm: Neuroradiologische Methoden und Befunde vor epilepsiechirurgischen Eingriffen im Kindesalter. Radiologe 33 (1993)189–197

Ostertun, B., L. Solymosi, J. Zentner, H. K. Wolf, C.E. Elger, J. Schramm et al.: MR- und CT-Charakteristik a glioneuronaler Hamartien. Klin. Neuroradiol. 4 (1994) 34–40

Ostertun, B., H. K. Wolf, M. G. Campos, C. Matus, L. Solymosi, C.E. Elger et al.: Dysembryoplastic neuroepithelial tumors: Mr and Ctevaluation. Amer. J. Neuroradiol. 17 (1996) 419–430

Ostertun, B.: Radiologische Diagnostik bei Epilepsie. Fortschr. Roentgenstr. 170 (1999) 235–245

Piepgras, U.: Neurokutane Systemerkrankungen. Thieme, Stuttgart 1984

Sato, Y.,S. C. S. Kao, W. L. Smith: Radiographic manifestations of anomalies of the brain. Radiol. Clin. n. Amer. 29 (1991) 179–194

Tessier, P.: Anatomical classification of facial craniofacial, and latero-facial clefts. J. max.-fac. Surg. 4 (1976) 69–92

Van Bogaert, P., P. David, C. A. Gillian et al.: Perisylvian dysgenesis. Clinical, EEG, MRI and glucose metabolism features in 10 patients. Brain 121 (1998) 2229–2238

Walter, M., M. Kiefer, W. I. Steudel, S. Leonhardt, R. Isermann: Modellbildung biologischer prozesse am Beispiel des Hirndrucks. Biomed. Techn. 43 (1998) 286–287

Wolf, H. K., M. G. Campos, J. Zentner, A. Hufnagel, J. Schramm, C. E. Elger et al.: Surgical pathology of temporal lobe epilepsy. Experience with 216 cases. J. Neuropathol. exp. Neurol. 52 (1993) 499–506

Zentner, J., A. Hufnagel, H. K. Wolf, B. Ostertun, E. Behrens, M. G. Campos et al.: Surgical treatment of temporal lobe epilepsy: clinical, radiological and histopathological findings in 178 patients. J. Neurol. Neurosurg. Psychiatry 58 (1995) 666–673

Zentner, J., H. K. Wolf, B. Ostertun, A. Hufnagel, M. G. Campos, L. Solymosi et al.: Gangliogliomas: clinical, radiological and histopathological findings in 51 patients. J. Neurol. Neurosurg. Psychiat. 57 (1994) 1497–1502

   Traumatic Lesions

General Pathology and Neurology

B. Haubitz

Brain injuries are classified as follows according to whether or not the dura mater has been breached:

This classification takes into account pathomorphologic and functional aspects, including the therapeutic measures that are derived from them. Because the brain is enclosed within a firm capsule composed of the meninges and neurocranial bone, the various patterns of injury result from the biomechanical interactions that take place between the brain and its coverings. Trauma can cause brain edema, which may be diffuse or focal depending on the mechanism of the injury. Focal edema also occurs after vascular injuries that are associated with a disruption of arterial or venous blood flow. Vasogenic edema typically shows a territorial distribution.

If the patient survives the trauma initially, numerous factors determine the prognosis, including:

• Hypoxemia due to blood loss and aspiration,

• Increased intracranial pressure due to brain edema,

• Inflammatory complications.

Brain abscesses can also develop, even after a period of years.

Other trauma-related conditions that are directly or indirectly life-threatening include:

• Pulmonary complications due to long-term ventilation,

• Pulmonary emboli originating from pelvic and lower extremity venous thrombosis,

• Cerebral seizures.

More is known about the effects of trauma on the gray matter (cerebral cortex, basal ganglia) than on the white matter. Many, but not all head injuries that are associated with neurologic deficits leave behind morphologic traces that are detectable by modern imaging methods.

Imaging Principles

B. Haubitz

Careful observation of the patient is essential for the timely and appropriate use of neuroimaging studies.

For example, it can be disastrous to withhold close clinical observation because a patient has normal-appearing skull films. The trauma history, posttrauma history, and current neurologic findings are critical in determining the further course of action. Special care should be taken when an adequate history is not available. Close cooperation with the neurologist and neurosurgeon is essential for an optimum neuroradiologie evaluation.

Improvements in emergency rescue procedures, along with the increased availability of CT, have substantially improved the survival rates and quality of life of patients with head injuries. The outstanding role of CT in primary diagnosis and follow-up is undisputed. Increasingly, however, MRI is also being used to evaluate craniocerebral trauma, particularly in the subacute phase and when it is necessary to resolve discrepancies between a trauma patient’s neurologic status and CT findings. It is unnecessary to perform CT in every patient admitted with a diagnosis of head trauma, as this would be costly and inefficient. Primary CT should be reserved for cases of acute craniocerebral trauma where there is impairment of consciousness not obviously attributable to substance use, where clouding of consciousness is observed, or where there are focal neurologic signs, a penetrating injury, or a palpable depressed fracture. CT is also used on a case-by-case basis following trauma that cannot definitely be classified as trivial.

With improvements in the transport of seriously injured patients, CT is sometimes performed so early that imaging abnormalities are notyet seen, despite the clinical presence of neurologic deficits. Contusional hemorrhages, for example, may not be detectable until 1 or 2 days after the injury. It should also be considered that traumatic extra-axial hematomas may not appear to require surgical evacuation based on initial CT findings but can undergo considerable expansion over the next several hours.

Craniocerebral Injuries

B. Haubitz

Brain Injuries

Closed craniocerebral trauma. A brain contusion (cortical contusion) is the most frequent diagnosis in patients with closed head injuries. The acute detection of contusional hemorrhage is usually accomplished with computed tomography (CT). The advantage of CT lies in the concomitant detection of bony injuries. One disadvantage is its limited value at the base of the brain due to partial volume averaging and bone-induced artifacts.

Blunt craniocerebral trauma. In the first systematic studies on the use of magnetic resonance imaging (MRI) in blunt head trauma, diffuse axonal lesions were found in approximately half of all cases, and cortical contusions were detected with the same frequency. Manifestations of deep graymatter lesions (basal ganglia) or primary brainstem injuries were found in only 4% of cases.

Brain contusions (cortical contusions). Brain contusions are caused by a force acting over a large area of the neurocranium. The mechanism may be abrupt deceleration after a fall or massive acute acceleration of the stationary skull. The initial effect is an isolated or confluent rhexis hemorrhage in the cerebral cortex or subcortical white matter. With passage of time, tissue liquefaction occurs and culminates in a cystic parenchymal defect. The detection of postcontusional states with imaging procedures such as CT and MRI is a common issue in disability assessments. Typical sites of occurrence are the frontobasal and temporal brain regions, which are at particular risk because of their proximity to sharp bony edges and the frequency of frontal head trauma. Contusion-related parenchymal hemorrhages are also found near the falx cerebri and tentorium. Because of the brain’s own inertia, a blow to the head tends to cause injury both at the site of the impact (coup) and on the side opposite the impact (contrecoup), and focal neurologic symptoms frequently occur immediately after the trauma. Traumatic cerebral hemorrhage cannot always be distinguished from spontaneous intracranial hemorrhage by imaging alone. This problem of differential diagnosis is most likely to arise when the history is unclear.

Fig. 1.59a, b Contusional hemorrhages.

a Initial scan shows hyperdense areas in the bifrontal and left temporal regions of the brain. Axial CT.

b Scan 13 hours later shows expansion of the hemorrhagic areas with the developmentof perifocal edema. Axial CT.

Fig. 1.60a, b Hemorrhagic contusion of the right temporal lobe in the acute to subacute stage.

a Axial CT shows a superficial hypodense area of brain parenchym a with small hyperdense components.

b T2-weighted axial MRI shows considerably more extensive damage with a larger hemorrhagic component. The high signal represents the contusion and associated edema. The very low signal is from contusional hemorrhages at the stage of deoxyhemoglobin or intracellular methemoglobin.

• Gliosis with or without hemorrhagic residues in the white matter after a shearing injury,

• Porencephalic parenchymal defects, with or without enlargement of adjacent CSF spaces,

• Various forms of elevated pressure hydrocephalus.

MRI can even detect ruptures of the optic nerve or optic chiasm and lacerations of other small structures. In patients who have already had surgery, one should consider whether all or part of the visible changes are a result of the surgical procedure. Secondary effects of vascular injuries and inflammatory complications should also be considered as the potential causes of some lesions.

With an acute subdural hematoma, the severity of the traumatizing force is greater than in the subacute or chronic form. As a result, the hematomas are often combined with other primary traumatic brain injuries. Hemorrhage into the subdural space is believed to be caused by the rupture of bridging veins, usually in response to a frontal blow to the head. This mechanism of injury is particularly common in elderly patients, whose brain is more mobile due to age-related atrophy and is, therefore, more susceptible to injury than in younger individuals.

Other causes of subdural hematoma:

• Dural sinus injury (with venous hemorrhage),

• Vascular anomalies (with spontaneous hemorrhage due to a dural fistula).

While acute subdural hematomas are typically unilateral, chronic hematomas are often bilateral. Chronic subdural hematoma is thought to be caused by repeated episodes of minor bleeding from injured vessels of varying size. The entry of CSF into the blood compartment is believed to contribute to the intracranial mass effect. Some chronic subdural hematomas can be interpreted as the late stage of an acute, traumatic subdural hematoma. In other cases the chronic hematoma is classified as “primary, ” at least by its clinical presentation. If hemorrhagic pachymeningitis is present, the collection is xanthochromic rather than bloody and is termed a subdural hygroma. This type of collection is particularly common in infants and small children and in patients over 50 with a degenerative consumptive systemic disease, atrophic brain changes, or alcoholism. Trauma has no causal significance in these cases, but even trivial injuries can induce subsequent hemorrhage leading to further clinical deterioration. In this type of chronic subdural hematoma, it is common to find fibrous membranes that loculate the fluid collection.

Fig. 1.65 Acute subdural hematoma in the right frontotemporal area. Swelling of the right cerebral hemisphere has exacerbated the shift of midline structures. Axial CT.

Fig. 1.66 Film-like acute subdural hematoma on the left side of the tentorium. Axial CT.

Fig. 1.67 Traumatic hygroma in the left frontotemporal area due to bifrontal contusional hemorrhage with perifocal edema, intraventricular hemorrhage, and subdural blood in the interhemispheric fissure. Axial CT.

Subacute subdural hematomas are easy to detect on MRI, as they usually display a high signal intensity in all sequences. With the conversion of deoxyhemoglobin to methemoglobin and lysis of the red blood cells, the paramagnetic properties of the methemoglobin become a dominant factor at this stage. On reaching a chronic stage, however, the hematoma may become isointense to brain tissue on T1-weighted images, and mass effects may provide the only imaging clue to the presence of an extra-axial fluid collection. Meanwhile, the effusion usually shows homogeneous high signal intensity on T2-weighted images, unless there is recurrent fresh bleeding so that methemoglobin and deoxyhemoglobin are present in the same collection. Contrast administration usually produces marked, uniform linear enhancement of the (thickened) meninges that surround the (subacute or chronic) subdural hematoma. Dura and arachnoid distant from the hemorrhage show the same response in many cases. Paramagnetic contrast enhancement is necessary only if doubt exists, however. The differential diagnosis includes various lesions that usually have a different clinical presentation:

• Meningioma en plaque,

• Neoplastic or inflammatory infiltration of the meninges and subdural space,

• Subdural pus (empyema).

While the former lesions show intense contrast enhancement extending to their core, a subdural empyema may closely resemble a subdural hematoma. It should be added that an untreated subdural hematoma that does not undergo complete reabsorption may calcify.

Epidural Hematoma

Epidural hematomas most commonly result from motor vehicle accidents. As with subdural hematomas, males predominate by a substantial ratio (5:1). The hemorrhage usually results from the traumatic rupture of a meningeal artery (especially the middle meningeal artery or one of its branches) or, less commonly, the rupture of a dural sinus. Blood collects in the space between the dura and the inner table of the calvaria.

Since the source of the bleeding is usually arterial, epidural hematomas tend to develop swiftly and produce clinical manifestations of rapid onset, for example:

• Vomiting

• Motor unrest

• Altered mental status

• Obtundation

• Ipsilateral dilatation of the pupil (indicating oculomotor nerve compression)

• Pyramidal tract lesion with contralateral hemiparesis

• Seizures.

In contrast, an acute subdural hematoma usually has a venous source and therefore develops over a longer period. While it is generally true that a brief interval between injury and symptom onset implies an epidural hematoma while a long interval suggests a subdural hematoma, exceptions to this general rule may occur.

If calvarial injury is present, the hematoma is generally located on the side of the fracture. CT or MRI will often disclose a scalp hematoma over the site of the injury. Epidural hematomas most commonly occur in the temporal and parietal areas. Frontal, occipital, and infratentorial locations are less common, and hematomas at these sites are usually smaller. Bilateral epidural hematomas are unusual. Occasionally an epidural hematoma coexists with a subdural hematoma or a traumatic subarachnoid hemorrhage. Epidural hematomas are relatively rare in children and the elderly, because the dura is more firmly attached to the calvaria in these age groups than in middle-aged persons.

With frontobasal skull fractures, there is a danger of traumatic CSF leak if the fracture lines enter the walls of the sphenoid sinus or ethmoid labyrinth. CSF leak can also occur with temporal bone fractures that involve the mastoid. The patient complains of fluid discharge from the nose, and recurrent bouts of bacterial meningitis can develop in severe cases. Plain skull films cannot demonstrate a CSF leak, but they may show a large intracranial air collection (pneumocephalus) signifying an abnormal communication between the CSF space and the environment. The best way to detect CSF leaks, which can also occur spontaneously, is by CT cisternography. In this study several milliliters of a myelographic contrast agent are injected intrathecally via lumbar puncture, and the table is moved to a head-down tilt to opacify the intracranial subarachnoid space. CT examination is then performed in the prone position using coronal scanning and high-resolution technique (HR CT). In the most favorable case, the study will demonstrate the bone defect in the skull base along with contrast leakage from the basal subarachnoid space into the nose, paranasal sinuses, or mastoid air cells. Recently, MRbased methods have also become available for the detection of CSF leak.

Injuries of the Facial Skeleton

G. Albrecht

Trauma to the facial skeleton can cause individual or combined fractures of the various bones of the midfacial region.

• Occipitofrontal and occipitonasal views to demonstrate the orbits,

• Submentovertical projection to demonstrate the zygomatic arches (“jug-handle view”),

• Panoramic tomogram (old term: orthopantomogram) to survey the maxilla and mandible.

But CT (preferably HR CT) is the imaging modality of choice for demonstrating midfacial injuries. Direct coronal scans are advantageous for evaluating fractures of the orbital floor or orbital roof and frontal skull base. Before the patient is positioned for coronal scanning (prone or supine with the head maximally extended), however, it is necessary to exclude cervical spinal injuries. The advantage of CT over conventional tomography is that it also defines the soft tissues and entails low radiation exposure. One drawback is that direct coronal scanning cannot be performed in seriously injured patients.

Central or pyramidal fractures. These are pyramid-shaped midfacial fractures in which the maxilla is separated from the rest of the facial skeleton. This group includes:

• Solitary nasal bone fracture or nasoethmoid fracture,

• Le Fort II fracture (through the nasal root, medial orbital wall including the lacrimal bone and duct, medial orbital floor, facial/posterolateral antral wall, and pterygoid process: the nasal bones may be spared).

Fig. 1.72 Le Fort I–III fractures. Bilateral transverse fractures of the maxill a with horizontal separation of the nasal and antral floor (Le Fort I fracture), bilateral fractures of the orbital floor and lamin a papyrace a (Le Fort II fracture), transverse fracture of the nasal root, and bilateral fractures through the frontozygomatic suture (Le Fort III fracture). Additional fractures involve the frontal process of the maxill a on both sides and the lateral orbital wall on the left side. the fractures are accompanied by bilateral maxillary hematosinus and intraorbital air collections on the left side. Foreign bodies (glass fragments) are embedded in the buccal soft tissues, especially on the left side. 3-D volume reconstruction.

Centrolateral fractures. These involve the separation of the facial skeleton from the skull base. They include:

• Le Fort III fracture (through the nasal root, medial and lateral orbital walls, zygomatic arch, and pterygoid process with separation of the frontozygomatic suture; may involve the orbital floor and medial antral wall instead of the nasal root).

Lateral fractures. These mainly affect the zygoma. In addition to an isolated zygomatic arch fracture, the orbital floor is always involved. An important type of lateral midfacial injury is:

General Pathology and Neurology

C. Zimmer and H. Traupe

Brain tumors are the third most common class of tumor after gastric and lung cancer, ranking with neoplasias of the hematopoietic/lymphatic tissue and colon. They are the second most common tumor group in children after the leukemias. Various systems have been devised for the classification of brain tumors. A widely used system is the 1993 revision of the World Health Organization (WHO) classification, which is based on tumor histogenesis, or the cells from which tumors originate. All cell types that occur in the brain can undergo neoplastic degeneration and form tumors. Additionally, every brain tumor is graded based on its biologic behavior. Four histologic grades (I-IV) are recognized in the WHO system:

Each of the WHO tumor grades is associated with a prognostic implication following complete macroscopic tumor removal:

Brain tumors. Brain tumors have special pathologic significance because even biologically benign forms can be extremely “malignant” because of their location in or adjacent to functionally important brain areas and the lack of space available for the displacement of brain parenchyma.

Because of this, a glioblastoma cannot be removed completely on both a macroscopic and microscopic level.

The majority of brain tumors are primary to the brain, but up to 40% of brain tumors detected in adults are found to be metastatic. While 70% of brain tumors in adults arise at supratentorial sites, infratentorial tumors are more prevalent in children. Many types of brain tumor have sites of predilection and characteristic incidences in different age groups. Nearly all types of brain tumor (except meningiomas and neurinomas) are more common in males than females. Brain tumors metastasize primarily along cerebrospinal fluid (CSF) pathways, usually to another site within the central nervous system (CNS), but occasional metastasis to locations outside the CNS has been described for almost all brain tumors.

The clinical manifestations depend on various factors, including:

Location and size of the mass

Tumor histology

Perifocal reaction.

The following initial symptoms are common:

Headache

Cerebral seizures

Altered mental status.

Later symptoms may include:

Signs of increased intracranial pressure,

Focal cerebral symptoms.

Very rarely, the tumor type can be inferred from its clinical manifestations, as in the case of cerebellopontine angle tumors.

The cause of brain tumors is still uncertain. Recent studies suggest that certain tumors, particularly gliomas, result from the mutation of numerous genes that regulate cellular proliferation and differentiation.

• Gene therapy

• Immunotherapy

• Antiangiogenesis therapy.

Imaging Principles

C. Zimmer and H. Traupe

The two most important imaging modalities for brain tumor detection are computed tomography (CT) and magnetic resonance imaging (MRI). Plain skull films and cerebral angiograms are useful today only as adjuncts.

Plain skull films. The following signs are indicative of raised intracranial pressure on plain skull radiographs:

• Widened cranial sutures and increased convolutional markings in children,

• Demineralization and cortical thinning of the dorsum sellae (caused by slow-growing tumors that expand the sella) in adults.

Cerebral angiography. About the only application of cerebral angiography today is to define tumor vascularity prior to surgery or to direct the preoperative devascularization of hypervascular tumors. The following are nonspecific angiographic tumor signs:

• Vascular displacement

• Neovascularization

• Changes in vascular calibers.

CT, MRI. CT and MRI often have complementary roles in the evaluation of brain tumors. Whenever a tumor is suspected, these modalities should be performed with and without contrast administration whenever possible. A radiologic tissue diagnosis employs basic criteria such as:

• Tumor location (including intra-axial/extraaxial differentiation),

• Prevalence of the tumor being considered,

• Patient’s age.

Additional important criteria are:

• Tumor size

• Tumor margins

• Perifocal reaction

• Unifocal or multifocal tumor occurrence

• Mass effect from the tumor

• Internal structure of the tumor, including necrosis, cysts, calcifications, and vascularity.

Contrast administration increases the sensitivity for tumor detection (especially small basal neoplasms) andyields information on the integrity of the blood-brain barrier (BBB). Contrast enhancement has two components that are frequently combined:

• Intravascular component, reflecting the increased vascularity of the tumor tissue,

• Extravascular component, reflecting a breakdown of the BBB.

It should be emphasized that tumor enhancement does not define the tumor boundaries, because tumor cells can be found a considerable distance past the margins of areas that show pathologic enhancement.

Functional imaging techniques. Besides static CT and MRI techniques that mainly depict morphologic details, dynamic and functional imaging techniques are increasingly being used in the evaluation of brain tumors. For example, functional magnetic resonance imaging (fMRI)