The cranium and the brain

Modalities for examinaton

Skull radiography

Radiographic examination of the skull has significantly diminished after the introduction and diffusion of CT and MRI. There are, however, many conditions that can be diagnosed by skull radiography directly or indirectly. Intracerebral calcifications can be normal as in the pineal gland and in the choroid plexus or pathologic as in infectious diseases, e.g. toxoplasmosis, cytomegalic inclusion disease and cysticercosis. Tumour calcifications are most frequently identified in oligodendrogliomas, meningiomas and craniopharyngeomas. Lateral displacement of the calcified pineal gland has been used as a sign of intracranial mass effect. Decalcification of the sella turcica can be a general sign of increased intracranial pressure. Enlargement of the sella turcica indicates a pituitary adenoma.

Osteolytic lesions occur in many diseases, both benign and malignant. Epidermoid, eosinophilic granuloma, multiple myeloma and metastases are some examples. Sclerotic lesions can represent meningiomas, osteomas, fibrous dysplasia, Paget's disease or metastases from prostatic carcinoma.

Skull fractures are usually of linear type. In depressed skull fractures tangential films are essential for determination of the degree of depression. A pineal shift indicates an intracranial mass effect and is a significant finding.

Computed tomography

When computed tomography was first presented to the scientific community in April 1972, the acquisition of two slices in the axial plane required more than 4 minutes, the matrix size was 64x64, and it was necessary to place a water bag around the head to allow the evaluation of subtle density differences within the head by the computer.

Nevertheless, it was immediately clear that a revolutionary diagnostic tool had become available to neuroradiologists; pneumoencephalography and ventriculography rapidly became extinct.

Nowadays, modern spiral CT acquires serial images of the whole head in less than one minute and the resolution has increased up to matrix size of 1024x 1024. Three dimensional reconstructions (3D) may be obtained routinely and almost instantaneously. CT angiography has become a reality following intravenous injection of contrast.

The basic physical principle of CT, however, has not changed: this technique is based on x-ray absorption and the capability to measure and localise different absorption coefficients exhibited by tissues with different atomic numbers.

In the MRI era, CT remains essential in the neuroradiological workup of almost all disease entities mainly because of its speed and accessibility. Patients can easily be monitored and even uncooperative patients may be examined.

In trauma cases CT is the first modality; in all patients with sudden, severe neurological deficit a CT scan provides the first most useful answers on the presence or absence of a surgically treatable condition.

MRI follows only if a more subtle evaluation of parenchymal abnormalities, a better topographical definition, or further attempts at gross pathological tissue characterisation are required. In a modem neuroradiological department, a high quality CT system is a minimum requirement, possibly with a second less sophisticated unit mainly for trauma, emergency cases and follow-up.

Magnetic resonance imaging

Magnetic resonance imaging has been used in neuroradiology for more than ten years. The method has come to be used widely during the past few years and its value in diagnostic neuroradiology is significant. MRI is superior to CT in the assessment of lesions in the white substance of the brain. In addition, it is the examination of choice in posterior fossa lesions, as well as in the examination of those lesions situated near the mid-line or the base of the skull, The main disadvantage of the method is the difficulty in visualizing fresh blood and calcification. For diagnostic neuroradiology access to both CT and MRI is essential. The brain's normal anatomy as shown by MRI is illustrated in Fig. 1.

It is important to be aware of the fact that MRI is mainly a morphological technique despite the fact that it is often claimed that the technique has great potential in the characterisation of tissues. This type of analysis has been found to be difficult and the technique's possibilities in this sphere are limited. This means that MRI still does not distinguish itself markedly from CT. As a purely morphological method is has, however, certain advantages over CT and provides extremely good analysis

of anatomical detail in all three planes. Other advantages of MRl include high tissue contrast without the use of additional contrast media and also the absence of artefacts caused by bone and air. If the patient is unable to lie still, serious movement artefacts result. Aneurysm clips made of magnetic material represent an absolute contraindication to examination by MRl.

The grey matter of the brain contains 10-15 % more water than the white matter while the white matter contains relatively more lipids. This difference in content contributes to the high contrast difference demonstrated by MRl between grey and white matter. Compared to white matter, grey matter has longer, and CSF much longer, relaxation times both with T1 and T2. Because pathological processes in the central nervous system in the main exhibit increased water content (oedema) and because MRI is more sensitive than CT in the demonstration of water content increase, MRI has high sensitivity. However, specificity is still a problem. Measurement of the relaxation times on T1 and T2 has not increased specificity to any great degree because there is significant overlapping of values thus obtained in different types of pathological processes. For example, it can still be difficult to differentiate between benign and malignant tumours and between tumours and inflammation. The paramagnetic contrast medium Gadolinium-DTP A can improve specificity but gives more or less the same type of information as contrast medium-enhanced CT.

Within the central nervous system, magnetic resonance spectroscopy (MRS) has still not established itself clinically. The reasons are several, including difficulties in precisely locating the lesion, poor signal-scatter ratios and low sensitivity.

Angiography

The basic principle of angiography has not changed since 1927 when Egas Moniz, a Portuguese neurosurgeon, first presented it by performing direct puncture of the carotid artery in the neck. Angiography, however, is nowadays performed almost universally via the femoral artery and the selective catheterization and injection of the arteries of interest. The images are electronically acquired with digital systems and subsequently photographed onto x-ray film (Digital Subtraction Angiography, DSA).

The technique of femoral puncture was first described by Seidinger in the 1950s; an introducer is nowadays usually placed in the right femoral artery and a preshaped catheter, usually 5 French, is directed under fluoroscopic control in the supra-aortic vessels.

Angiography remains necessary in many conditions but it is usually preceded by CT and/or MR.

Angiography must always be performed in the case of subarachnoid hemorrhage, or when an AVM must be defined in all its aspects (feeding arteries, nidus, draining veins) before surgery or interventional procedures can be performed. Angiography is frequently needed in the preoperative evaluation of tumours, particularly meningiomas, in case of arthritis or when the vessels in the neck must be demonstrated before surgery or angioplasty for atherosclerotic disease.

Magnetic resonance angiography (MRA) will probably further reduce the need for catheter angiography; it seems, however, that this technique will always remain in the neuroradiological armamentarium, particularly because of the growth of intravascular interventional techniques, whose indications are expanding to include the treatment of subarachnoid aneurysms.

Nuclear medicine methods

Nuclear medicine techniques using isotopes give primarily functional information and, because spatial resolution is not nearly as good as with

Figure 2. Cerebral blood flow with SPECT in the investigation of epilepsy. Hypoperfusion in the left temporal lobe (arrow).

CT and MRI, morphological information is limited. They are capable, however, of identifying the region of interest. The value of isotope methods as compared with CT and MRI varies depending on the type of lesion being examined. There are examples of lesions which can only be visualized with these functional methods. Of course, there are also examples of the opposite. Isotope techniques are becoming increasingly important as a result of the development of tracing molecules and other technical advances.

SPECT (single photon emission computed tomography) examinations are performed with a gamma camera. The radiopharmaceuticals are lipophilic and because of this pass the intact blood-brain barrier, and are extracted from the blood into the brain substance in proportion to blood flow in the region (rCBF). Blood flow is considered to be related to metabolic activity. Both hypo- and hyperperfusion can be demonstrated. Clinically SPECT is used in the localization of epileptic foci (Fig. 2), in the assessment of ischaemic conditions and in the investigation of dementia.

Basic static brain scintigraphy, in which 99m Technetium sodium pertechnetate (half-life six hours) is injected into a peripheral vein, has become less important. The technique depends on the fact that many intracerebral lesions cause a defect in the blood-brain barrier through which

Figure 3. PET with 18FDG (glucose analogue) in a patient with partial complex epilepsy. Reduced radioactive uptake indicating reduced glucose metabolism in the region of the epileptic focus in the left temporal lobe (arrow).

the radiopharmaceutical passes into the lesion, and is then detected with the gamma camera. The scintigram exhibits poor resolution, low specificity and varying sensitivity.

PET (positron emission tomography) is a complex form and extension of the classical tracing molecule technique. The technique depends on special synthesis techniques in which different substances are labelled with short lived positron-emitting radionuclides, for example 11C with a half-life of 20 minutes, which are produced in a cyclotron. A very large number of substances can be labelled, such as amino acids, carbohydrates, signal substances and drugs. After injection of the preparation into the patient, its distribution in time and space is examined with the help of the positron camera. Other short-lived positron-emitting radionuclides are 15O (half-life 2 minutes), 13N (l0 minutes) and 18F (110 minutes). The method gives regional quantitative functional and biochemical information. This information can be difficult or impossible to obtain in any other way. Blood volume, blood flow, metabolism, receptor and enzyme kinetics and pH can all be studied with PET. The technique improves the diagnosis and monitoring of treatment in a number of large groups of illness, for example tumours, infarctions, epilepsy (Fig. 3), skull injuries, psychiatric, movement and metabolic disorders. PET is capable of contributing enormously in the diagnosis and characterisation of central nervous system disorders. It is capable of shedding light on pathophysiology and is used in the development of new treatment methods. The method is still to a large extent used for research purposes but is increasingly being used clinically.

As far as the availability of isotope techniques goes, SPECT is used widely while the use of PET is limited to rather few hospitals worldwide. PET is a markedly more expansive technique than SPECT but is of great interest because of the potential which derives from the enormous variety of tracer molecules and quantification possibilities.

Kjell Bergstrom and Giuseppe Scotti