In the two decades since computed tomography (CT) was introduced by Sir Godfrey Hounsfield it has become established as a powerful diagnostic tool and one that is relevant to many branches of surgery. Used appropriately, CT is capable of making a major impact on management decisions.

The technology of CT is outside the scope of this chapter and the interested reader is referred to Hounsfield (1973) or to Pullan (1979). In essence, the scanner rotates an X-ray tube around the patient in an arc and the emergent radiation beam is measured by photoelectric detectors. A computer is used to display the measurements as an image representing a cross-sectional ‘slice’ of the patient, based on the density of tissues to X-rays and their ‘attenuation value’ (Fig. 1) 119. The image can be thought of as a cross-sectional radiograph but unlike radiography there is no superimposition of structures and the detector/computer system makes the technique very sensitive. The principal advantages of CT over conventional techniques stem from this fact.

From the patient's point of view CT examination is simple. In the majority of examinations all the patient has to do is lie on the couch while the machine makes the readings (Fig. 2) 120. Preparation is minimal; some departments prefer to starve their patients for a few hours beforehand if intravenous contrast media are to be given.

Patients undergoing examination of the abdomen and pelvis are required to drink a dilute contrast medium in order to opacify the bowel and distinguish this from local structures with similar attenuation values (Fig. 1) 119. This contrast medium is also administered rectally for examinations of the pelvic viscera. Examination for gynaecological indications requires a full bladder and the use of a vaginal tampon as an anatomical marker.

Barium retained in the colon after previous gastrointestinal radiology produces artefacts on CT and barium studies should be deferred if CT is required. Large metal prostheses such as hip transplants may cause sufficient interference to prevent successful imaging around them. Metal clips used for haemostasis or for marking the margins of tumours also produce significant artefacts.

Sedation is only rarely required for patients undergoing CT. Premedication may be useful for those in pain so that they are able to lie still. General anaesthesia is required if involuntary or uncontrollable movement may be a problem, particularly in children, although many infants can be examined satisfactorily after a feed or under sedation.

Examination is usually carried out with the patient supine, although specialized indications may require specific positions. Exposures typically last a few seconds, and suspended respiration is required when the chest or abdomen are being examined; a diagnostic examination may not be possible if the patient has difficulty holding their breath, as respiratory movement may cause artefact. Other areas can be studied during quiet respiration. Where exposures of the abdomen last longer than 5 s an antiperistaltic agent (e.g. hyoscine butylbromide, Buscopan, or glucagon) reduces movement artefact caused by peristalsis.

Most CT machines can produce a digital radiograph of the examination area (Fig. 3) 121 at the start of the examination. This allows the examination to be planned accurately and also provides a computerized record of the sections.

The examination produces sequential sections taken through the area of interest, the section thickness and inter-slice interval being determined by the size of the organ under investigation. Routine examinations of the chest, abdomen, or pelvis usually use sections 8- to 10- mm thick, whereas a specific examination of the adrenal glands may require 5- mm sections, and 2- mm sections may be needed to display the auditory ossicles. Axial sections are standard, although if the patient can be positioned appropriately in the gantry, sections can be obtained in coronal or other planes; this is required in only a minority of examinations.

The CT section is a cross-sectional radiograph in which the tissues are displayed on a grey scale according to their attenuation value. Dense structures such as cortical bone appear light, whereas low density areas like air appear dark, as is the case with conventional radiographs. The attenuation of tissues on CT is displayed on a wider scale than can be shown effectively on one image, however. The display console therefore allows the image to be manipulated so that areas at different points on the attenuation scale can be examined. This allows discrimination of, for example, the fine detail of the lungs, at the lower end of the attenuation scale (Fig. 4) 122,123.

The basic image data can also be manipulated in other ways. Measurements from a small area can be reprocessed to give a high resolution image; this is useful for demonstrating small structures such as the auditory ossicles or trabecular detail in bone (Fig. 5) 124. Information from contiguous thin sections may be reformatted in different planes, or in three-dimensional perspective views, to provide a more anatomical display. Such images are helpful in communicating the orientation of lesions but rarely add further diagnostic information.

Enhancement refers to the commonly used technique of scanning following the intravenous administration of iodine-containing contrast medium. This increases the attenuation of areas which have a circulation, and allows their differentiation from avascular structures such as abscesses, cysts and devascularized tissue. However it is commonly used to aid diagnosis by increasing the contrast between normal and abnormal tissues, hence the term ‘enhancement’. This technique is routine in the examination of the liver, kidney and brain, and in the investigation of abdominal sepsis or trauma (Fig. 6) 125.

As circulating blood increases in attenuation after enhancement, examination of blood vessels can be made by taking sections soon after injection, usually during the first or second circulation of the injection bolus. This may be used to distinguish tortuous vessels from other pathology and also to demonstrate vascular disease such as aortic aneurysm or dissection (Fig. 7) 126. Examination during direct injection of local veins or through an arterial catheter (CT arteriography) is also recommended in specific indications, although the standard technique suffices for most vascular studies.

Many CT machines allow rapidly repeated exposures to be made at one level following an intravascular bolus of contrast medium, with the resulting alteration plotted by the computer against time. This technique, ‘dynamic CT’, produces an accurate measurement of tissue perfusion but is little used in practice, although it is a useful research tool. Combined with incremental table movement, dynamic CT gives an overview of an organ in one phase of perfusion. This is sometimes used in examination of vessels and in a detailed search for small space-occupying lesions in the liver.

The cross-sectional display produced by CT has also been applied to other contrast procedures in radiology in order to obtain more information. Examples include examination after intrathecal or intra-articular injection (CT myelography and CT arthrography, respectively). In both cases cross-sectional definition is used to produce images of areas which are difficult to assess by the corresponding conventional technique.

The precise tissue map of CT images has been used successfully to direct treatment beams in radiotherapy and it is now common for radiotherapy to be planned on the basis of CT images, using specially constructed computer hardware.

The principal advantage of CT is that it provides a clear, accurate display of tissues without superimposition of structures. Disease processes may be detected at an earlier stage than is possible with other techniques, and lesions may be detected in areas which are difficult to assess with conventional imaging. The technique is not limited to specific organs: since all of the tissues in a body section are displayed it can be used to search for disease sites.

The clinical advantages of CT are well illustrated by its role in neurosurgery, one of its first areas of application. The technique offered for the first time the ability to image cerebral anatomy and pathology directly. Invasive and indirect tests such as arteriography and encephalography were largely replaced, and more accurate diagnosis became possible.

Although CT is effective in disease detection and localization, characterization of lesions is more difficult, since many have similar attenuation characteristics. For example, it may not be possible to distinguish fibrotic masses from benign or malignant neoplasms on the basis of their CT appearance alone. Biopsy is therefore usually required for definitive diagnosis.

The other main disadvantages of CT are the high capital cost of the equipment and the fact that it employs ionizing radiation. The absorbed radiation dose from CT varies according to examination technique, but it is generally similar to that encountered with other major radiological procedures such as angiography or barium enema. CT is therefore used with care around radiosensitive structures such as the eye, or in children and young people, and only for over-riding indications in pregnant women.

In the abdomen, pelvis, and musculoskeletal soft tissues, ultrasound offers an alternative to CT as a sectional imaging technique. The relative strengths of the two are too complicated to discuss in detail here, but in general, if a good quality image is obtained by ultrasound, the two techniques are usually comparable in application. However, the results of ultrasound, unlike CT, are limited by the presence of bowel gas and bone, and if these prevent good images being obtained, CT is more reliable. Ultrasound is also attenuated by fat, making the technique more suitable for slim patients, whereas a moderate amount of body fat improves image quality in CT. Ultrasound is of limited use in the chest.

The technique which most resembles CT is magnetic resonance imaging (MRI). Like CT, MRI is a cross-sectional technique, but unlike CT, examination in any plane is possible. MRI has the advantage of not employing radiation, and it discriminates between soft tissues to a degree unequalled by any other technique. However MRI is expensive and of limited availability. Movement artefact is a problem in studying the abdomen, because scan times are long and there is, as yet, no generally available contrast medium to label the bowel, as there is in CT. Little signal is obtained from lung and MRI does not compare with CT in this area.

It is the golden rule of investigational medicine that no patient is examined unless the results influence clinical management. This is particularly true of CT: examinations should always be tailored to the clinical problem. However, when used appropriately, CT has proved to be a powerful factor in clinical management decisions in a wide range of applications. All surgical specialities are major users of CT services. Table 1 114 lists the clinical indications for which CT is currently recommended; these are divided according to clinical subspeciality although there is some overlap.

In recent years the indications for CT have had to be reassessed in the light of growing experience in MRI and its availability. MRI has become an established investigational tool in the neurosciences and in orthopaedics, and to a lesser extent in gynaecology (Table 1) 114. For technical reasons MRI is unlikely ever to be applicable to the diagnosis of conditions affecting the lungs or cortical bone. Haemorrhage can produce confusing appearances on MRI and CT therefore remains the technique of choice in patients who have suffered trauma.

The main advantages of CT in clinical practice stem not only from its accuracy in detecting disease but also from the fact that a convincing normal examination virtually excludes the presence of lesions of any size. When an abdominal mass is suspected, for example, CT is the most accurate technique for demonstrating disease and indicating its organ of origin (Fig. 8) 127,128, but is also more reliable than ultrasound in excluding disease when there is none present. The technique can therefore be used to select patients who require surgical intervention. Similar observations apply to patients with suspected intra-abdominal abscesses (Fig. 6) 125, in whom lesions close to the bowel are difficult to detect by ultrasound; CT more reliably confirms or excludes a focal collection. CT is also a reliable technique for excluding significant damage to intra-abdominal organs in patients who have suffered abdominal trauma (Fig. 9) 129. In all three instances the exclusion of disease has a major effect on the clinical management of the patient.

The ability of CT to delineate masses accurately has produced a major advance in the management of malignant disease. In addition to the diagnostic role outlined above, the technique is recommended for assessing the local extent of the majority of solid tumours and distinguishes reliably between patients with resectable disease and those in whom an attempted resection is pointless (Fig. 8) 127,128. In addition, CT is used to detect malignant lymph node enlargement in the chest, abdomen, and pelvis (but cannot, unlike lymphography, demonstrate small tumour deposits in nodes of normal size) and has become the technique of choice for demonstrating metastases to the brain, lungs (Fig. 4) 122,123, liver, and adrenal glands. Disease staging protocols based on CT results have now been defined for the majority of malignant tumours, with the aim of excluding disease spread and therefore identifying those patients suitable for radical treatment.

The combined cross-sectional display of bone and soft tissue has made CT an important technique in orthopaedics, allowing the assessment of stability of spinal fractures, demonstrating the disposition of fracture fragments prior to surgical fixation (Fig. 5) 124, and planning corrective surgery in joint disease.

A wide range of percutaneous therapeutic procedures are now performed under CT control. The principal advantage of the technique is that it permits the operator to site an instrument with confidence and safety, even in relatively inaccessible areas of the body. The most common technique is CT-guided biopsy (Fig. 10) 130: aspiration for diagnostic cytology or cutting needle biopsy for histological diagnosis can be undertaken in virtually any area of the body. CT-guided drainage can also be used in the treatment of most deep-seated abscesses and other pathological fluid collections. Guided neurolysis, tumour lysis by alcohol injection, and laser therapy are also possible.

Although the capital and running costs of CT are high, the technique is undoubtedly cost effective. It can be used to achieve an early diagnosis in patients who would otherwise need to undergo a large number of alternative investigations, and it can be performed on an outpatient basis, reducing costs for inpatient investigation. Moreover, the diagnostic and therapeutic applications of CT frequently replace exploratory laparotomy, or other major surgical procedures. Maximization of these cost benefits is heavily dependent on good patient selection, and calls for close liaison between the surgeon and the radiologist.

In the future the clinical role of CT will need to be reassessed as MRI develops, particularly in examination of the abdomen, where further development of MRI is to be expected. For the present, CT is the mainstay of cross-sectional imaging.

Genant HK. Symposium on computed tomography. Orth Clin N Am 1985; 16: 357–89.

Golding SJ. Computed tomography and tumour staging. Curr Imaging 1990; 2: 2–8.

Golding SJ, Husband JE. CT-Guided interventional techniques. Interventional Radiol 1990; 5: 101–23.

Hounsfield GN. Computerised transverse axial scanning. Br J Radiol 1973; 46: 1016–22.

Moore AT, Dixon AK, Wheeler T. Cost-benefit evaluation of computed tomography. Health Trends 1987; 19:8–12.

Muller PR, Simeone JF. Intraabdominal abscess: diagnosis by sonography and computed tomography. Radiol Clin N Am 1983; 21: 425–43.

Pullan BR. The scientific basis of computerised tomography. In: Lodge T, Steiner R, eds. Recent Advances in Radiology No. 6. Edinburgh: Churchill Livingstone, 1979; 1–15.

Williams MP, Scott IHK, Dixon AK. Computer tomography in 101 patients with a palpable abdominal mass. Clin Radiol 1984; 35: 293–6.

Wing VW, Federle MP, Morris JA, Jeffrey RB, Bluth R. The clinical impact of CT for blunt abdominal trauma. Am J Roentgenol 1985; 145: 1191–4.

Wittenberg J, Fineberg HV, Ferrucci JT. et al. Clinical efficacy of computed body tomography. Am J Roentegenol 1980; 134: 1111–20.