• Substantial cost saving on film and processing chemicals

• Reduced physical space needed for storage

• Reduced staff costs—no need to file or retrieve films

• Better availability of images—no need to obtain physical films; several viewers can see the same images simultaneously in their place of work

• Substantially reduced numbers of missing examinations

• Easy transmission of images from one institution to another—by portable storage media (CD-ROM) or electronic links

Chest X-ray: Interpreting plain chest X-rays requires a methodical approach. Several learned documents have been written on the subject (see e.g. http://student.bmj.com/student/view-article.html?id=sbmj04018), and no attempt will be made to produce an incomplete version here.

Plain abdominal radiology: Most abdominal films are taken with the patient supine. Bowel is visible when it contains gas (Figs 5.3 and 5.4); normal small bowel is less than 3 cm wide and tends to occupy the centre of the abdomen. When dilated, it shows transverse folds (plicae circulares) which completely cross the lumen. The colon usually lies peripherally and has haustrations; these folds only partly traverse the lumen (Fig. 5.3). Normal colon is less than 6 cm wide and often contains faecal lumps with a mottled appearance. Further reading about this topic is available from: http://www.studentbmj.com/topics/clinical/imaging_techniques.php.

The kidneys may be outlined by a border of perinephric fat but outlines are often obscured by bowel gas and faeces. After nephrectomy, the ‘renal outline’ may persist and the kidney appears to be present. Small urinary tract stones are easily obscured and the liver may be visible but its size cannot be estimated.

When examining an abdominal X-ray, important features to look for are:

The limitations of plain abdominal radiography are summarised in Box 5.1.

Bone X-rays: Conventional radiographs of bones continue to play an important role in diagnosis of bone disease, particularly fractures, but also infection, neoplasia and degenerative conditions. Other investigations including bone scintigraphy, CT, MRI and ultrasound are often used, particularly if conventional X-rays are normal. If in doubt it is best to discuss with a radiologist.

Contrast materials: Barium sulphate is the best agent for directly outlining the GI tract. It is insoluble in water and is not absorbed. An aqueous suspension is non-irritant and very radiodense. Barium investigations of the upper and lower gastrointestinal tracts have been largely superseded by endoscopy or cross-sectional imaging but are still occasionally employed.

Water-soluble iodinated benzoic acid derivatives can be injected into arteries or veins to opacify them (arteriography or venography). Injectable contrast media have improved over the years to make them safer but they remain potentially nephrotoxic in patients with renal impairment.

Direct venography has largely been replaced by colour duplex ultrasound, and CT or MR angiography is used more than direct studies for diagnosis.

Examples of contrast radiology:

Complications of barium contrast studies: The limitations of barium contrast studies are summarised in Box 5.2. There is a risk that contrast material may be aspirated into the bronchial tree, causing aspiration pneumonitis, so care must be taken when giving oral barium to patients with swallowing difficulties. CT scanning is safer and more likely to provide useful information such as the level and cause of obstruction, and associated pathology such as lung or liver metastases.

Magnetic resonance cholangio-pancreatography (MRCP) (Fig. 5.6 e and f): MRCP now produces images that rival the quality of ERCP. MRI differentiates tissues and organs by their varying water content. Bile and pancreatic juice are mostly water, hence MRCP gives clear images of bile in the gall bladder and ducts and outlines the pancreatic duct. It reveals filling defects caused by stones or tumours. MRCP can identify bile leaks, gallstones in the bile ducts, and duct obstruction from any cause. There are no known hazards. MRCP is increasingly used prior to ERCP for pancreatico-biliary investigation to reduce the number of patients requiring the more invasive investigation.

Indications for MRCP include:

Percutaneous transhepatic cholangiography (PTC): Percutaneous cholangiography is rarely used for diagnosis but has occasional therapeutic indications. A long fine (22 G) ‘Chiba’ needle is passed percutaneously, directly into dilated intrahepatic ducts, and contrast injected to display the duct system (see Fig. 5.6c). This shows the configuration of extrahepatic duct obstruction from the proximal direction. Occasionally, PTC is employed to drain obstructed ducts after failed endoscopic placement of a drain or stent. In jaundiced patients, disordered blood clotting is likely so a clotting screen and platelet count should be performed before PTC, and any clotting disorder corrected.

Endoscopic retrograde cholangio-pancreatography (ERCP): This is described below (see Diagnostic and therapeutic duodenoscopy); its use in obstructive jaundice is described in detail in Chapter 18. The basic technique is illustrated in Figure 5.6 a and b.

Operative cholangiography and choledochoscopy: It is standard practice to perform operative cholangiography during open cholecystectomy. For laparoscopic cholecystectomy, some surgeons routinely perform operative cholangiography whilst others prefer no imaging at all for selected cases or else preoperative assessment using MRCP for cases deemed likely to have duct stones.

Operative cholangiography allows the (highly variable) biliary anatomy to be displayed, it demonstrates stones in the bile ducts and it shows whether contrast flows freely into the duodenum. A fine plastic cannula is introduced into a small cystic duct incision and passed into the common bile duct. Water-soluble contrast material is injected to outline the duct system and fluoroscopic images or X-ray films are taken. If duct stones are demonstrated, they are often retrieved surgically at the same operation. At open cholecystectomy, this is via a longitudinal incision in the common bile duct (exploration of the common bile duct). At laparoscopic surgery, the same technique is used with a small transverse or longitudinal choledochotomy, depending on the duct size and the stone size to be retrieved. A flexible 5 mm or 3 mm endoscope called a choledochoscope is passed into the bile duct and stones may be retrieved using a range of techniques including snares, baskets, balloons, or they can even be shattered with lithotripsy probes. The choledochoscope enables the bile and intrahepatic ducts to be inspected afterwards to see that all stones have been removed. A further cholangiogram is often done afterwards to ensure the duct is clear. Stone removal may be deferred and performed later at ERCP although it carries risk of complications, including biliary leakage and acute pancreatitis.

T-tube cholangiography: Following exploration of bile ducts for stones, a T-tube is often left in situ to drain the duct. The transverse limb of the T lies in the duct and the long limb drains out through an opening in the duct to the skin where it is connected to a drainage bag. About 1 week after operation, contrast can be injected along the T-tube to outline the biliary tree and show abnormalities such as residual stones, bile leakage and duct stenoses as well as confirming free drainage into the duodenum. If residual stones are still present they can be retrieved either at ERCP or sometimes by the radiologists via the T-tube itself.

Percutaneous transluminal angioplasty (PTA) or balloon angioplasty: Angioplasty under local anaesthesia is a less invasive alternative to surgery for treating many peripheral and coronary arterial stenoses. In general, short stenoses in large vessels are most suitable. The method is particularly useful for lower limb atherosclerosis (especially iliac and superficial femoral arteries) and for coronary artery disease and, to a lesser extent, renal artery and mesenteric artery stenoses. Carotid artery disease is also suitable for angioplasty in some cases; filters are usually placed above the stenotic segment before dilatation to reduce the risk of embolism to the brain.

Major complications of angioplasty are rare in experienced hands but there is a small risk of precipitating acute ischaemia due to distal embolisation or dissection. Thus, surgical salvage should be readily available should complications develop. Unfortunately, 25–40% of angioplastied lesions undergo restenosis or occlusion within 1 year, but the process can usually be repeated. Where stenoses fail to remain open at angioplasty, expandable stents can be placed within the treated lesion. Peripheral arterial stents are most often used to treat iliac occlusive disease and improve long-term patency. These stents are either mounted on an angioplasty balloon or are self-expanding on withdrawal of the introducer sheath. Stents are being increasingly used in superficial femoral artery stenoses and occlusions. Overall, angioplasty causes minimal interventional and anaesthetic stress to the patient and is often performed on a day-case basis.

Techniques of percutaneous angioplasty: Angioplasty is usually performed under local anaesthesia (Fig. 5.7). A needle is first inserted into an accessible artery and a short flexible guide-wire passed through it into the artery and the needle removed. A working sheath with a valved side-arm is passed over the guide-wire and advanced into the artery. A catheter is inserted and a long guide-wire is then substituted for the first and manipulated up to and through the stenosis using contrast injections and fluoroscopic control. An angioplasty catheter with a plastic inflatable balloon at its end is then passed over the guide-wire and manipulated across the stenosis. Angioplasty balloons are now no wider than the catheter before inflation, and inflate to a fixed diameter at a given pressure. It is possible to measure the arterial pressure above and below the stenosis to determine the pressure gradient before and after angioplasty, although this is not performed routinely. The balloon is then inflated to a pressure of typically 6 and 15 atmospheres, depending on the balloon type, to dilate the stenosis, then further contrast is injected to check the result. Angioplasty techniques and equipment have progressively improved and many patients now return to near-normal life after minimal intervention.

Many patients with ischaemic legs or coronary heart disease who might not have been suitable for reconstructive surgery can now have angioplasty because of its minimally invasive nature and low complication rates.

Local arterial thrombolytic therapy: An artery freshly occluded by thrombosis and causing ischaemia can be recanalised by local intra-arterial infusion of thrombolytic agents. High local concentrations with limited systemic spill-over were intended to avoid the serious bleeding and allergic complications of systemic thrombolysis. However, experience has shown that the risk of major haemorrhage still exists, some episodes of which have been fatal. Examples include intracerebral haemorrhagic strokes and bleeding from recent surgical wounds or from the gastrointestinal tract. For this reason, the treatment has limited indications, i.e. recently occluded native arteries or bypass grafts. The thrombolytic agent usually used is recombinant tissue plasminogen activator (R-tPa). R-tPa acts more quickly and does not have the frequent allergic effects of the older agents, e.g. streptokinase.

For recent acute embolic ischaemia, surgical embolectomy remains the best treatment.

Therapeutic embolisation: Highly vascular lesions that would be difficult or impossible to treat by surgery alone, such as some arteriovenous malformations, can have their arterial supply reduced or obliterated by embolisation. The main supplying artery is identified by selective arteriography and a catheter manoeuvred into it, close to the lesion. The materials chosen for embolisation depend on the nature of the lesion but include gelatin foam, polyvinyl alcohol particles, minute steel coils, plugs, cyanoacrylate glue or liquid polymers. The process can be repeated for all the feeding vessels.

Embolisation is particularly useful in the treatment of gastrointestinal haemorrhage, some pseudoaneurysms, uterine fibroids and internal iliac arteries prior to endovascular aneurysm repair (EVAR). The technique is also used to treat hepatic tumours (primary and secondary) and some bone metastases prior to surgery.

Minimal access graft placement: EVAR is now widely used to treat both abdominal and thoracic aortic aneurysms. There are several different devices available, composed of metal stents and graft material, and technical improvements are continuing. The device is correctly positioned in the aorta, and then deployment consists of unsheathing the device under fluoroscopic guidance. Graft limbs are used to extend the device into the iliac arteries. There are now stent grafts capable of treating more complex aneurysms with side branches/grafts into the renal, mesenteric, internal iliac or arch arteries. One disadvantage of the technique includes endoleakage, i.e. continued slow bleeding into the aneurysm sac because of an inadequate seal or, more commonly, from lumbar arteries or the inferior mesenteric artery. Other complications include graft migration, limb occlusion or limb dislocation. Although reintervention rates are higher than with open aneurysm repair, these techniques have allowed many patients to be treated who would never be fit for open surgery. The late complication rate is about 10% per year, considerably greater than for open aneurysm grafting, but rates are falling with improving techniques.

Venography: Colour duplex Doppler ultrasound scanning has replaced contrast venography for diagnosing deep vein thrombosis (DVT) as well as for demonstrating reflux from deep to superficial vessels in varicose veins. A skilled operator can demonstrate the patency or otherwise of all the lower limb veins and the presence of fresh or old thrombus. Valve competence can be shown in deep and superficial veins and perforating veins. Vein wall irregularity caused by previous DVT can also be displayed. Contrast venography can sometimes be helpful in defining the anatomy of complex superficial varicose veins and very occasionally to diagnose or exclude calf vein DVT where duplex is inconclusive.

Placement of vena caval filters: After venous thromboembolism, a few patients experience recurrent pulmonary embolism despite adequate anticoagulation. In others, anticoagulation therapy is contraindicated and other methods of preventing pulmonary embolism must be found, e.g. in pregnancy, after a haemorrhagic stroke, in patients with a high risk of falling or those with certain bleeding disorders. In these groups, the risk of pulmonary embolism can be markedly reduced by placing a filter in the inferior vena cava, above or below the renal veins. This still allows venous blood to return to the heart but traps substantial embolic material in the flowing blood. Since the mid-1960s a range of filtration devices have been developed that can be inserted relatively simply using a catheter via the femoral or jugular vein. Typical examples are the Celect or Optease filter. Some filters can be retrieved by percutaneous techniques if necessary within a limited time period. Complications are uncommon and include caval occlusion, migration and strut fracture.

Minimally invasive treatment of varicose veins: Several methods of ablating the long or short saphenous vein have appeared recently including foam sclerotherapy and laser or radiofrequency ablation. These are described in Chapter 43.

General principles of urography: Urography is a radiological technique for examining the kidneys and urinary collecting systems using intravenous contrast concentrated and excreted by the kidneys. A plain abdominal control film is taken beforehand so calcified opacities can be compared with post-contrast images. This helps identify if an opacity lies within the renal tract and hence whether it is likely to be a stone (see Fig. 5.8a).

In intravenous urography (IVU), films are taken at intervals after injection. The renal parenchyma normally opacifies almost immediately and contrast then flows into the renal pelvis, ureters and bladder.

Special precautions with intravenous urography: Intravenous contrast is potentially nephrotoxic in patients with impaired renal function. Risk factors should be specified to help the radiologist plan the safest investigation. Alternatives such as ultrasound, unenhanced CT or occasionally MR may be considered. Important risk disorders are:

CT urography: Unenhanced CT of the renal tract is increasingly used instead of IVU to diagnose renal or ureteric colic. It is more sensitive for detecting stones and is quicker to perform but usually gives a higher radiation dose. CT with intravenous contrast is frequently used to investigate persistent haematuria when other tests are normal; in some centres, it is used earlier in the diagnostic pathway.

Percutaneous therapeutic techniques: The renal pelvis can be punctured percutaneously with a needle guided by ultrasound or CT scanning. The tract is then dilated to allow tubes of various types and sizes to be inserted. Gaining access to the kidney in this way is known as percutaneous nephrostomy. It can be employed to remove stones from the renal pelvis, to urgently but temporarily relieve acute distal urinary obstruction and drain the kidney over a few days ahead of a definitive procedure.

Special transducers: Probes have been developed for use via different body orifices, and for body cavities via endoscopic instruments, percutaneous cannulae and laparoscopes. These devices are placed closer to the structures being examined than surface probes. This allows higher-frequency sound to be used which has lower penetration but greater spectral resolution and gives a more detailed display. These transducers are often combined with biopsy devices to enable tissue sampling.

Rectal probes are used for examining the rectal wall and prostate gland in detail and vaginal probes for investigating the pelvic organs. Endoscopic probes (e.g. transoesophageal) can examine and monitor the heart, upper GI organs and adjacent tissues. Laparoscopic probes can be applied directly to viscera to seek the extent of tumour spread or metastases, e.g. pancreas, liver. They provide a more reliable diagnosis of liver metastases than percutaneous ultrasound or cross-sectional imaging.

Doppler-shifted ultrasound: Ultrasound can detect and study blood flow by applying the Doppler principle. Simple hand-held equipment is cheap and portable, and is invaluable in the vascular clinic (see Fig. 5.9). Using a probe coupled to the skin with conduction gel, a beam of ultrasound is directed at an artery or vein. Ultrasound reflects from moving red cells and causes a shift in sound frequency related to blood velocity. Reflected ultrasound is used to generate an audible signal (for detecting blood flow) or is processed to reveal information about the nature of flow. The audio pitch is related to blood velocity and provides some qualitative assessment of whether flow is normal or abnormal.

General principles of CT scanning: Computerised tomography involves X-raying a series of thin transverse ‘slices’ of the patient’s head, body or limbs. A precise fan-shaped beam of X-rays is repeatedly pulsed from successive angles around the circumference of each slice and the transmitted radiation is electronically recorded on the opposite side (see Fig. 5.10).

Since CT was first introduced in the 1970s, the pace of development has been rapid. Modern machines capture images in a continuous spiral around the patient (spiral CT), usually as multislice CT which enables several slices to be captured simultaneously with each revolution of the X-ray tube. All modern machines now generate at least 64 slices for each turn. The result is rapid image capture of thinner slices at higher resolution. Spiral multislice CT machines can now produce images of both chest and abdomen in less than 1 minute. The data quality also allows images to be accurately reconstructed in three dimensions or in any chosen plane, e.g. sagittal, coronal or oblique, to improve the detection of abnormalities.

Note that the best CT images are obtained in well-nourished patients because some fat lies between the organs, enabling them to be differentiated more precisely than in very thin patients.

Further information can often be gained by performing CT after or during contrast, air or CO₂ enhancement. For example, oral or rectal contrast clearly outlines bowel, whilst intravenous contrast can show blood vessels, kidneys, damage to the blood–brain barrier or areas of absent blood flow in pancreatic necrosis.

CT and MRI are complementary techniques, each having advantages and disadvantages. The decision to use one or other depends on patient factors and availability of equipment as well as the suspected disease.

Applications of CT scanning: Pathological anatomy can be studied in great detail and a huge array of information can be obtained non-invasively to assist surgical diagnosis. Often the information is more accurate than could be obtained by exploratory operation. This is outstandingly so in brain injury after trauma where the management of serious head injuries has been transformed by head scanning. The technique enables timely and appropriate surgical intervention and avoids unnecessary exploratory operations.

As CT becomes more available, cheaper and technically better, it is often being used earlier in the diagnostic process, particularly in emergency cases, and for an increasing range of clinical conditions, though it is important to use clinical method and clinical common sense before ordering such tests.

Some indications for CT scanning are:

Principles of magnetic resonance imaging: Magnetic resonance imaging (MRI) was introduced into clinical practice in the early 1980s. MRI involves applying a powerful magnetic field to the body which causes the protons of hydrogen nuclei to become aligned. The protons are then excited by pulses of radio waves transmitted at a frequency that causes them to resonate and emit radio signals; these are recorded electronically. Sophisticated computation then produces images which can be viewed in any plane, transversely, longitudinally or at any obliquity (see Fig. 5.11).

Lipids have particularly high hydrogen content and are clearly seen on MRI. For this reason, the initial applications of MRI were in examining the brain and spinal cord. The technique is increasingly employed for investigating joints such as the knee, shoulder, hip and ankle, and in some cases replaces the need for arthroscopy. Examination times have fallen and good-quality images of chest, abdomen and pelvis can now be obtained. However, MRI is unsuitable for imaging gas-filled organs and dense bone.

Disadvantages of MRI: Some patients are unable to tolerate MRI because of claustrophobia but the main technical drawback is the effect of magnetism on metal. Thus metallic foreign bodies, especially in the eyes, may move in the magnetic field and cause damage. MRI cannot be used with cardiac pacemakers and some metallic implants, and patients being artificially ventilated require machines made of non-ferrous materials. Scanning times are still longer than for CT (though falling) and so MRI is less suitable for young children, the elderly or confused, patients in pain, ventilated patients and emergency patients with active bleeding. The diagnostic applications of MRI continue to expand in parallel with the technical improvements.

Applications of magnetic resonance imaging:

Fine-needle aspiration cytology (FNA) and core biopsy: A fine needle (22 gauge) can be safely passed through most organs or small bowel to aspirate fragments of tissue from a suspicious lesion. Larger-diameter needles can be used for direct core biopsy of masses. The depth and direction of the needle can be accurately guided by ultrasound or CT to ensure a representative sample is taken. For example, pancreatic masses can be reached by transfixing bowel lying in front of the pancreas; this causes remarkably few side-effects. Where practicable, many surgeons and pathologists prefer the larger specimens obtainable with core biopsy techniques using special needles such as the Trucut, available in various configurations and dimensions.

Minimally invasive applications in breast disease include ultrasound- or mammographically-guided fine-needle aspiration (FNA) or core biopsy of asymptomatic abnormalities detected on mammography, including those found on screening. Stereotactic apparatus can be employed to make this process more accurate. Mammographic guidance is also sometimes used to place a hooked wire close to an impalpable abnormality to locate it before surgical excision (mammographic localisation).

Guided core biopsy or FNA techniques are also important in the diagnosis of thyroid lumps, for sampling liver nodules and for taking renal biopsies in diffuse renal disease.

Lung scanning: Lung scanning was widely used in the diagnosis of pulmonary embolism but is used much less since multislice CT became widely available.

Bone scanning: Phosphate-based agents (phosphates or bisphosphonates) labelled with technetium are usually employed. The tracer is taken up in areas of increased bone deposition and resorption, indicating sites of bone growth and repair (see Fig. 5.13). These include growth plates, some primary tumours, secondary tumours, foci of bone infection, healing fracture sites, active arthritis and Paget’s disease. Bone scanning is highly sensitive but interpretation of the scans requires caution because it lacks specificity. They are usually interpreted alongside plain radiographs to improve specificity.

The tracer agent is injected intravenously and becomes distributed throughout all body fluids. The highest concentration collects at sites of osteogenesis about 6 hours later and the patient is then scanned. The tracer is also taken up in areas of dystrophic calcification and may sometimes reveal an unsuspected carcinoma of breast, an old myocardial infarction scar or a uterine fibroid.

The main indications for bone scanning are:

Renal scans: Renal scanning is an important method of investigating the urinary tract. It can obtain information not available from any other source, is quick and simple to perform and allows the function of each kidney to be assessed individually.

There are three main varieties of scan, using different isotopes. DTPA (diethylene tetramine penta-acetic acid) is excreted in the urine like urographic contrast, while DMSA (dimercapto-succinic acid) and MAG3 remain in cortical tissue. (Aide-mémoire: DT ‘Pee’ A, excreted in urine; D ‘Meat’ SA, retained in cortical tissue.) Examples are shown in Figure 5.14.

DTPA scanning is used to follow up children with reflux nephropathy. The isotope is instilled into the bladder; the child then voids urine while being scanned and any vesico-ureteric reflux is demonstrated. DTPA is also used to diagnose ureteric obstruction and to distinguish obstructed from merely capacious non-obstructed renal tracts.

When unilateral renal parenchymal disease is being investigated, both DMSA and DTPA can give an estimate of excretory activity. The two agents can be used to estimate differential renal function when investigating renal artery stenosis or the function of a transplanted kidney. DMSA is used specifically to image the renal parenchyma to demonstrate renal scars or tumours.

Scanning for gastrointestinal bleeding: Scanning using the patient’s own isotopically labelled red cells may be employed to locate a source of continuing or intermittent gastrointestinal bleeding. This is useful where the rate of bleeding is relatively slow or in a patient with recurrent haemorrhage, particularly where a source cannot be identified by endoscopy or arteriography.

The patient’s blood is labelled with radioactive technetium and reinjected, then the abdomen is scanned at intervals over the next 24 hours or so for ‘hot spots’ indicating accumulating gastrointestinal haemorrhage. If the rate of bleeding is more than about 0.1 ml per minute, the scan usually reveals activity concentrated in one part of the bowel. This indicates the general area of haemorrhage and enables the surgical search to be focused, for example on the distal stomach or right side of the colon. Radionuclide scanning can detect blood accumulating over a period, whereas selective angiography requires a higher rate of bleeding at the moment of injection; however, it can reveal the site more precisely.

In children, rectal bleeding may be due to bleeding from a Meckel’s diverticulum caused by ulceration of ectopic gastric mucosa. A radionuclide compound of technetium concentrated in gastric mucosa may reveal the source.

Leucocyte scanning for inflammation and infection: When an abscess or other infected focus is suspected but cannot be localised, the patient’s white blood cells can be labelled with indium-111 or technetium-99m, then reinjected and scanned. Typical indications are patients with a high swinging pyrexia after operation, or with sepsis of unknown origin. The process is expensive but has a high degree of specificity and sensitivity; however, there is a small proportion of false negatives where an occult abscess is not revealed by the scan.

Leucocyte scanning can determine the extent of bowel involvement in inflammatory bowel disease, both ulcerative colitis and Crohn’s disease. Tc-hexamethylpropyleneamine oxime (HMPAO)-labelled leucocytes migrate towards areas of inflamed bowel which are then revealed on imaging.

Thyroid scans: Thyroid scanning is described in Chapter 49. Its use is declining in favour of fine-needle aspiration and cytology except in specific disorders of thyroid function.

Cardiovascular imaging: A multiple gated acquisition (MUGA) scan can provide information about ventricular function. This can be useful following myocardial infarction or for patients receiving doxorubicin (Adriamycin) chemotherapy which can damage heart muscle. Radionuclide lymphangiography in chronic lymphoedema can demonstrate the patency and capacity of lower limb lymphatics.

Liver and spleen scans:

Fibreoptic illumination: Rigid endoscopes and early flexible instruments were illuminated by tiny incandescent bulbs prone to failure. These were superseded by fibreoptic light guides in both rigid and flexible endoscopes. Fibreoptic light guides channel light from a powerful, remote fan-cooled light source to the distal end of an endoscope. They are made up of thousands of glass fibres, each with total internal reflection, so very little light is lost and no heat is transmitted. A powerful, cool light beam emerges from the distal end of even the longest endoscope.

Image transmission: The next development crucial to the design of early flexible endoscopes was the invention of coherent viewing bundles where the orientation of fibres at the distal end exactly matched the proximal viewing end. Each fibre transmitted a tiny part of the distal scene to the viewing end. The distal end of most endoscopes allowed a viewing angle of over 100°, and lenses gave a remarkable depth of focus. Accurate diagnosis could often be made on inspection alone.

A later development was the charge-coupled device (CCD) video camera, a light- and colour-sensitive microchip positioned at the distal end. The image is transmitted to a colour monitor and the system has now replaced the older direct viewing method.

Structure of flexible endoscopes: Most flexible endoscopes include a mechanism to steer the distal end in four directions (except for specialised ultra-slender scopes), a distal imaging chip for video endoscopy, one or two fibreoptic light guides, a suction channel and a channel for inflating the hollow viscus under inspection with air, doubling as a lens washing channel (see Fig. 5.15). The suction channel is also used to pass slender flexible operating tools such as tiny forceps for biopsies, grasping forceps for retrieving foreign bodies, laser guides for therapy (haemostasis or tumour destruction), snares for excision of polyps, diathermy wires, scissors for cutting sutures and needles for injecting haemostatic agents.

Diagnostic upper gastrointestinal endoscopy: Oesophago-gastro-duodenoscopy, also known as OGD or gastroscopy, involves inspecting the upper GI mucosa using a steerable, flexible endoscope. It is usually carried out under intravenous sedation and local anaesthetic spray on a day-case basis. Among other applications, gastroscopy enables the whole area prone to peptic ulcer disease and cancer to be directly and comprehensively examined.

Flexible endoscopy has the following advantages over GI contrast radiology:

Therapeutic upper gastrointestinal endoscopy:

Diagnostic and therapeutic duodenoscopy: A side-viewing duodenoscope can be used to inspect the duodenal papilla and guide insertion of a cannula or therapeutic tools. Cannulation allows injection of contrast material into the common bile duct and separately into the pancreatic duct. The technique is known as ERCP (endoscopic retrograde cholangio-pancreatography) (see Fig. 5.17 and Ch. 11) and is an important part of gastroenterological investigation.

ERCP can be both diagnostic and therapeutic. Indications for diagnostic ERCP may be decreasing as newer and safer techniques such as MRCP (magnetic resonance cholangio-pancreatography) become available. Therapeutic ERCP allows many bile duct disorders that would previously have required difficult, time-consuming and dangerous operations to be managed by minimal access techniques, with short hospital stays. For example, bile duct stones can often be removed endoscopically by slitting the sphincter at the lower end (sphincterotomy) and retrieving them with a balloon catheter or a Dormia basket. Other therapeutic measures include inserting bile duct stents for palliating malignant biliary obstruction (cancer of pancreatic head, bile duct or duodenum) and for managing postoperative bile leaks.

Enteroscopy: Barium follow-through, CT and MRI have low rates of positive diagnosis in small bowel disorders. Direct small bowel visualisation used to be achieved by ‘push’ enteroscopy (with a 2 m endoscope that could examine up to a metre beyond the duodeno-jejunal flexure) or by operative enteroscopy via a laparotomy. Later, a technique of double-balloon enteroscopy appeared, with an endoscope passed via the mouth and then coaxed along the small bowel using attached balloons as counter-traction. None of these methods was convenient or reliable.

A newer development for small bowel investigation is capsule endoscopy, introduced in 1999. This enables the entire 3–5 metres of small bowel to be visualised with relative ease. The patient swallows a capsule which is propelled through the gastrointestinal tract by peristalsis. An imaging device continually transmits images to sensors on the abdominal wall and the capsule-camera then passes in the stool.

One device, the PillCam SB capsule endoscope, is only 26 × 11 mm, weighs under 4 g. It contains a battery, light-emitting diodes, an imaging chip camera that captures images at 2 frames a second and a radio transmitter that passes images to a sensor array for up to 8 hours. The camera has an image field of 140°. If obstruction is suspected, a different model with a body made of lactose can be used. This disintegrates in less than 48 hours if arrested.

Capsule endoscopy has a positive diagnosis rate of around 65% compared with around 20% for other methods. In obscure gastrointestinal bleeding, there is a positive diagnostic yield of 45–75% in patients who have already had negative upper and lower GI endoscopy. Typical findings include angiodysplasia, tumours, varices and ulcers. Other indications include suspected small bowel Crohn’s disease, particularly in children, assessment of coeliac disease, screening in familial polyposis syndromes and diagnosis of Barrett’s oesophagus (by attaching a string). Biopsies cannot yet be taken.

Large bowel endoscopy (colonoscopy): Flexible endoscopes of different lengths are available for large bowel examination (see Fig. 5.18). The shortest, the fibreoptic sigmoidoscope, is about 60 cm long. It is simple to use and allows examination of the rectum, sigmoid colon and descending colon with minimal bowel preparation. Longer colonoscopes enable the entire large bowel to be inspected, and vary in stiffness to assist intubation to the caecum. Other techniques also help reach the caecum, including insufflation with carbon dioxide rather than air, and releasing seed oil from the tip to lubricate the instrument.

Colonoscopy allows inspection of pathological lesions, biopsy of suspicious lesions and resection of lesions such as polyps. Colonoscopy is also employed for surveillance and follow-up of patients treated for colorectal cancer or polyps. New tumours or polyps (metachronous lesions) are looked for and the original site of surgery can be examined. Similar examinations are also used for surveillance of patients with longstanding ulcerative colitis; multiple biopsies are taken to examine for dysplasia, and the entire large bowel is inspected for adenomas or carcinomas. Acutely bleeding angiodysplastic lesions in the large bowel can be treated with diathermy.

Colonoscopy is the most reliable method of screening asymptomatic people for colorectal carcinoma. However, its application is limited by the lack of trained endoscopists, by cost and by lack of patient compliance.

Urological endoscopy: Endoscopic urology is a progressively larger part of urological surgery with flexible instruments used for diagnostic cystoscopy and ureteroscopy.

Cystourethroscopy (cystoscopy) using a rigid instrument is the main diagnostic and therapeutic tool for disease of the urethra, prostate and bladder. Transurethral resection of the prostate has virtually eliminated the need for open retropubic prostatectomy and most early bladder tumours can be treated endoscopically. Laser enucleation of prostatic adenomas is an exciting development. An instrument similar to the cystoscope (but longer), the ureteroscope, can now be used to retrieve stones from the lower half of the ureter.

Endoscopic methods of percutaneous stone removal from the renal pelvis have become widely available (percutaneous nephrolithotomy). These involve creating a channel from the skin into the renal pelvis and dilating it until an endoscope can be passed. When the stone is seen, various instruments can be used to fragment it and achieve its removal.

Diagnostic and therapeutic laparoscopy: This is covered in Chapter 10.