Radiological examination may provide invaluable diagnostic information in cases of amoebic liver abscess. Since the descriptions of Munk in 1944,¹ various authors have emphasized radiological abnormalities in amoebic infection of the liver. In 1958, Lamont and Pooler² included these abnormalities as one of the diagnostic criteria in the syndrome of hepatic amoebiasis.
The commonly used radiological investigations include fluoroscopy, teleradiography, the use of contrast media, tomography and angiography. The procedure most likely to yield useful information can be judged only from clinical experience.
As in any radiological survey, the liver, the diaphragm, the pleura and the lungs must be studied separately. It is also important to keep somatic and gonadal radiation to a minimum, with its judicious use during pregnancy.

Fluoroscopy
The fluoroscopic examination of the upper abdomen and chest should, if possible, be carried out with an image intensifier.
The advantage of fluoroscopy over radiography is that the movements of the diaphragm can be visualised .³
The most frequently encountered fluoroscopic sign is elevation of the right dome of the diaphragm. In fact, as early as 1945, Isaac⁴ stated, that elevation of the hemidiaphragm with diminished mobility or complete immobilisation is pathognomonic of a frank liver abscess. This was again emphasized by Chaves.⁵ Fluoroscopy is also useful in demonstrating a pericardial effusion due to rupture of an amoebic abscess of the left lobe of the liver. In these cases there is an enlargement of the cardiac shadow and pulsations are usually absent.

X-ray chest
The roentgenographic signs of amoebic liver abscess are numerous. They vary considerably according to the stage of the disease, the location of the abscess and the extent and degree of hepatic and thoracic involvement.⁶ The frequency of abnormal findings in chest X-rays varies markedly in different series from 64.8 to 83.5%.⁷
An important sign of amoebic liver abscess is the elevation of the right dome of the diaphragm (Figs. 23 a,b). This is diagnosed when the difference between the two hemidiaphragms is 2.5 cm. or more. Ramachandran et al⁸ feel that the figure of 3.5 cm. will prevent over-reading of the X-rays, as elevation of the right hemidiaphragm may also be seen in a small proportion of healthy young adults, a proportion of patients with cardiac disease, cirrhosis of liver. gastroenteritis,⁹ deformities of the thoracic cage and in right phrenic nerve lesions.
In 1963, Rowland¹⁰ described persistence of this radiological abnormality in a small proportion of patients of 'hepatic amoebiasis' and attributed it to the fact that evacuation of pus was not carried out. However, in 1975, Ramachandran et al⁸ reported persistently elevated hemidiaphragm even after the pus was evacuated. They suggested that this sign may be due to presence of sizeable silent abscesses in some; and the occurrence of a non-specific reactive hepatitis or chronic intestinal amoebiasis in others. On the other hand, Subramaniam and Madangopalan¹¹ thought that persistence of elevation did not necessarily mean persistence of the abscess. Unfortunately all these studies were done without the help of liver scans. Therefore, most of these conclusions cannot be accepted. Although we have done no prospective study, according to our observations, in majority of the cases, contours of diaphragm go hand in hand with the size and healing of the abscess. However, on a few occasions we have observed persistence of radiological signs for a few years (Fig. 24) of an abscess has healed as proved by a normal scan. These changes are possibly due to adhesions resulting in permanent elevation and immobility of the diaphragm.
The postero-anterior and the right lateral views of chest should be asked for in patient suspected of having an amoebic liver abscess. A number of such cases would be found to have elevation of the right hemidiaphragm in right lateral view, without a significant elevation in the postero-anterior view.¹¹
In my opinion, contour of the right dome is more important than elevation. A "hazy and indistinct' contour is more significant in the diagnosis (Fig. 25) Others have also often seen a flat elevated right dome.⁶ The flattened appearance is possibly due to obliteration of costo- and cardiophrenic angles (Figs. 26 a,b). A localised bulging or tenting of the right hemidiaphragm into the right lower lung field denotes a superior surface abscess⁶ (Figs. 27 a,b). An abscess situated supero-anteriorly causes bulge in the anteromedial part of the right hemidiaphragm, with obliteration of the cardiophrenic angle in the posteroanterior view and the anterior costophrenic gutter in the lateral view. Localised elevations are more significant because rarely, a generalised elevation can be produced by any hepatomegaly.
Plate atelectases shadows are a common finding at the right base and are quite diagnostic in presence of an elevated right dome (Figs. 28 a,b,c). These are transverse linear opacities and are possibly due to linear collapse of lung tissue because of compression.
An effusion of the median fissure is not uncommon in amoebic liver abscess.¹² It produces a triangular opacity with the base directed towards the diaphragm and the apex to the right hilum. Such triangular shadows in lateral views have also been seen in cases of hepatobronchial fistulas¹³ (Fig. 29). Pleural effusions and adhesions may obliterate costo- and cardiophrenic angles.
In 1946, Galloway¹⁴ stated that a subphrenic abscess complicating an amoebic liver abscess caused obliteration of cardiophrenic angle medially and anteriorly. In subphrenic abscess per se, the obliteration is seen more commonly in the posterior costophrenic space in the lateral view. I have not found these signs useful enough to be specific for a differential diagnosis.
A right lobe lateral surface abscess may cause widening of the right lower intercostal spaces.
Inflammatory changes are commonly seen at the base of the right lung. Pneumonitis associated with elevated right dome of the diaphragm is probably secondary to amoebic liver abscess, if other causes of elevation can be ruled out.
Pleuropulmonary complications may occur. In such an event, the radiological signs may include diffuse or patchy pneumonitis, solitary or multiple lung abscesses, a broncho-pleural fistula or massive pleural effusions. Menon¹⁵ described three types of shadows in hepatopulmonary amoebiasis-

1. A homogenous, soft, dome shaped opacity with its base super-imposed on and covering a variable extent of the right dome of the diaphragm and demarcated from it by the denser hepatic shadow below.
2. A fluffy, irregular opacity overlying the right dome of the diaphragm imperceptibly merging with the hepatic shadow below.
3. A small, dense, triangular shadow in the right costophrenic angle and/or cardiophrenic angle (Fig. 30).

A pleural reaction in the right costophrenic angle may indicate rupture of the abscess into the pleural cavity.⁴ However, it could also be a sympathetic pleural effusion. Aspiration alone can differentiate between the two.
Uncommonly, left lobe abscess may perforate the left dome of the diaphragm to enter the pericardium. An X-ray chest then will show radiological signs of pericardial effusion with or without shadows of pleural or pulmonary amoebiasis on the left side. These have been discussed elsewhere in detail.

Plain radiography of the upper abdomen
The liver casts an appreciable shadow by virtue of the right lobe merging in the right hemidiaphragm above, the right lateral edge and the lower border sloping upward and medially. The left lobe, however, cannot be seen so easily because of its more central position and smaller size.
In amoebic liver abscess the plain X-ray abdomen may show an enlarged liver and the diagnosis may be suspected by the asymmetry of the liver margin as outlined by the lung or intestinal gas shadow. In an inferior surface abscess of the right lobe, a depressed hepatic flexure, transverse colon (Fig. 31) or right kidney may be visualised with edge of the liver lower than normal.
Margulies and Stoane¹⁶ included loss of hepatic angle as an important roentgen finding in amoebic liver abscess. Usually, presence of peritoneal fluid can be diagnosed on a plain X-ray of abdomen, when the right inferior margin of the liver is obscured in the supine position, otherwise seen clearly in normal persons. In 7 cases of amoebic liver abscess of Schmidt,¹⁶ the hepatic angle was lost both in supine and erect positions. This was because inflammatory changes in the liver were thought to obscure the adjacent omental and pericolic fat which normally delineate the hepatic angle.
Schmidt¹⁶ also detected, in addition to hepatomegaly, a dense distended gall bladder indenting the intestinal shadow attributed to associated ileus of the gall bladder, secondary to intense inflammatory changes in the liver. Interestingly this finding was not observed in his cases of acute cholecystitis or subdiaphragmatic abscess on the right side.
Though X-ray of chest and plain X-ray of the upper abdomen may be useful adjuncts in the diagnosis of amoebic infection of the liver, more precise radiological procedures are usually required to delineate the liver, with a detailed study of the abscess cavity. Pneumoperitoneum tomography, angiography and the use of contrast media, with intracavitary instillation of the same are some of these methods.

Intracavitary radiographic methods
Air. The injection of air into an amoebic abscess cavity following aspiration is an innocuous procedure. In 1953 Sloan and Freedman¹⁷ used it as aid in diagnosis and to evaluate the efficacy of therapy. In many hospitals this simple procedure is still being used. Although authorities like Wilmot mock at this procedure, their value cannot be underestimated in rural areas where liver scanning facility is not available. Figures 32 a,b,c, show the X-rays taken following the injection of air, after aspiration, delineating the abscess cavity.
lodised oil. Gupta and Khanna¹³ used Dionosil (oily or aqueous) or DiaginNol viscous as a contrast medium. Other authors have utilised Lipiodol or Myodil (Fig. 33). The medium can be injected into the abscess cavity and the films taken under fluoroscopic control. The visualisation of the amoebic liver abscess was labelled as "cavernogram" by Gupta and Khanna. They also injected a case of chronic empyema through a drainage tube and called it "pleurogram". Cavernogram is a useful procedure to derive information about the size, location and shape of the amoebic liver abscess. It also outlines a hepato-bronchial or a hepato-pleural fistula.
Barium sulphate. In 1970, Harding et al¹⁸ used a sterile suspension of a micro-opaque barium sulphate for radiographic visualisation of the amoebic liver abscess. The contrast medium is taken up by macrophages in the wall of the abscess cavity. It thus allows localisation of the abscess, observation of changes in the size, and an early detection of rupture. Bhasin et al¹⁹ have used micro-opaque barium sulphate to study the resolution time of the liver abscess.
Radioactive media. In 1971 Viranuvatti et al²⁰ used intracavitary instillation of radioactive ¹³¹I-BSP in five cases of amoebic liver abscess and compared its use with other media. After aspiration, 2 ml. Of ¹³¹ I labelled BSP and 40 ml. of 70% microlipiodol and air were instilled into the cavity. Subsequently the patients were scanned and X-rayed. Thus, a comparative study of visualisation by instillation of radio-opaque substance into the cavity versus intra-cavitary ¹³¹I-BSP liver scanning was carried out. They found that radioactive substances gave the best visualisation of abscess cavity and in detail and were superior to other contrast media. They were also more advantageous in determining the "resolution time" of the abscess, since their watery solution could flow easily and outline the cavity better than the oily radio-opaque iodized oil.
Thus, both positive contrast media such as lipiodol etc. and negative contrast media such as air have been used extensively by various workers to delineate, with greater accuracy, the cavity of an amoebic liver abscess. In fact a double contrast method of using air and lipiodol has also been employed. These methods have been claimed to be of diagnostic value and thus lead to effective therapy in such patients.
However, there has been strong opposition from other authors who condemn these methods labelling them to be of limited value and potentially dangerous in the management of an amoebic liver abscess.¹⁸ Apart from the theoretical risk of embolisation and secondary infection, it is felt that positive contrast media by their physical presence in the abscess cavity may interfere with the healing of the abscess.
However, as long as liver scanning procedures are not available at all the centres, these procedures can be relied upon as they certainly serve as guidelines to the clinicians regarding the site and the size of the abscess cavity in problem patients.

Pneumo-peritoneum. The primary object of pneumo-peritoneum is to distinguish between supradiaphragmatic extension of aQ amoebic liver abscess and primary pulmonary or pleural disease, by demonstrating adhesions between the superior surface of the liver and the diaphragm.^13,19-23 It is also of value as a diagnostic procedure in delineating the abnormalities in the hepatic contour caused by an amoebic abscess.
400-1000 ml. of air or carbon-dioxide is introduced into the peritoneal cavity. In 1921, Alvarez²² introduced the use of carbon-dioxide therapy eliminating one of the major hazards of the procedure-namely fatal air embolism. The procedure usually does not cause the subject any discomfort and gives adequate contrast for demonstarting adhesions²³ (Fig. 34).

Studies with contrast media such as barium-meal examination, intravenous pyelography and cholecystography may also be used to delineate the hepatic contour.
Barium meal studies. Displacement of adjacent organs is common with large abscesses of either the right or left lobe. A barium filled transverse colon may well be seen displaced downwards and posteriorly by a huge left lobe abscess. Baker and Murray²⁴ have reported oesophageal obstruction in a case of amoebic liver abscess. In some cases, a displacement of an elongated intra-abdominal portion of the oesophagus has been reported.²⁵ Radiological changes produced by the enlarged left lobe are discussed elsewhere in detail.
Nephrogram may occasionally demonstrate distortion of the calyceal pattern of the right kidney simulating a renal neoplasm.²⁶ Likewise, the right ureter may be displaced by a right lobe abscess simulating a retroperitoneal cyst.²⁷
Cholecystogram. An inferior abscess of the right lobe of liver may produce a distended gall bladder or its displacement downwards and laterally.²⁸

Total body opacification
In 1974, Chang et al²⁹ while studying the right kidney, accidentally discovered that sodium diatrizoate infusion in hepatic tomograms demonstrated the presence of an amoebic liver abscess and the visualisation of the walls of the abscess was found to be of value. Morin et al³⁰ ascertained the presence of hepatic abscess in two adults by total body opacification. This procedure identified the hypovascular nature of the abscess, thus differing from vascular hepatomas. McCroft³¹also used the technique of total body opacification and tomography to prove the existence and location of amoebic liver abscess in two of his patients.

Selective hepatic arteriography
Amoebic liver abscess is a low-grade inflammatory process which is seen as an avascular lesion at angiography.³² Arteries and veins are displaced and a rim of compressed parenchyma surrounds a lucency in the hepatogram phase. The injection of thorium dioxide previously used for hepatography has, now fallen into disrepute.
Instead of hepatic angiography, coeliac arteriography, splenoportography and inferior- venocavography have been used by some authors.³³ Wallace also tried hepatosplenography after thorotrast injection to delineate amoebic liver abscess.³⁴
Thus selective angiography is an effective method of visualizing the liver. However, would any one opt for such a traumatic and invasive investigatior in centres where modern, non-invasive investigations are available?

In conclusion it is likely that with the advent of isotope liver scanning and ultrasonography, many of the radiological investigations may soon become obsolete. However, a simple, non-invasive and universally available investigation like X-ray chest, would even today, quite often prove of immeasurable value, in making a diagnosis of a superior surface amoebic liver abscess.

References

1. Munk, L 1, Brit. J Radiol., 1944, 17, 48
2. Lamont, N M, and Pooler, N R. Quart. J. Med 1958, 27, 390
3. Shanks, S C, and Kerley, P. Textbook of X-Ray Diagnosis, 5th Edition, H K Lewis & Co., 1976.
4. Isaac, F. Radiol, 1945. 45. 581
5. Chaves, F J Z C, Am. J. Gastroent., 1977, 68, 273.
6. Schorr, S. and Schwattz, A, Am. J Roentgen.,1951, 66, 546.
7. Ramachandran, S. Iayawardhana, D L N. et al, Post Grad Med. J, 1971, 47, 615.
8. Ramachandran, S. Goonatillake, H D, et al. J Trop Med. Hyg., 1975, 78, 133.
9. Ramachandran, S. Jayawardhana, D L N. et al, Post-Grad Med. J. 1976, 52, 154.
10. Rowland, H A K, J. Trop Med. Hyg., 1963, 66, 113
11. Subramaniam, R. and Madangopalan, N. Amoebiasis, Sandoz's Monograph, 1970.
12. Monroe, L S. (IN) Bockus, Gastroenterology, Vol IV, 3rd Edition, W 13 Saunders & Co Philad 1976.
13. Gupta, S K, and Khanna, M N. Clin. Radiol, 1972, 23, 219.
14. Galloway (as quoted in Textbook of X-ray Diagnosis Shanks & Kerley). 1970.
15. Menon, N K, Dis Chest, 1964, 46, 219.
16. Margulies and Stoane (as quoted hy Schmidt,A G. and Stockton, Am. J Roentgen, 1969, 107, 47).
17. Sloan, S. and Freedman, T. Arch. Int. Med., 1953, 91, 550
18. Harding, T. Lewis, E A, et al, Clin. Radiol., 1970,21, 68.
19. Bhasin, A S. Mittal, K P. et al, J.Trop. Med. Hyg., 1977, 80, 169.
20. Viranuvatti, V, Suwanik, R. et al, Am. 1. Gastroent., 1971, 55, 13.
21. Ellman, B. Macleod, I N. et al, Brit. Med. J ,1965, 2, 1406.
22. Alvarez, W C, Am. J. Roentgen., 1921, 8, 71.
23. Clark, R H P. Bercovitz, Z T. et al; Am. J. Trop. Med. Hyg., 1948, 28, 545.
24. Baker, H M, and Murray, J A, Cent. Afr J. Med., 1969, 51, 129.
25. Taveras Jaran and Ross Golder. Roentgen. Of Abdomen, 1st Edition, Williams & Wilkins Co..Baltimore, USA. 1961, 1450.
26. Ramakrishnan, as quoted by Shanks, 1976.
27. Saratchandra and Gairaj, as quoted by Shanks, 1976.
28. Meschan, I, Analysis of Roentgen Signs in General Radiologv. W B Saunders & Co., Philad., 1973.
29. Chang, S. Chang, S. et al, Am, J. Trop. Med., 1974, 23, 31.
30. Morin, M E, Baker, D A, et al, Am, Med. Ass., 1976, 236, 1607.
31. McCroft, J J. West / Med. 1976, 124, 426.
32. Reuter, S R. and Redman, H C, Gastrointestinal Angiography, W B Saunders & Co., Philad., 1977.
33. Balasegaram, M, Ann Surg., 1972, 175, 528.
34. Wallace, M V, J Am Med. Ass, 1944, 125, 775.