Diagnostics in Liver Diseases 6 Sonography

6 Sonography

Page:

1 Physical principles 126

2 General principles and echo types 126

3 Artefacts 126

4 Section planes 127

5 Examination of the liver 127

5.1 Normal findings 127

5.2 Indications 128

6 Diffuse liver diseases 128

6.1 Congestive liver 129

6.2 Fatty liver 129

6.3 Liver cirrhosis 129

6.4 Portal hypertension 130

6.5 Portal vein thrombosis 130

6.6 Obstruction of hepatic veins 130

6.7 Jaundice 131

6.8 Ascites 131

7 Circumscribed liver diseases 131

7.1 Echofree lesions 131

⫺ cysts, abscesses

7.2 Hypoechoic lesions 132

⫺ adenomas, focal fat accumulation, FNH, lymphadenopathy, etc.

7.3 Hyperechoic lesions 133

⫺ haemangiomas, NRH, echinococcus, HCC, metastases

8 Examination of the spleen 135

9 Special sonographic techniques 135

9.1 Sonography-guided puncture 135

9.2 Limitations and complementary procedures 136 9.3 Colour-encoded Doppler sonography 136

9.4 Endoscopic sonography 137

앫 References (1⫺123) 138

(Figures 6.1 ⫺6.20; tables 6.1⫺6.6)

Ultrasonography is a routine examination measure in the diagnosis of hepatobiliary diseases.

After years of experience with ultrasonography, a well- trained sonographer will produce highly reliable results in the field of liver imaging. • The method neither in- conveniences nor harms the patient. Thus there is a broad range of indications for examination by ultra- sonography in patients with hepatobiliary diseases.

There are no contraindications. Specific diagnostic statements, however, are rarely possible.

1 Physical principles

Mechanical oscillations of > 18,000 Hz are referred to as ultra- sound. In solid bodies, sound waves spread longitudinally as well as transversally; in fluids, gases or body tissue, however, waves only spread longitudinally. The average velocity of sound conduction (v) in tissues is approximately 1,500 m/sec.

v (m/sec) reflection

Air 331 99.88%

Fat 1,476 0.12%

Water 1,496 0.00%

Muscle 1,568 0.48%

Liver 1,570 0.30%

Bone 3,360 46.00%

Ultrasonic waves are generated by a piezoelectric crystal and emit- ted via a sound-conductive medium. These waves are reflected, broken, dispersed and absorbed by boundary layers. The piezo- electric crystal also acts as a sound-wave receiver and registers the modified ultrasonic waves ( ⫽ reciprocal piezoelectric effect). They are then converted into an electric signal and displayed by means of oscilloscopic imaging.

This will result in wavelengths of 0.1 to 1.5 mm in the diagnostic frequency range of 1 to 20 MHz. The 3.5 MHz range has proved suitable for abdominal sonography based on the B mode ( ⫽ brightness mode). A 5 MHz transducer will increase the resolution (0.6 mm axially, 1.2 mm laterally) but will also decrease the depth of penetration of sound waves (10 cm). Resolution in tissues is approx. 1 mm towards the sound beam ( ⫽ axial) and 2⫺6 mm perpendicular to the axis of the sound beam ( ⫽ lateral). Reso- lution thus depends on ultrasonic frequency and focus. An increase in ultrasonic frequency therefore produces higher resolution and greater penetration depth. • A linear scanner applies parallel sound waves, while a sectorial scanner emits waves of a focal or curved source. Both of these techniques rapidly produce images at a rate of 30/sec ( ⫽ real time). This allows (1.) direct monitoring of in- ternal physical motion and (2.) mobile handling of the transducer for dynamic portrayal and analysis of the image. In this way, the image can be frozen at any stage for documentation purposes, or the process of examination can be videotaped. (1)

2 General principles and echo types

The sonographic technique is based on reflection; it works by measuring and portraying differences of im- pedance. Differences in acoustic characteristics of structures and tissues are the basis of sonographic differentiation.

The product of sonic speed in the respective tissue, together with the specific tissue density, yields the acoustic impedance (⫽ acous- tic resistance). Each medium has its own value with regard to sonic resistance. The partition between two media of different acoustic impedance is termed acoustic interface (which should not be con- fused with the anatomical interface). The magnitude of difference of impedance determines the acoustic effectiveness of this in- terface. • Continuous technical development in grading the degrees of brightness of image points has made it possible to define the different shades of grey between the two extremes of reflection, black and white. Various organs, including the liver, exhibit charac- teristic shades of grey. Connective tissue, fat deposition, liquids, air or cellular infiltration will each alter the grey scale. This has considerably improved the differentiation of both the organs and the findings ⫺ thus a more exact diagnosis is guaranteed. (1)

Ultrasound impulses emitted into the body are reflected by interfaces, so that boundaries of organs are received as contour echoes, whereas internal heterogenicities of organs are received as structure echoes. Analysis of the echo structure is seen as a specific feature of ultra- sonography. The sonic energy reflected by interfaces is visible on the screen in the form of bright spots. • As a rule, the echo represents an acoustic interface within a solid structure. Sonography can differentiate between various zones: (1.) hyperechoic, (2.) isoechoic, (3.) hypo- echoic, and (4.) echofree. Echofree areas point to the absence of acoustic interfaces, which as a rule corres- ponds to a “liquid” consistency. (s. figs. 6.4, 6.11) 䉴 The terms “hypodense”, “isodense” and “hyperdense”

depict pathological changes observed in CT. For physical reasons, they are reserved for CT examinations and should not be used in sonography. (s. p. 171)

3 Artefacts

Sonographic artefacts may at times impede the assess-

ment considerably. (47) Artefacts are mainly associated

with the method applied and hence can often be elimin-

ated simply by changing the examination technique. The

occurrence of artefacts is largely fostered by extensive

differences in acoustic impedance within a particular

substrate, especially in the abdomen. In some 5% of

cases, artefacts render sonographic examination impos-

sible. They may appear in multiple forms:

1. amplification artefacts 5. mirror artefacts 2. arched artefacts 6. precursory artefacts 3. artificial sedimentation 7. repetitive echoes 4. contour aberrations 8. shadow zones

4 Section planes

The ultrasonographic examination of the liver is per- formed on the basis of section planes. This investigation is best carried out by adhering to a well-established system.

Movement, i. e. repositioning the transducer, should be reduced in favour of tilting and requiring the patient to breathe in deeply. (s. fig. 6.1)

Fig. 6.1: Section planes in the sonographic examination of the liver: longitudinal sections (I-IV), transverse section (V), subcostal section (VI), intercostal section (VII)

With the help of these different planes, it is possible to dis- play numerous anatomical formations and structures. (s.

tab. 6.1)

5 Examination of the liver

5.1 Normal findings

The assessment of the liver is based on knowledge of the sonographic anatomy as well as on external, internal and dynamic criteria. Evaluation of sonographic results must allow for the fact that normal findings vary with body height, body weight, shape of the thorax, habitus and age. (32 ⫺34, 81, 89) The normal liver parenchyma reveals equal or marginally higher echogenicity than the (right) renal parenchyma. (64) From the age of 60 ⫺65 onwards, the liver loses up to a third of its weight (“se- nile atrophy”, F.T h. Frerichs, 1858 ). (s. tabs. 6.2, 6.3) (s.

fig. 6.2)

Close to the round ligament of the liver, the lower liver margin is somewhat drawn in and rounded off. There is a circular to oval, precisely circumscribed hyperechoic

1. Longitudinal sections size, shape and internal

I, II structure as well as mo-

tility and compressibility of the left lobe of liver 2. Longitudinal sections size, shape and internal III, IV structure as well as mo- tility and compressibility of the right lobe of liver;

gall bladder, porta hepatis, portal vein, inferior vena cava, common hepatic duct 3. Transverse section V porta hepatis, falciform

ligament, left lobe of liver 4. Subcostal section VI internal structure of the

right lobe of liver with portal vein, portal vein branches and hepatic veins; gall bladder, intra- hepatic bile ducts 5. Intercostal section right lobe of liver with

VII subphrenic region

Tab. 6.1: Sonographic diagram of hepatobiliary structures and for- mations using five section planes

Fig. 6.2: Normal liver with portal vein (VP), inferior vena cava (VC) and right branch of the portal vein (RHA). Subcostal plane:

1 ⫽ left branch of VP; 2 ⫽ right branch of hepatic vein; 3 ⫽ cranial diaphragm

change of structure, marking the border between the right and left lobe of liver. The higher the fatty tissue content of the round ligament of the liver, the more striking this normal finding appears to be. (s. p. 171!) Dorsal to this finding, a shadow zone may be generated by the absorption of sound waves due to connective tis- sue present at this site. • Likewise, the structure of the falciform ligament as shown by ultrasonography must be observed and correctly interpreted. (39) (s. figs. 6.3;

8.1!) • The papillary process of the caudate lobe ( ⫽ seg-

ment I) may be misinterpreted as being a space-occupy-

ing proliferation. (43) • An enlarged Riedel’s lobe shaped

like a tongue (as a variation of the norm) may extend

as far as the pelvic region. • The sonographic definition of liver segments ⫺ particularly by means of laparos- copic sonography ⫺ has proved valuable for surgical procedures. (27, 37, 46, 80) (s. p. 138)

1. External criteria

앫 Shape (dorsal and ventral contour) 앫 Margin (inferior border of the liver) 앫 Surface

앫 Size (measuring of the right liver lobe in the MCL) 앫 Position (e.g. downward displacement due to pulmonary

emphysema) 2. Internal criteria

앫 Circumscribed alterations 앫 Internal structure 앫 Sound conduction 앫 Vascular structures 앫 Anatomical segmentation 3. Dynamic criteria

앫 Compressibility 앫 Pain upon palpation 앫 Respiration-related motility 앫 Vascular pulsations

Tab. 6.2: Criteria for the sonographic examination of the liver

Size 12 ±3 cm in the MCL (vertical) 9 ±1 cm in depth

Shape wedge-shaped

Margin right ⱕ75°, left ⱕ45°, sharp-edged Surface smooth, slightly concave

Contour dorsally and ventrally extended Internal structure fine homogeneous reflexes; somewhat

more echogenic than the kidney paren- chyma, slightly less echogenic than the pancreas; rather “dark” and “sharply contrasted”; the area of the caudate lobe is usually more hypoechoic

Quadrate lobe 43 ±8 mm

Liver veins easily depictable as far as the periphery,

< 6 mm, no wall echoes, lumen fluctu- ations during Valsalva’s manoeuvre Portal vein mostly < 12 mm, ventral to the inferior

vena cava, marked wall echoes, branch- ing distribution from the porta hepatis, no respiratory modulation

Hepatic artery branching with pulsations in the porta hepatis

Bile ducts thin, bright reflex bands, mostly demon- strable (< 4 mm; < 50% of the ac- companying portal vein branch) Ductus choledochus ⱕ7 mm; following cholecystectomy and

in old age ⱕ11 mm

Tab. 6.3: Normal ultrasound findings of the liver. (The liver is an organ rich in variation and thus the figures given should be seen merely as points of reference to be treated individually)

5.2 Indications

In the case of suspected liver or biliary diseases, US is always indicated. Sonography has become a routine examination technique. (56, 75) (s. tab. 6.4)

Fig. 6.3: Falciform ligament of the liver: transverse and longitudinal image (simulating “liver metastasis” with concomitant rectal car- cinoma) (s. fig. 8.1!)

1. Diffuse liver diseases

앫 diagnostic confirmation ⫹⫹

앫 detection of complications ⫹⫹

앫 differentiation ⫹

앫 malignant degeneration ( ⫹)

2. Circumscribed liver diseases

앫 diagnostic confirmation ⫹ ⫺ ⫹⫹

앫 follow-up monitoring ⫹ ⫺ ⫹⫹

앫 biopsy aid ⫹⫹⫹

앫 differentiation ⫹

앫 diagnosis by exclusion ( ⫹)

3. Jaundice and cholestasis

앫 differentiation ⫹⫹

4. Endoscopy or surgery

앫 prior clarification ⫹⫹

앫 follow-up examination ⫹⫹

5. Screening after abdominal trauma ⫹⫹

6. Assessment of portal veins and hepatic veins ⫹ 7. Liver biopsy

앫 targeted biopsy ⫹⫹⫹

앫 targeted fine-needle biopsy ⫹⫹⫹

8. Perihepatic lymphadenopathy ⫹

Tab. 6.4: Indications for sonography of the liver (accuracy or val- idity: ⫹⫹⫹ ⫽ very high (>90%), ⫹⫹ ⫽ high (>70%), ⫹ ⫽ mod- erate (> 50%), ( ⫹) ⫽ minimal, helpful in individual cases)

6 Diffuse liver diseases

A normal sonographic finding does not necessarily rule

out a diffuse liver disease (30% of cases). • The morpho-

logically normal liver exhibits 20 ⫺25% false-positive

ultrasonographic findings in terms of a diffuse liver dis-

ease. Indeed, in 35 ⫺40% of all ultrasound findings in the

liver, results can be classified as diffuse parenchymal pro-

cesses. Ultrasonography is no substitute for histology. •

Neither subclassification nor aetiological classification

of the various diffuse diseases of the liver is possible by

means of ultrasonography. There are, however, criteria

and constellations of findings which are typical of some

forms of disease. (6, 36, 42, 58, 87, 92)

The increase in echogenicity (i.e. increase in density) of a homogeneous, frequently coarsened structure with de- creased sound conduction in the form of a “structurally dense liver” is a reliable indicator of a diffuse liver dis- ease. This increase in intensity and frequency of echoes yields the image of a “bright (white) liver” (e. g. fatty liver, haemachromatosis). A diffuse liver disease can also be accompanied by a decrease in echogenicity, which is why the hypoechoic liver is known as a “dark liver” (e. g. acute liver congestion, acute viral hepatitis, amyloidosis). (s. fig. 6.4)

1. reduced 1. even

2. irregular

3. speckled

4. dotted 2. loose

3. moderately enhanced

4. clearly enhanced

Reflex density Reflex distribution

Fig. 6.4: Structural changes in the hepatic parenchyma with regard to reflex density and reflex distribution

䉴 In terms of ultrasonography, various liver diseases appear as diffuse changes in the parenchyma. They usu- ally exhibit uncharacteristic findings. In cases where the diagnosis is uncertain, particularly after employing CT, liver biopsy (or laparoscopy) is then indicated:

1. acute hepatitis 6. congestive liver

2. chronic hepatitis 7. systemic haematological diseases 3. fatty liver 8. liver cirrhosis

4. metabolic diseases 9. granulomatous liver diseases 5. infectious diseases 10. diffuse metastatic spread

6.1 Congestive liver

The congestive liver is enlarged and tender upon pres- sure. The hepatic veins are dilated (normal diameter of the right hepatic vein is < 6 mm) and may be traced to the periphery. The inferior vena cava is widened and no longer exhibits the typical respiration-dependent double pulsation. Characteristic of the chronic congestive liver is the increase in echogenicity, so that during the pro- gressive course of the disease, the findings can resemble those of a micronodular cirrhosis. (see chapter 39.1)

6.2 Fatty liver

The markedly enhanced echogenicity of the fatty liver (“bright liver” or “large white liver”) is due to the high number of water/fat interfaces. (s. fig. 6.5) • False-posi- tive findings occur in roughly 20% of cases. In decreas- ing hepatic steatosis (< 30%), the sensitivity of recogni- tion is reduced accordingly. Sonographically, a fatty de- posit is evidenced in > 20% of hepatocytes in about 90%

of cases. There is a good correlation between ultrasono- graphy and histological results, as there is between hypertriglyceridaemia and diabetes. The liver volume measured by ultrasonography is elevated, a fact which is also suggested by the rounded liver margin. The interior structure is inhomogeneously coarsened (in contrast to the kidney parenchyma), sound conduction continu- ously decreases dorsad, and hepatic veins are poorly imaged. In cases of pronounced fatty liver, the portal vein can be dilated as a result of portal hypertension. In 20 ⫺30% of cases involving a fatty liver, the sonographic findings are mistaken for liver fibrosis, haemochro- matosis, micronodular cirrhosis, chronic hepatitis C, thesaurismoses and systemic diseases. However, an acute fatty liver of pregnancy usually produces a normal sonographic image. • Occasionally, the distribution of fat shows focal disorders. (s. figs. 8.2 ⫺8.4) A focal accu- mulation of fat in the liver appears as a hyperechoic area. In contrast, a focal reduction of fat within a fatty liver displays a hypoechoic area, primarily ventral to the hilus branching of the portal vein, often triangular in shape, but also with a rounded form. (64) (s. figs. 6.12, 6.13) (see chapter 31.3)

Fig. 6.5: Pronounced fatty liver (so-called enlarged white liver)

6.3 Liver cirrhosis

Sonographic findings in liver cirrhosis are polymorphic on account of the vastly differing individual hepatic structures (loss of parenchyma, regenerative nodes, con- nective tissue, fat depositions, compressed hepatic veins, etc.). (s. figs. 6.6, 6.7)

The size of the cirrhotic liver varies between hepato-

megaly, normal finding and atrophy. The regular pro-

Fig. 6.6: Liver cirrhosis with ascites (longitudinal section): the left lobe of liver is rounded and plump; intrahepatic vessels are re- duced. Irregular and inhomogeneous structure. Clear undulatory limitation (arrow) on the underside due to nodular transformation.

Wide hypoechoic fringe due to ascites

portion of both liver lobes may have changed, resulting in an asymmetric enlargement or reduction of the liver.

As a rule, ultrasonography shows the left lobe of liver to be affected to a greater extent and to be larger than the right lobe. The surface appears finely or roughly coarsened ⫺ particularly when applying a 5 or 10 MHz transducer. The contours can be seen as rounded edges and are frequently shaped like a bird’s head. Echogenic- ity is increased with reduced sound conduction. The quadrate lobe (segment IV) often shows a reduced dia- meter (< 30 mm). Since the caudate lobe is less affected by this cirrhosis-related transformation due to its own special blood supply, it appears enlarged, it is strikingly hypoechoic and its contours are clearly delineated ⫺ a finding which always points to cirrhosis. Prominent ar- teries may represent the “pseudo double-barrelled-gun”

phenomenon. (cf. fig. 6.8) • The Doppler perfusion index (DPI) is usually increased (E. L een et. al., 1993) . In cases of pronounced cirrhosis, the sensitivity of sonography is 85 ⫺90% and it has a specificity of 80⫺95%. (2, 9, 16, 26, 34, 45, 59, 77, 79, 93, 105, 106) (see chapter 35)

6.4 Portal hypertension

Detection of portal hypertension (see chapter 14) is of crucial importance in the diagnosis of cirrhosis. The sensitivity is 76 ⫺80% and the specificity 100%. The fol- lowing findings may be present: (1.) dilation of the por- tal vein (> 1.5 cm), (2.) calibre leap between the extra- and intrahepatic segments of the portal vein in the porta hepatis (so-called portal vein amputation), (3.) dilation of the splenic vein (> 1.5 cm), (4.) widening of the hep-

atic artery, (5.) splenomegaly with contact between the liver and spleen (s. p. 135), (6.) dilation and/or rigid cali- bre of the superior mesenteric vein, (7.) detection of collateral vessels, particularly in both the splenic and the hepatic porta as well as in the anterior wall of the stomach, (8.) thickening of the stomach wall to > 22 mm

(73) and of the gall-bladder wall (74) , and (9.) recanal- ization of the umbilical vein (s. fig. 6.7), which can be traced to the umbilical region ( ⫽ Cruveilhier-von Baumgarten syndrome). (s. fig. 7.4) (23, 35, 93, 98)

Fig. 6.7: Alcohol-induced cirrhosis with recanalized umbilical vein ( ⫽ hypoechoic band: see arrow); (PV ⫽ left branch of the portal vein; CPA ⫽ cockade of pyloric antrum) (s. fig. 14.12)

6.5 Portal vein thrombosis

Portal vein thrombosis (see chapter 39.3.3) with gradual or incomplete obstruction merely produces anastomoses and develops asymptomatically. After complete obstruc- tion, ultrasonography will show paraportal, angiomatous anastomoses, predominantly in the porta hepatis ( ⫽ cavernous transformation). (95) Splenomegaly is evi- denced with a normal-sized liver. The portal vein is not detectable. • Colour Doppler sonography has become of paramount importance in portal vein system diagnostics.

6.6 Obstruction of hepatic veins

The Budd-Chiari syndrome (see chapter 39.2.1) as well

as hepatic vein obstruction (type VOD) are discernible by

the total absence of hepatic veins or by their diminished

lumen. The residual lumen of a vein often shows hyper-

echoic thrombotic material. Sometimes membraneous

webs of varying lengths consisting of hyperechoic struc-

tures with acoustic shadows are visible in the lumen of the

inferior vena cava. The liver is enlarged. Due to its own

venous drainage system, the caudate lobe ( ⫽ segment I) is

noticeably augmented (normal 6.8 ± 1.3 cm) and usually

presents a hypoechoic structure. In most cases, ascites is

also present. (40, 71, 95)

6.7 Jaundice

Diagnostic accuracy in differentiating between obstruct- ive and nonobstructive jaundice (see chapter 12) can be achieved in up to 90% of cases. (s. tab. 12.1) • Ultra- sonography will reveal the cause of an obstruction in 30 ⫺60% of patients and localize this occlusion in al- most all cases. (s. fig. 25.8!) Ultrasound shows the con- gested intrahepatic bile ducts as resembling vessels which run parallel to the portal vein branches and are mostly found in a ventral position (so-called “double- barrelled-gun phenomenon”). (s. fig. 6.8) These congested bile ducts can also take on a “lakeland plain” pattern.

(s. fig. 6.9) Close to the porta hepatis, dilated bile ducts may also appear in a radial pattern (so-called “wheel- spoke phenomenon”). An extrahepatic occlusion is gen- erally easier to recognize if there is a change in the lumen of the common bile duct when Valsalva’s man- oeuvre is applied. (67)

Fig. 6.8: Double-barrelled-gun phenomenon in obstructive jaun- dice (see arrow)

6.8 Ascites

Ascites (see chapter 16) can be successfully diagnosed even in small quantities (< 200 ml). Given a favourable lo- cation (preferred sites), it is even possible to detect fluid volumes of < 50 ml. • Such a preferred site of ascites is the so-called Morrison’s pouch: the inferior surface of the right lobe is anterior to the colon, and posterior to the perirenal fatty tissue. It is separated from the latter by this peritoneal recess, where a small amount of ascitic fluid generally collects. Here it can be detected sonographically at an early stage. • Ascites coats abdominal organs, es- pecially the liver and the spleen, and is shown as an echo- free fringe zone. (s. fig. 6.6) Occasionally, intestinal loops

Fig. 6.9: Blockage and dilation of the intrahepatic bile ducts (“lakeland plain”) in obstructive jaundice

float freely in the ascitic fluid, revealing the mesenteric at- tachment (so-called “sea-anemone phenomenon”). In add- ition, concomitant pleural effusion may be visible as an echofree zone above the diaphragm. Similarly, pericardial effusion can appear as an echofree zone at the cardiac apex. As a result of adhesions, ascites is possibly discern- ible in honeycomb-like compartments of various shapes and sizes. (101)

7 Circumscribed liver diseases

Focal changes are detected by ultrasonography with a sen- sitivity of 60 ⫺95%. The frequency of focal hepatic lesions is 3 to 5% (up to 10%) of all ultrasonic scans. Sonography does not allow a distinction to be made between benign and malignant lesions. (19, 41, 55, 58, 61, 69, 72, 94, 99) • Focal changes in echogenicity give rise to an inhomogeneous he- patic structure. This can be caused by focal intrahepatic findings, vascular abnormalities, protrusions of the liver surface or circumscribed congestion of the bile ducts. The greater the difference in acoustic characteristics and the clearer the demarcation between focal changes and the surrounding liver parenchyma, the easier it becomes to detect such a focus by means of ultrasonography.

Calcium-dense foci of a diameter of 3 ⫺5 mm, cystic for- mations of 5 ⫺8 mm and solid foci of 5⫺10 mm can there- fore be identified. Such irregular permeation of the liver with different tissue structures will result in a homo- geneously irregular image in ultrasonography, resembling a diffuse liver disease. • Detection of small foci close to the capsule is easier if a 5 MHz transducer is used.

7.1 Echofree lesions

Echofree lesions are: (1.) cysts, (2.) abscesses as well as

necrotic colliquation of metastases, and (3.) fresh or re-

liquefied haematomas. (48, 85) • Caroli’s syndrome (83) , Osler’s syndrome (95) and peliosis hepatis (54) may be assigned to this class as well. These echofree lesions ex- hibit distal sound amplification. (s. fig. 9.4)

Cysts (see chapter 36.4.14) are typically echofree, round, with a smooth, thin, limiting wall; they show a subsequent (dorsal) amplification of echo. Sizes of 0.5 cm and more can be clearly identified using ultrasound.

The prevalence of (dysontogenetic) congenital hepatic cysts in the general population is 3 ⫺5%. They are mostly found in the right liver lobe, more frequently in women than in men, and are multiple in around 25% of cases. The cysts are usually 1 ⫺3 cm in size. However, they develop at differing rates and can grow to as much as 20 cm in diameter. When they reach a certain size, the cysts can have suppression effects within the liver itself and also on adjacent organs. (s. figs. 6.10; 36.9) Bleeding into the cysts can cause internal sedimentary echoes in the cyst lumen. (11) • In autosomal dominant polycystic liver disease (96) , varying numbers of different- sized cysts are found in the liver and in the kidneys, sometimes also in the pancreas and the spleen. The overall frequency is 0.5 ⫺0.8%. (s. figs. 8.6; 36.12) • Hy- datid cysts often comprise daughter cysts with a doub- ling of the wall contour. The echo image may be ex- tremely heterogeneous. The cyst walls are sometimes hyperechoic and thicker (with occasional calcification);

they mostly show a smooth contour. (20, 52) A typical finding is the presence of smaller daughter cysts inside the mother cyst, e. g. honeycomb pattern in type II b or III. (s. figs. 25.15, 25.16) For purposes of sonographic classification, it has proved beneficial to have five categ- ories. (30)

Fig. 6.10: Liver cyst. (C ⫽ cyst; HV ⫽ hepatic vein; VC ⫽ inferior vena cava; L ⫽ right lobe of liver; arrows ⫽ hyperechoic area be- low the cyst) (s. figs. 8.6; 36.9)

Abscesses (see chapter 27) display fine interior echoes, irregular shapes and blurred walls. There are repetitive echoes (so-called “comet-tail phenomenon”). Differen- tiation between an amoebic and pyogenic abscess is not possible with sonography. In the light of the multiple

aspects of ultrasound findings regarding abscesses, a three-type classification system has been drawn up in line with the respective internal structures, which has been broken down further into types IIIb, IV and V. (38, 68, 91) (s. pp 175, 513) (s. figs. 6.11; 9.2; 25.1)

Type I: echofree Type II: hypoechoic

Type III b: hyperechoic without sound amplification Type III a: hyperechoic

with sound amplification

Type IV: mixed type Type V: reflective Different types of abscesses

Fig. 6.11: Synoptic view of 5 different types of abscesses as shown by means of sonographic examination (38)

7.2 Hypoechoic lesions

The following lesions may be hypoechoic: (1.) metas- tases, (2.) liver cell carcinoma, (3.) adenomas, (4.) focal nodular hyperplasia, (5.) abscesses, (6.) haematomas, (7.) early liver infarction, (8.) foci showing reduced fatty infiltration, (9.) lymphomas, and (10.) lipomas. In indi- vidual cases, differentiation between a benign and a ma- lignant structural defect may cause considerable diffi- culties. (59) (s. fig. 9.4)

Adenomas (see chapter 36.4.1) are encapsulated hyper- vascularized tumours. They have a smooth wall, are variable in size and are round to oval. As a rule, they are solitary (right more than left). They can be both hypoechoic as well as (more rarely) hyperechoic, pos- sibly due to bleeding within the adenoma.

Focal accumulation of fat is predominantly found either

near to the falciform ligament of the liver or to the por-

tal branches as well as in segment IV. It appears as an irregularly limited hyperechoic focus, which is often dif- ficult to classify by differential diagnosis and can even be misinterpreted as being malignant. Blood vessels run normally. (31, 78, 97) (s. figs. 8.2 ⫺8.4) • Focal reduction of fat in a fatty liver (in alcohol abuse, diabetes mellitus, chemotherapy) is predominantly found close to the gall bladder or the caudate lobe. Mostly it has a triangular shape, and there is a vascular association. In rare cases, it shows a round shape. (13) (s. figs. 6.12, 6.13)

Fig. 6.12: Focal reduction of fat in a fatty liver: hypoechoic, tri- angular shape (see arrow) near to the right branch of the portal vein (PV) and the gall bladder (GB); (VC ⫽ inferior vena cava)

Fig. 6.13: Focal reduction of fat in a fatty liver: hypoechoic round shape (see arrows)

Focal nodular hyperplasia (see chapter 36.4.2) ⫺ as a benign hepatocellular tumour ⫺ is mostly hypoechoic.

Occasionally, it can be differentiated as a protruding contour or a pediculate liver tumour. Compression of the surrounding liver tissue may be the cause of a visible

“capsule”, which actually has no anatomical structure of its own. At a size of > 3 cm, fibrous septa and arteries

can resemble a typical wheelspoke structure, often with a clearly visible central core. • The blood circulation within the tumour can be visualized by means of con- trast medium: early arterial hyperperfusion is usually in evidence. As a hypervascularized lesion, FNH some- times has afferent arteries, which may surround the lesion like a basket. (s. fig. 6.14)

Fig. 6.14: FNH: Hypervascularized internal structure with hyper- echoic star-shaped scar ( ⫽ wheele-spoke pattern)

Macronodular tuberculosis ⁽⁵¹⁾ , granulomatosis of the liver (60) and particularly liver metastases may likewise appear as focal liver diseases with differing echogenicity.

Haematomas are initially hyperechoic, but turn hypo- echoic when they liquefy within a few days, and even become echofree at a later stage. “Older” haematomas usually revert to being hyperechoic. (48, 85)

Lymphadenopathy: Enlarged abdominal lymph nodes are frequently detectable in acute viral hepatitis (A ⫺D) and chronic hepatitis (B, C) ⫺ occasionally also in auto- immune hepatitis, primary biliary cirrhosis and primary sclerosing cholangitis. (12) Lymphadenopathy corres- ponds to the histological stage of chronic hepatitis C.

(21) Sonographic determination of perihepatic lymph- adenopathy predominantly in the hepatoduodenal ligament (57, 118) is not regarded as a pathognomonic sign of a underlying malignant disease.

7.3 Hyperechoic lesions

Hyperechoic lesions may be (1.) metastases, (2.) liver carcinoma, (3.) cholangiocarcinoma, (4.) haemangioma, (5.) hamartoma, (6.) nodular regenerative hyperplasia, (7.) old haematoma, (8.) focal fatty infiltration, (9.) echinococcus alveolaris, (10.) scarring, (11.) calcification (s. fig. 9.4), and (12.) aerobilia (s. fig. 32.6).

Haemangioma (see chapter 36.4.4) ⫺ the most common

benign finding ⫺ is usually detected incidentally as a

homogeneous and frequently circular focal lesion. It is

found in solitary or multiple (10 ⫺30%) form. The fre- quency is 8 times higher in women than in men. The preferred localization is close to the diaphragm or larger vessels. A haemangioma does not produce any bio- chemical variations from the norm. Occasionally, there is an afferent or efferent vessel. • The typical haemangi- oma (approx. 90%) has a diameter of 1 ⫺4 cm; it is uni- formly hyperechoic (a white tumour like a “snowball”).

It forms a clearly defined yet irregularly shaped (“lobu- lated”) surface with a distal amplification of sound. It is often found close to a hepatic vein. It has no halo. A centripetal filling in a high-flow haemangioma (like the CT “iris diaphragm phenomenon”) can be detected using echo contrast CHI. In a small number of cases (< 0.5%) it may happen that a malignant tumour is hid- den beneath the image of a haemangioma, and thus goes unnoticed. (50) (s. fig. 6.15) • The atypical haeman- gioma is indistinctly circumscribed and is usually larger than 4 cm. The inhomogeneous hyperechoic internal structure is caused by multiple acoustic interfaces (cavernous cavities of blood, bleedings, thrombosis, or- ganized tissues, etc.). Large haemangiomas may exhibit the “chameleon phenomenon”: when the patient assumes a different position, they change their echogenicity. (100)

Fig. 6.15: Typical haemangioma: hyperechoic density like a white tumour (“snowball”) near to a hepatic vein ( ⫽ red-coloured) (s.

figs. 8.5, 8.9; 36.6, 36.7)

Nodular regenerative hyperplasia (see chapter 36.4.3) is also deemed to be a benign hyperechoic structural de- ficiency. These multiple foci resemble grapes in their ap- pearance and are found in the direct vicinity of vessels, possibly causing portal hypertension. Foci are 1 ⫺3 cm in size and may easily be confused with metastases.

Echinococcus alveolaris (see chapter 25.2.3.2) is a solid and irregularly shaped focus. Inside, small cysts can be differentiated. It exhibits infiltrating growth and can imitate a malignant process.

Hepatocellular carcinoma (see chapter 37.3) is five times more frequent in men than in women. Hypoechoic as well as hyperechoic structural defects are observed, with a variety of different sound qualities present in a single tumour. The carcinoma may appear as a solitary or

multiple phenomenon, at times also infiltrating dif- fusely. HCC can be differentiated from regeneration nodes by enhanced early-arterial perfusion of the con- trast medium. However, there is a “chaotic” vascular pattern (as opposed to that witnessed in FNH). The rate of detection is 35 ⫺40% (<1 cm diameter of the tumour) and approx. 90% (> 1 ⫺2 cm in size). When applying endoscopic or intraoperative ultrasonography (s. p.

138), the respective values are 86% and 98%! Most fre- quently, hepatocellular carcinoma develops in cases of liver cirrhosis (ca. 75%) or haemochromatosis (ca. 20%), which further complicates the diagnostic process with ultrasonography. (8, 22, 59, 76, 99, 100) (s. fig. 6.16)

Fig. 6.16: Hepatocellular carcinoma (subcostal section) with ex- tended hypoechoic carcinoma (small arrows). Infiltration of the portal vein (intrahepatic) (large arrow) as a pathognomonic sign of liver cell carcinoma ( 씮)

Liver metastases (see chapter 37.10) are depicted with extreme variability under ultrasonography. They mostly occur as multiple and less frequently as solitary tu- mours. The spread of multiple small metastases is the cause of a diffuse heterogeneous echoic pattern in large parts of the liver without any distinct focal findings. The echoic pattern of liver metastases varies between hypo- echoic and hyperechoic ⫺ even within a single metastasis.

This spectrum of variation depends largely on the vas- cularity and/or respective development of the tumour.

Occult metastases can be detected by measuring the

Doppler perfusion index (DPI). (49) Occasionally, there

are separate hypoechoic and hyperechoic structural de-

ficiencies with and without a halonated (hypoechoic)

margin, possibly due to the varying age of the tumour

with its intermittent metastatic spread. In some cases,

signs of infiltration are found in the adjacent liver tissue,

biliary tract and vessels. The hypoechoic fringe embrac- ing a more hyperechoic centre (the so-called “halo”) is termed “target phenomenon”. By contrast, the “bull’s- eye phenomenon” displays a hyperechoic fringe and a hypoechoic or echofree (i. e. liquid) centre due to central colliquation of a metastasis. (s. fig. 6.17) Calcified me- tastases (particularly of a mucinous colorectal carci- noma) are hyperechoic with sound extinction. (s. figs.

37.24, 37.25) Hyperechoic metastases are often found in rectal or colon carcinoma, while hypoechoic metastases are more often identified in mammary, pancreatic and bronchial carcinoma. • There is no relation between the sonographic criteria of a metastasis and the localization or cell type of the primary tumour.

Fig. 6.17: Liver metastasis (subcostal section): hypoechoic fringe (arrow); wide, more hyperechoic margin and hypoechoic central colliquation (“bull’s-eye”). Structural inhomogeneity of the remaining hepatic segments

Humps on the surface of the liver, displacement or com- pression of intrahepatic vessels as well as bile ducts with cholestasis are all known to be important signs of a tu- mour. Central necrosis, bleeding, indistinct outlines and the formation of branches in the vicinity are signs of a high degree of malignancy or rapid tumourous growth.

• The precision of detection regarding metastases by means of ultrasonography ranges from 44% to more than 90%, with sensitivity varying from 40 ⫺80% and specificity from 62 ⫺100%. Accuracy does not only de- pend on the size and location of metastases, but also on the type of the primary tumour. Metastases of 0.5 cm upwards are detectable by ultrasonography. Tumourous foci in the area of the left lobe of liver (as well as subphrenic or right-sided foci) are easily overlooked.

Differential diagnosis between haemangioma or colo- rectal metastasis may be difficult. (4, 15, 56, 99, 100)

This synopsis of circumscribed structural defects clearly shows that solid focal alterations may be hypoechoic or hyperechoic as well as benign or ma- lignant. • The value of ultrasonography is the actual detection and not the definite specific diagnosis of such focal findings!

The strategy generally applied in the diagnostic clari- fication in the case of suspected “liver tumour” is shown later in a flow diagram. (s. p. 196) (s. fig. 9.4)

8 Examination of the spleen

The spleen is hidden underneath the left costal arch.

Examination of the normal spleen (4 ⫻7⫻11 cm) using ultrasonography (from the left side, through the inter- costal window) is therefore quite difficult. The hump of the spleen may be covered by aeriferous lung segments in the phrenicocostal sinus. The weight of the spleen de- pends on its respective blood content as well as on the anatomy and gender of the individual; it varies from 120 to 200 g. The pole of the spleen has a length of up to 11 cm, and the spleen has a diameter of up to 5 cm. The echogenicity is fine and homogeneous; the pattern can be either looser or more dense than that of the liver. The nor- mal spleen may show indentations or even flaps. Variable sectioning facilitates the determination of the size of the spleen in splenomegaly. (s. figs. 6.18; 11.1)

Fig. 6.18: Splenomegaly (S) in chronic myeloplastic leukaemia (length ⫽ 21 cm, depth ⫽ 7.4 cm)

The most useful parameter for the diagnosis of spleno-

megaly using sonography is a splenic width of > 5 cm

(80% accuracy) or > 7 cm (100% sensitivity). The length

of the spleen in splenomegaly (> 11 cm) is used for

diagnostic orientation. In long-standing splenomegaly,

the echo structure is coarsened. In liver cirrhosis, the spleen is as a rule significantly enlarged. However, the absence of splenomegaly in individual cases excludes neither portal hypertension nor cirrhosis, since the release of pressure may have been effected via collateral vessels.

Patency of the splenic vein or the portal vein excludes thrombosis as a cause of splenomegaly. (see chapter 11)

9 Special sonographic techniques

9.1 Sonography-guided puncture

Liver changes of uncertain aetiology, whether of a diffuse or local nature, require histological clarification. This is particularly important if therapeutic consequences or prognostic statements are to be derived. For this purpose, the following possibilities exist: (1.) fine-needle biopsy with cytological or microbiological examination (3, 10, 24, 25, 82) or (2.) thick-needle biopsy with histological or histochemical evaluation. (8, 65, 102) For diagnostic and especially for therapeutic purposes, there is also (3.) punc- ture of lesions (cysts, abscesses) with the possible place- ment of a drainage and the application of antibiotic lavage as well as the injection of cytostatics or alcohol into the tumour region, or the injection of substances to promote the sclerosing of cysts. (s. tab. 7.3)

The external diameter of the customary fine needle is

< 1 mm. A Chiba needle is generally used, although this is only suitable for obtaining cytological material. Diag- nostic assessment of the bioptic material requires special experience of cytology. • Using the incisive biopsy can- nula (diameter ca. 1 mm), a narrow biopsy cylinder can be extracted, which is processed and assessed in the usual histological manner. As opposed to Chiba cyto- logy, this bioptic method therefore allows a more reli- able assessment of the benignancy or malignancy of a liver tumour.

Method

There are three ways of proceeding:

(1.) Free puncture: Puncture or biopsy needle and sound trans- ducer are not connected mechanically.

(2.) Linked puncture: Biopsy is performed using needle forceps attached to the side of a sound transducer. The passage of the needle and the targeted focus are shown on a monitor using a sight line. The puncture path is at an angle (adjustable as re- quired) to the propagation of sound, i. e. the sound follows a different path from the biopsy needle.

(3.) Linked puncture with central perforation of the linear sound transducer. Needle and sound follow the same route. The visi- bility of the needle is not as good as when the biopsy needle is introduced into the sound field from the side.

For the assessment of focal hepatic alterations, it is pref- erable to use needles with which one can obtain cyto- logical as well as histological material in one session, namely with the help of fine-needle technology (e. g.

Trucut bioptic gun, 21 G). A further development is to

be seen in the self-activating vacuum needle, which, as a fine needle, ensures a high degree of accuracy in ob- taining tissue for histological examination. Use of the Chiba needle, which was widespread in the past, now tends to be less common.

Contraindications include coagulation disorders, superfi- cial hyperechoic foci (e. g. haemangioma), vascular an- eurysm, portal hypertension, obstructive jaundice and hydatid cysts. Care should also be taken not to damage any unknown structures along the path of the biopsy needle.

Possible complications are (1.) bleeding, (2.) spread of tumour cells into the puncture channel (0.05%), (3.) mismanaged biopsy as well as possible injury or perfor- ation of other organs, (4.) spread of infectious materials from abscesses, and (5.) bile peritonitis. The frequency of these complications is approximately 0.5% and the lethality approximately 0.08%. Hepatic haematoma re- sulting from puncture can be detected by ultrasono- graphy provided it is large enough and subsequently monitored in the follow-up. (48, 85) Occurrence of pain in the region of the liver after puncturing is a strong indication for ultrasonography.

䉴 Percutaneous “blind” biopsy, formerly used in Men- ghini’s technique, is today considered obsolete; it has been replaced by sonography-assisted liver biopsy. As a consequence, the safety of biopsy has been significantly improved. (s. p. 149)

9.2 Limitations and complementary procedures The results presented so far in the literature concerning the sensitivity and specificity of ultrasonography un- doubtedly reflect the success of the respective investi- gators following many years of experience. The accuracy of such results (given as a percentage) should, however, al- ways be seen in relative terms, and any general statements made in this connection should be qualified to obtain a more realistic mean value.

䉴 Diagnostic expectations regarding ultrasonography in liver diseases are quite frequently exaggerated. Hence the critical interpretation of results is all too often neg- lected, and false-positive or false-negative conclusions are drawn all too hastily. Even when there is the slightest doubt regarding the interpretation of findings, the examination should be carried out again: repeated in- vestigations will improve the accuracy.

䉴 Under optimal conditions, focal lesions of a diameter of 5 mm upwards can be identified. Liver metastases at varying stages can only be detected in 50 ⫺60% of cases.

• If the assessment of focal hepatic changes is not pos-

sible with ultrasonography ⫺ even after colour-encoded

duplex sonography ⫺ the indication for computer tomo-

graphy (CT) is given. Indeed, the new spiral CT genera-

tion with i.v. injection of contrast medium provides ex-

cellent vascular representation. The differentiation of fo- cal and malignant as well as vascular alterations of the liver is thus achieved in a way previously not thought possible. (s. p. 173) • Magnetic resonance imaging (MRI) has gained increasing importance since the introduction of T

contrast medium (iron-[II, III]-oxides) as well as of MRI angiography; this imaging technique needs to be redefined from the hepatological point of view. (s. p.

177) • For the differential diagnosis of focal nodular hyperplasia versus adenoma, a scintiscan is indicated. (s.

p. 194) • If diagnostic explanation has not been provided by means of these complementary examinations, the in- dication for laparoscopy with targeted biopsy (possibly with forceps biopsy) has to be considered. (s. p. 150) 9.3 Colour-encoded Doppler sonography

When reflected by mobile particles, sound waves undergo a shift of frequency proportional to the velocity of these particles ( ⫽ Doppler effect).

1. Signs of liver cirrhosis

2. Diagnosis of portal hypertension (with detection of hepatofugal flow as well as the slowing down of blood flow in the portal vein)

3. Confirmation of collateral circulation (e. g. um- bilical vein, veins of the gall-bladder wall) or flow reversal with centrifugal refluxes

4. Unclear splenomegaly

5. Suspected partial or complete thromboses of the splenic-portal vascular bed

6. Budd-Chiari syndrome or veno-occlusive disease 7. Vascular system in liver tumours, tumour com-

pression

8. Diagnosis of focal hepatic lesions (low or high number of blood vessels)

9. Assessment of a portosystemic shunt 10. Gastrointestinal bleeding of unknown cause 11. Internal or external head of Medusa

12. Assessment of the portal vein system before and after liver transplantation

13. Variations or malformations in the portal vein system (or visceral arteries)

14. Suspected Cruveilhier-von Baumgarten syndrome 15. Suspected cavernous transformation (formation of multiple venous collaterals close to the hepa- tic porta)

16. Therapeutic assessment of the vascularization of liver tumours after surgical interventions, chemo- embolization or parenteral chemotherapy

Tab. 6.5: Indications for colour-encoded Doppler sonography

The development of Duplex sonography is based upon the combination of conventional 2D grey-scale sonography with the 1D Doppler effect. This facilitates the one-

dimensional determination of the direction, velocity and volume of flow. (18, 44, 66, 70, 88, 90, 103, 104, 106) • The intro- duction of colour-encoded Doppler sonography (CEDS) in 1987 can be attributed to a combination of 2D Doppler sonography with real-time grey-scale sonography. Both the volume and the direction of the blood flow are repre- sented semiquantitatively by way of colour encoding (with different shades of brightness). The method is de- pendent upon the relationship between the vascular course and the transducer (red ⫽ blood flow towards the probe, blue ⫽ blood flow away from the probe). ^{(5, 7, 28,}

29, 62, 63, 84) (s. fig. 6.19) • The portal vein blood flow is centripetal and slightly pulsatile; the mean velocity of flow was measured as 15.2 ⫾2.6 cm/sec, representing a flow volume of 693 ⫾235 ml/min. (s. fig. 6.20)

Fig. 6.19: Portal vein (PV) and division into the right (RP) and left (LP) portal vein trunk (subcostal section)

Fig. 6.20: Portal vein flow depicted by means of spectral analysis (velocity of flow ⫽ 11cm/sec); hepatopetal flow is shown in red

The volume only increases slightly during expiration, but

significantly after eating (880 ⫾269 ml/min). Portal hy-

pertension can be most reliably determined by measuring

the flow volume prior to and following a test meal. (53) •

In the case of suspected pathological findings in the area

of the splanchnic vessels, colour-encoded Doppler sonog-

raphy is indicated. This investigation method will incorp-

orate the portal vein system, hepatic veins, splenic vein,

superior mesenteric vein and coeliac artery as well as the arterial vascular system. (s. tab. 6.5) (s. figs. 6.14; 14.13;

35.12)

The use of signal amplification with the help of ultra- sound contrast medium provides a considerably improved depiction of portohepatic vessels as well as the perfusion of hepatic tumours. • Microbubbles are used as a contrast medium. The echo signal is amplified by increased back- scattering of ultrasound waves caused by the microbub- bles. This technique is based on observations made by the cardiologist C.R. J oyner, jr. (1966) . Up to now, the medium of choice has been a galactose solution (99.9%) sus- pended in palmitic acid, the latter being used to stabilize the microbubbles. It is administered as a bolus injection or intravenously. (17, 100)

Particularly colour-power sonography makes it easier to identify small and deeply-set vessels as well as poorly vascularized focal tumours reliably. The procedure does not replace CEDS, but serves to complement it. (14, 86)

The following sonographic procedures can be applied, depending on the specific tasks and objectives (s. tab.

6.6):

1. Conventional sonography 2. Doppler sonography

⫺ Duplex sonography

(combination of 2D conventional sonography with 1D Doppler sonography)

⫺ colour-encoded sonography

(combination of real-time grey-scale sonography with 2D colour-encoded real-time Doppler sonography)

⫺ Doppler-power sonography

⫺ contrast medium signal-enhanced Doppler sonography 3. 3D Doppler sonography

Tab. 6.6: Sonographic examination procedures

9.4 Endoscopic sonography

Endoscopic ultrasonography promises greater accuracy as it involves placing a specialized sound transducer dir- ectly on the liver by means of laparoscopy or surgery.

Laparoscopic sonography: As early as 1975, ^{D. L} ook et al. showed the gall bladder, using the A-scan mode of ultrasonography via the laparoscope. In 1980 Y. F uru- kawa et al. described laparoscopic ultrasonography by employing the B-scan mode. Initially, optic lenses were employed, before being substituted later by linear probes. In 1989 F. F ornari et al. reported on the laparo- scopic application of a rotating sectorial scanner. (108)

Here a linear sound transducer (7 MHz) attached to a rigid shaft by means of a flexible segment produced optimum results. This routine has the advantage of a wider scope of motility and improved orientation inside the gas-filled abdomen ⫺ without significant additional inconvenience to the patient. • Laparoscopic ultrasono- graphy not only detects intrahepatic foci ⫺ with the pos-

sibility of targeted biopsy ⫺ but it also allows better matching of a focal lesion with the respective segment as well as more reliable preoperative tumour staging. • This procedure is also of advantage for the intraoper- ative application of ultrasonography in the detailed search for intrahepatic foci as well as in the intraoper- ative depiction of segmental borders. (107 ⫺118)

Intraductal ultrasonography: IDUS involving flexible miniature ultrasound probes with a diameter of approxi- mately 1.5 ⫺2 mm (7.5⫺20 MHz) will expand the diag- nostic spectrum concerning pathological processes in the larger bile ducts. The choledochus can usually be examined up to the hepatic duct. The identification rate of malignant or benign findings is high (accuracy 92%, sensitivity 90%, and specificity 93%. (119 ⫺123)

References:

1. Abbitt, P.L.: Ultrasonography-update on liver technique. Radiol. Clin.

N. Amer. 1998; 36: 299 ⫺307

2. Aube´, C., Oberti, F., Korali, N., Namour, M.-A., Loisel, D., Tanguy, J.- Y., Valesia, E., Pilette, C., Rousselet, M.C., Bedossa, P., Rifflet, H., Maiga, M.Y., Penneau-Fontbonne, D., Caron, C., Cales, P.: Ultrasono- graphic diagnosis of hepatic fibrosis or cirrhosis. J. Hepatol. 1999; 30:

472 ⫺478

3. Badea, R., Badea, Ch., Galtar, N., Cracium, C., Marin, M., Buchen, M.: Die Bedeutung der aspirativen Feinnadelbiopsie in der Präzisierung ultraschalldiagnostizierter Lebertumoren. Untersuchungen an 165 Patienten. Z. Klin. Med. 1991; 46: 1489 ⫺1493

4. Becker, D., Strobel, D., Hahn, E.G.: Tissue harmonic imaging and con- trast harmonic imaging. Improving the diagnosis of liver metastasis.

Internist 2000; 41: 17 ⫺23

5. Blank, W., Braun, B.: Gewebsdiagnostik durch Dopplersonographie.

Bildgebung 1995; 62 (Suppl. 1): 31 ⫺35

6. Bleck, J.S., Gebel, M., Manns, M.P.: Quantitative sonography. Clinical role and perspectives. Internist 2000; 41: 10 ⫺16

7. Bolondi, L., Gaiani, S., Li Bassi, S., Zironi, G., Bonino, F., Brunetto, M., Barabara, L.: Diagnosis of Budd-Chiari syndrome by pulsed Doppler ultrasound. Gastroenterology 1991; 100: 1324 ⫺1331

8. Bolondi, L., Gaiani, S., Benzi, G., Rigamonti, A., Fusconi, F., Barbara, L.: Ultrasonography and guided biopsy in the diagnosis of hepatocellu- lar carcinoma. Ital. J. Gastroenterol. 1992; 24: 46 ⫺49

9. Bolondi, L., Zironi, G., Gaiani, S., Li Bassi, S., Benzi, G., Barbara, L.:

Caliber of splenic and hepatic arteries and spleen size in cirrhosis of different etiology. Liver 1991; 11: 198 ⫺205

10. Buscarini, L., Fornari, F., Bolondi, L., Colombo, P., Livraghi, T., Mag- nolfi, F., Rapaccini, G.L., Salmi, A.: Ultrasound-guided fine-needle bi- opsy of focal liver lesions: techniques, diagnostic accuracy and compli- cations. A retrospective study on 2091 biopsies. J. Hepatol. 1990; 11:

344 ⫺348

11. Caremani, M., Vincenti, A., Benci, A., Sassoli, S., Tacconi, D.: Eco- graphic epidemiology of non-parasitic hepatic cysts. J. Clin. Ultrasound 1993; 21: 115 ⫺118

12. Cassani, F., Zoli, M., Battoni, L., Cordiani, M.R., Brunori, A., Bianchi, F.B., Pisi, E.: Prevalence and significance of abdominal lymphadenopa- thy in patients with chronic liver disease: an ultrasound study. J. Clin.

Gastroenterol. 1990; 12: 42 ⫺46

13. Caturelli, E., Squillante, M.M., Andriulli, A., Cedrone, A., Cellerino, C., Pompili, M., Manoja, E.R., Rapaccini, G.L.: Hypoechoic lesions in the

“bright liver”: a reliable indicator of fatty change. A prospective study.

J. Gastroenterol. Hepatol. 1992; 7: 469 ⫺472

14. Cedrone, A., Pompili, M., Sallustio, G., Lorenzelli, G.P., Gasbarrini, G., Rapaccini, G.L.: Comparison between color power Doppler ultrasound with echo-enhancer and spiral computed tomography in the evaluation of hepatocellular carcinoma vascularization before and after ablation procedures. Amer. J. Gastroenterol. 2001; 96: 1854 ⫺1859

15. Charnley, R.M., Morris, D.L., Dennison, A.R., Amar, S.S., Hardcastle, J.D.: Detection of colorectal liver metastases using intraoperative ultra- sonography. Brit. J. Surg. 1991; 78: 45 ⫺48

16. Colli, A., Fraquelli, M., Andreoletti, M., Marino, B., Zuccoli, E., Conte, D.: Severe liver fibrosis or cirrhosis: accuracy of US for detection.

Analysis of 300 cases. Radiology 2003; 227: 89 ⫺94

17. Cosgrove, D.O., Blomley, M.J.K., Jayaram, V., Nihoyannopoulos, P.:

Echo-enhancing (contrast) agents. Ultrasound Quart. 1998; 14: 66 ⫺75 18. D’Alimonte, P., Cioni, G., Cristani, A., Ferrari, A., Ventura, E., Romag-

noli, R.: Duplex-Doppler ultrasonography in the assessment of portal hypertension. Utility of the measurement of maximum portal flow vel- ocity. Radiology 1993; 17: 126 ⫺129

19. Delorme, S., Kaick, van, G.: Sonographie fokaler Leberveränderungen.

Praktische Hinweise für die Differentialdiagnostik. Radiologe 1992;

32: 198 ⫺206