B-mode ultrasonography (B-US), a standard method of surveillance of hepatocellular carcinoma (HCC), has a fair sensitivity of 63% in detecting early stage HCC. Sonazoid, a contrast agent for ultrasonography, has reported to have superior sensitivity in detecting focal liver lesion since it has ability of Kupffer phase imaging as well as vascular phase imaging. So our aim is to compare the detection rate of early stage HCC and false referral rate of HCC between B-mode US and Sonazoid-enhanced ultrasonography (S-US) within the same prospective data group. Our hypothesis is that S-US has superior detection rate of early stage HCC (5%) than that of B-mode US (3%).

Inclusion Criteria:

1. Patient 20 years old or older,
2. Patient at high risk of HCC and is supposed to take ultrasonography,
3. Patient who has no suspicious HCC on previous examinations or patient who undergoes the first surveillance,
4. Patient who has liver cirrhosis proven by one of the following methods-liver biopsy (METAVIR score 4), identification of esophageal or gastric varix by endoscopically or radiologically, surface nodularity on ultrasonography, CT or MRI, platelet count less than 100,000/mm3, serum albumin less than 3.5 g/dl, or prothrombin time higher than 1.3 INR,
5. Patient without contraindication for sonazoid,
6. Patient willing to sign the informed consent

Exclusion Criteria:

1. Patient on pregnancy or breast feeding,
2. Patient with allergy to egg,
3. Patient with left-to-right shunt, respiratory insufficiency, or severe pulmonary hypertension,
4. Patient with history of HCC,
5. Patient with history of malignancy other than HCC -