In patients with a suspicion of HCC (such as rising alpha-fetoprotein and protein induced by vitamin K absence or antagonist-II levels), one of the best methods of diagnosis involves a CT scan of the liver using intravenous (IV) contrast agent and a four-phase scanning (unenhanced, arterial phase, portal-venous phase and equilibrium phase) to increase the ability of the diagnostic radiologists to detect small or subtle HCCs.

CT is now widely available and represents an important method for the diagnostics of liver lesions and planning of therapy. Dynamic MDCT with an intravenous bolus injection of contrast medium (CM) is essential to diagnose liver tumor [12, 13]. Diagnostic accuracy of the liver tumors has improved, because MDCT has higher spatial and temporal resolutions by the increase of both gantry rotation speed and the number of detector rows. But rapid scan speed with a MDCT scanner sometimes increases the difficulty to image hypervascular HCC during the arterial phase after CM injection. The optimal technique for IV injection of CM is very important to diagnose HCC. Therefore, we should know the optimal scanning protocol to start in the arterial phase after IV injection of CM.

The arterial phase imaging is useful to detect hypervascular HCC. On arterial phase of liver dynamic CT, there is a good correlation between tumor vascularity on CT and angiography. The portal-venous and equilibrium phases imaging are useful for the differential diagnosis of HCC, because the washout of CM from the tumor in these phases is a typical finding of hypervascular HCC (Fig. 2). On portal-venous and equilibrium phases, a minimum enhancement of 50 Hounsfield units (HU) should be achieved as adequate liver enhancement to obtain high conspicuity of low-attenuated hepatic lesions at the portal-venous and equilibrium phases [14–16].

However, arterial parenchyma enhancement due to arterioportal venous (AP) shunt may become a false-positive lesion (mimic HCC) on dynamic MDCT that evaluates hemodynamics of liver tumor, and it may sometimes reduce specificity. The finding of washout pattern is useful to distinguish non-tumorous AP shunt from hypervascular HCC, because AP shunt substantially showed no washout of the liver at the portal-venous and equilibrium phases. In other words, HCC shows both focal arterial enhancement and the corresponding washout in the portal and equilibrium phases, but AP shunt shows arterial enhancement alone and wedge-shaped appearance. However, the differentiation between HCC and AP shunt is sometimes difficult because some HCCs (especially small HCCs) do not demonstrate washout.

Scanning through the upper abdomen can be performed in less than 2–3 s by using state-of-the-art MDCT scanners with more than 64 channels, even though a spatial resolution of 0.6 mm is employed for both longitudinal and short axis of the body (transverse slice thickness), so-called isotropic voxel volume imaging. Three-dimensional (3D) images can be reconstructed from the isotropic voxel imaging data by using multiplanar reconstruction (MPR), volume rendering (VR) (Fig. 3), and maximum intensity projection techniques. These 3D images are useful for preoperative anatomical evaluation for surgeons and preoperative explanations for patients [17].

The diagnostic algorithm for suspected HCC by the AASLD guideline states that an imaging diagnosis of HCC should be made if a lesion larger than 1 cm shows a typical vascular enhancement pattern on a single dynamic imaging study (Fig. 4) [36]. For hepatic nodules smaller than 1 cm, repeated sonography at 3 months is recommended to investigate the nodular size. Four-phase MDCT and dynamic contrast-enhanced MRI are recommended in this algorithm as imaging modalities to detect arterial hypervascularity and washout at portal-venous and/or equilibrium phase for hepatic nodules larger than 1 cm. Therefore, confirmation of arterial enhancement and the corresponding washout in the portal and equilibrium phases on dynamic CT is important to diagnose HCC after sonographic detection of hepatic nodule on the screening program. A biopsy is not needed to confirm the diagnosis of HCC if certain imaging criteria are met. When typical findings of HCC, such as arterial hypervascularity and washout at portal-venous and/or equilibrium phase on liver dynamic CT, could not be obtained, other contrast-enhanced imagings or biopsy of hepatic nodules are recommended.

The CT appearance of HCC depends on tumor size and histologic grade. Small HCC is defined as nodules smaller than 2–3 cm in size, whereas lesions larger than 3 cm in size are often heterogeneous with a mosaic or mixed pattern arising from a necrosis, fibrosis, fatty degeneration, and hemorrhage [50].

A characteristic enhancement pattern consisting of hyperenhancement in the arterial phase and washout in the portal-venous or equilibrium phase is associated with high specificity (nearly 100 %) for the diagnosis of HCC. But dynamic CT with extracellular CM has some diagnostic limitations related to tumor size. Hepatic nodules larger than 1 cm in diameter show high accuracy of hypervascular HCCs in a cirrhotic liver, although it is difficult to diagnose subcentimeter HCCs [36]. Therefore, the confirmation by biopsy is sometimes required for questionable hepatic nodules to determine the diagnosis. Thus, the key characteristics of HCC on CT are hypervascularity in the arterial phase scans, washout in the portal and equilibrium phase studies, a pseudocapsule, and a mosaic pattern. Both calcifications and intralesional fat may be appreciated.

There is insufficient diagnostic performance for the early detection of both small and early HCCs even with state-of-the-art CT. Gadoxetic acid disodium (gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid, Gd-EOB-DTPA; PRIMOVIST^Ⓡ Bayer Schering Pharma AG, Berlin, Germany) is a hepatocyte-specific MR contrast agent. Gd-EOB-DTPA-enhanced MRI is a promising examination to diagnose small HCC [51] and depicts early HCCs, such as well-differentiated HCC, in the early stage of hepatocarcinogenesis, because Gd-EOB-DTPA has the properties of an extracellular gadolinium chelate, as well as being a hepatocyte-targeting agent. Gd-EOB-DTPA-enhanced MRI enables hepatocyte-phase imaging to start 10–20 min postinjection. Therefore, additional dynamic imaging modality, such as Gd-EOB-DTPA-enhanced MRI, is recommended, when liver dynamic CT shows indeterminable result in the diagnostic process of HCCs [36].