Three-Phase vs Four-Phase Liver CT: Protocol Selection
For hepatocellular carcinoma (HCC) detection and characterization, use a three-phase CT protocol (late arterial, portal venous, and delayed phases) as it provides equivalent diagnostic accuracy to four-phase imaging while reducing radiation exposure.
Standard Protocol: Three-Phase CT
The ACR Appropriateness Criteria explicitly recommend multiphase CT with IV contrast as the standard approach for liver lesion characterization 1, 2. The essential phases are:
- Late arterial phase (preferred over early arterial)
- Portal venous phase
- Delayed phase
Why Three Phases Are Sufficient
The diagnostic hallmark of HCC—arterial hyperenhancement with delayed washout—has 89% sensitivity and 96% specificity 1. This pattern is optimally captured with late arterial and delayed phases, making these phases critical for diagnosis.
Research directly comparing protocols confirms that adding an early arterial phase (creating a four-phase protocol) does not improve HCC detection. A study of 33 HCCs found no statistically significant difference in sensitivity (94% vs 93%), specificity (99% vs 98%), or area under the ROC curve (0.971 vs 0.976) between three-phase and four-phase protocols 3.
When to Consider Four-Phase CT
Add a precontrast (noncontrast) phase only in specific circumstances:
- Post-treatment surveillance: Patients who have undergone prior liver-directed therapy may benefit from a noncontrast series to assess treatment response 1
- Characterization of specific lesions: When distinguishing hemorrhage, calcification, or fat content is clinically necessary
Do not routinely add an early arterial phase, as it increases radiation dose without improving diagnostic accuracy 3.
Key Clinical Considerations
Radiation Dose Implications
The effective dose difference is substantial:
- Three-phase CT: 20.6 mSv
- Four-phase CT: 27.7 mSv 4
This represents a 34% increase in radiation exposure with four-phase imaging, without corresponding diagnostic benefit.
Protocol Optimization
Deep learning analysis confirms that removing the precontrast phase (not the portal venous phase) maintains diagnostic accuracy while reducing radiation 5. The portal venous phase is particularly critical—lesions showing hypoattenuation below 136 HU in this phase have 87% accuracy for predicting HCC 6.
Common Pitfall to Avoid
Do not use single-phase CT for liver lesion characterization—it does not allow adequate assessment of enhancement patterns and is explicitly inadequate per ACR guidelines 1. If a lesion is detected on single-phase imaging, repeat with multiphase protocol.
Practical Algorithm
- Standard liver imaging: Use three-phase protocol (late arterial, portal venous, delayed)
- Post-treatment follow-up: Add precontrast phase (four-phase total)
- Incidental finding on single-phase CT: Repeat with three-phase protocol
- Never add early arterial phase routinely—no diagnostic benefit
The three-phase protocol is the evidence-based standard that balances diagnostic accuracy with radiation safety 1, 2, 3.