Spectral CT in patients with small HCC: investigation of image quality and diagnostic accuracy.

OBJECTIVE

To assess image quality and diagnostic accuracy of monochromatic imaging from spectral CT in patients with small HCC (≤3 cm).

METHODS

Twenty-seven patients with 31 HCC underwent spectral CT to generate conventional 140-kVp polychromatic images (group A) and monochromatic images with energy levels from 40 to 140 keV (group B) during the late arterial phase (LAP) and the portal venous phase (PVP). Two-sample t tests compared the tumour-to-liver contrast-to-noise ratio (CNR) and mean image noise. Lesion detection for LAP, reader confidence and readers' subjective evaluations of image quality were recorded.

RESULTS

Highest CNRs in group B were distributed at 40, 50 and 70 keV. Higher CNR values and lesion conspicuity scores (LCS) were obtained in group B than in group A (CNR 3.36 ± 2.07 vs. 1.47 ± 0.89 in LAP; 2.29 ± 2.26 vs. 1.58 ± 1.75 in PVP; LCS 2.82, 2.84, 2.63 and 2.53 at 40-70 keV, respectively, vs. 1.95) (P < 0.001). Lowest image noise for group B was at 70 keV, resulting in higher image quality than that in group A (4.70 vs. 4.07; P < 0.001).

CONCLUSION

Monochromatic energy levels of 40-70 keV can increase detectability in small HCC and this increase might not result in image quality degradation.

KEY POINTS

• Spectral computed tomography may help the detection of small hepatocellular carcinoma. • Monochromatic energy levels of 40-70 keV increase the sensitivity for detection. • Prospective study showed that monochromatic imaging provides greater diagnostic confidence. • Monochromatic energy level of 70 keV improves the overall image quality.