[Value of spectral CT-based quantitative analysis in differential diagnosis of liver cancer and liver abscess].

Objective: To investigate the value of spectral CT-based quantitative analysis in the differential diagnosis of liver cancer and liver abscess. Methods: A total of 70 patients with space-occupying lesions in the liver(45 with liver cancer and 25 with liver abscess)underwent spectral CT scans to obtain spectral images in the arterial phase and portal venous phase. The solid constituents of lesions and the iodine and water concentrations in necrotic or cystic parts of lesions, normal hepatic tissue, and abdominal aorta in the arterial phase and portal venous phase were measured, and the normalized iodine concentration(NIC)and lesion-to-normal hepatic tissue ratio(LNR)of iodine concentration were calculated. The two samples t-test and the receiver operating characteristic(ROC)curve analysis were performed for the quantitative indices above. Results: The patients with liver cancer had higher NIC and LNR in solid constituents in the arterial phase than those with liver abscess(NIC: 0.15±0.06 mg/ml vs 0.14±0.02 mg/ml, P > 0.05; LNR: 2.78±0.65 vs 1.45±0.88, P < 0.001). The patients with liver abscess had significantly higher NIC and LNR in solid constituents in the portal venous phase than those with liver cancer(NIC: 0.65±0.08 mg/ml vs 0.52±0.08 mg/ml, P≤0.001; LNR: 1.22±0.23 vs 0.95±0.15, P≤0.001). There were no significant differences in NIC in the arterial phase or NIC and LNR in the portal venous phase in necrotic or cystic parts of lesions between the patients with liver cancer and liver abscess(P > 0.05). The optimal quantitative value for the differential diagnosis of liver cancer and liver abscess was LNR in arterial phase, and the cut-off value of 1.53 had a sensitivity of 100% and a specificity of 92%. Conclusion: Quantitative iodine concentration analysis in spectral CT imaging has a certain value in the differential diagnosis of liver cancer and liver abscess and can improve the accuracy of diagnosis.