Abstract:

Objective To explore the value of energy spectrum CT imaging in differential diagnosis of pulmonary tuberculosis nodules and silicosis nodules. Methods Thirty-five cases of pulmonary tuberculosis (127 nodules; PTB group) and 30 cases of silicosis (180 nodules; silicosis group) diagnosed by energy spectrum CT scan from February to December 2018 in the Fifth People’s Hospital of Suzhou were enrolled in this study. The concentration of SiO₂ (H₂O) based substances, effective atomic number (EAN) and slope of curve in 40-70 keV range were obtained by energy spectrum analysis. Results The median (quartile) concentration of SiO₂ (H₂O) based substances in PTB group and in silicosis group were 170 (101,241) mg/cm ³ and 353 (260,454) mg/cm ³ respectively, with statistically significant difference (Z=-24.95, P=0.000); the mean EAN in PTB group and silicosis group were 9.29±0.83 and 10.68±0.98 respectively, with statistically significant difference (t=29.96, P=0.000); the average slope of the energy spectrum curve in PTB group and silicosis group were 5.95±2.19 and 10.80±2.48, the difference between those two groups was statistically significant (t=-17.96, P=0.000). When the threshold of the concentration of SiO₂ (H₂O) based substances was 275 mg/cm ³, the area under the ROC curve was the largest (0.95), and the sensitivity and specificity were 87.40% and 90.56% respectively. Conclusion The concentration of SiO₂(H₂O)based substances, EAN and slope of energy spectrum curve in energy spectrum CT are helpful for differential diagnosis of PTB nodules and silicosis nodules, among them the concentration of SiO₂ (H₂O) based substances is of higher value especially.

Key words: Tuberculosis, pulmonary, Silicosis, Multiple pulmonary nodules, Tomography, X-ray compute, Diagnosis, differential, Energy spectrum CT