PURPOSE

To evaluate the accuracy of pulmonary nodule (PN) detection in overweight or obese adult patients using ultralow-dose computed tomography (ULDCT) with tin filtration at 100 kV and advanced model-based iterative reconstruction (ADMIRE).

METHODS

Eighty-one patients with body mass indices of ≥25 kg/m were enrolled. All patients underwent low-dose chest CT (LDCT), followed by ULDCT. Two radiologists experienced in LDCT established the standard of reference (SOR) for PNs. The number, type, size, and location of PNs were identified in the SOR. Effective dose, objective image quality (IQ), and subjective IQ based on two radiologists' scores were compared between ULDCT and LDCT. The detection performances of radiologists based on ULDCT were calculated according to the nodule analyses. Logistic regression was used to test for independent predictors of PN detection sensitivity.

RESULTS

Both the effective dose and objective IQ were lower for ULDCT than for LDCT (both p < 0.001). Both radiologists rated the subjective IQ of the overall IQ on ULDCT to be diagnostically sufficient. In total, 234 nodules (mean diameter, 3.4 ± 1.9 mm) were classified into 32 subsolid, 149 solid, and 53 calcified nodules according to the SOR. The overall sensitivity of ULDCT for nodule detection was 93.6%. Based on multivariate analyses, the nodule types (p = 0.015) and sizes (p = 0.013) were independent predictors of nodule detection.

CONCLUSIONS

Compared with LDCT, ULDCT with tin filtration at 100 kV and ADMIRE could significantly reduce the radiation dose in overweight or obese patients while maintaining good sensitivity for nodule detection.