Volumetric measurement of pulmonary nodules at low-dose chest CT: effect of reconstruction setting on measurement variability.

<bold>Objective: </bold>To assess volumetric measurement variability in pulmonary nodules detected at low-dose chest CT with three reconstruction settings.<bold>Methods: </bold>The volume of 200 solid pulmonary nodules was measured three times using commercially available semi-automated software of low-dose chest CT data-sets reconstructed with 1 mm section thickness and a soft kernel (A), 2 mm and a soft kernel (B), and 2 mm and a sharp kernel (C), respectively. Repeatability coefficients of the three measurements within each setting were calculated by the Bland and Altman method. A three-level model was applied to test the impact of reconstruction setting on the measured volume.<bold>Results: </bold>The repeatability coefficients were 8.9, 22.5 and 37.5% for settings A, B and C. Three-level analysis showed that settings A and C yielded a 1.29 times higher estimate of nodule volume compared with setting B (P = 0.03). The significant interaction among setting, nodule location and morphology demonstrated that the effect of the reconstruction setting was different for different types of nodules. Low-dose CT reconstructed with 1 mm section thickness and a soft kernel provided the most repeatable volume measurement.<bold>Conclusion: </bold>A wide, nodule-type-dependent range of agreement between volume measurements with different reconstruction settings suggests strict consistency is required for serial CT studies. [ABSTRACT FROM AUTHOR]