Pulmonary nodules: detection with thick-section spiral CT versus conventional CT.

Abstract

Spiral volumetric computed tomography (CT) with single breath-hold technique was compared with conventional sequential CT in 39 patients. The spiral CT protocol consisted of a 10 mm/sec table feed during a 24-second breath hold at 145 mA, with reconstruction of images at 10-mm intervals; one (n = 21) or two (n = 18) sequences were necessary to screen the complete lung. Conventional CT was performed with a 1-second scan time, 145 mA, and contiguous 10-mm-thick sections. In the 39 patients studied with each technique, no lung nodule was detected in three, two had a single nodule, and 29 had multiple nodules. Two patients with normal findings at chest radiography and three with a solitary pulmonary nodule at conventional CT had multiple nodules at spiral CT. Mean number of nodules per patient was significantly higher with spiral versus conventional CT (18 +/- 4.5 vs 12.6 +/- 3.2 [mean +/- standard error of the mean], P = .01) as were the number of nodules less than 5 mm in diameter per patient (12.7 +/- 3.7 vs 8.4 +/- 2.3, P < .05) and 5-10 mm in diameter (2.9 +/- 0.9 vs 2.4 +/- 0.8, P < .05). Respiratory motion artifacts were never observed with spiral CT although they were present on four conventional CT scans.