This study was designed to determine lung volumes using inspiratory and expiratory helical CT with two-dimensional (2D) and three-dimensional (3D) postprocessing and to compare the accuracy of those measurements with pulmonary function test results.

Seventy-two patients with suspected pulmonary disease underwent unenhanced helical CT (slice thickness, 8 mm; pitch, 2; increment, 8 mm) at deep inspiration and expiration. Lung volumes were determined using either a 2D approach (semiautomatic segmentation; thresholds, -1024 and -200 H) or a 3D technique (double-threshold seeded volumes of interest; thresholds, -1024 H [lower] and -900, -500, 400, -300, or -200 H [upper]). Pulmonary function tests were available for correlation in all cases.

Using inspiratory helical CT, we underestimated total lung capacity by 12%, which had a good correlation (r = .89) with static lung volumes. Volume revealed by expiratory helical CT was equivalent to intrathoracic gas volume, which also exhibited a good correlation (r = .88). However, using expiratory helical CT, we overestimated residual volume by 850 ml with a rather good correlation (r = .77). An emphysema index revealed moderate correlation with the relative forced expiratory volume in 1 sec (inspiration, r = -.66; expiration, r = -.54), whereas the expired volume showed a moderate correlation with the absolute forced expiratory volume in 1 sec (r = .65). The 2D approach showed lower absolute volumes than the 3D technique (mean, 3.6%; r = .99). In the 3D technique, lower upper thresholds led to reduced volumes (170 ml/100 H).

Inspiratory and expiratory helical CT show high correlation with static lung volumes. The 3D technique (-1024 to -200 H) is recommended for absolute estimation of lung volumes.