Spiral CT optimization for measurement of bronchial lumen diameter using an experimental model.

The aim of the study was to determine optimal parameters for demonstrating sublobar bronchi on spiral CT. Measurements were obtained from five parallel polyethylene tubes embedded in foam matrix with similar radiographic characteristics to segmental and subsegmental bronchi and to lung parenchyma, respectively. Collimation widths of 1.5, 2, 3 and 4 mm were used, with a pitch of 1 or 1.5 and a reconstruction interval of 1 mm or 2 mm. Various slice planes were used. Images acquired orthogonally were viewed normally. Images acquired in planes oblique or parallel to the long axes of the tubes were reformatted into a plane orthogonal to the long axes of the tubes to be comparable with the directly acquired orthogonal images. Tube diameters were measured at lung window settings (L, -400; W, 1300) and compared with known true inner and outer tube diameters. Measurements from images acquired orthogonal to the tube long axes were accurate regardless of slice thickness. Images acquired obliquely or parallel only produced accurate measurements at the lowest slice thickness (1.5 mm). Pitch and reconstruction interval had no effect on measurement error in any scan plane. It is concluded that a slice thickness of 1.5 mm or less, with a pitch of 1.5, should be used when acquiring images at angles other than orthogonal to the long axes of experimental tubes equivalent to the segmental and subsegmental bronchi. It is suggested that similar parameters should be used in vivo and that the examination should be targeted to the area of the bronchial tree in question to reduce patient dose and length of breath-hold.