To facilitate systematic calibration and validation of quantitative airway measurements on CT for COPD diagnosis, an
acrylic plastic phantom has been designed with an array of cylindrical tubes varying lumen diameter and wall thickness
in a systematic way, which can be manufactured by inexpensive 3D-printing. Accuracy and reproducibility of the 3Dprinting
have been confirmed by CT measurements. The multipliable, unobtrusive phantom can be scanned simultaneously
with the patient for each exam, allows scan-specific calibration, and can thus improve multicenter study
comparability between differing clinical imaging protocols.
Three different methods to correct the bronchial measurements for partial volume effect and scanner blur were tested.
The correction methods were variants of the physically motivated method suggested by Weinheimer et al. which
integrates the Hounsfield densities in a certain wall area, and derives the corrected values assuming a characteristic
constant Hounsfield density of the bronchial wall. The alternative methods compared here differ in the choice of the
integration boundaries. Analysis of CT scans showed high agreement and good noise robustness of all correction
methods on the one hand, but significant dependency on the choice of the CT reconstruction filter on the other hand,
which emphasizes the benefits of scan-specific calibration.