Inter-scan repeatability of CT-based lung densitometry in the surveillance of emphysema in a lung cancer screening setting

Purpose: To investigate the inter-scan repeatability of CT-based lung densitometry protocols in the surveillance of emphysema in a lung cancer screening setting.

Materials and methods: Fifty-two healthy subjects who underwent low-dose chest CT and subsequent follow-up scan within a 16 month interval were retrospectively evaluated. Inter-scan repeatabilities were assessed for 9 different CT-based lung densitometry protocols with standard deviation (SD) of inter-scan differences. Susceptibility to inspiratory level was additionally assessed for each protocol, and volume adjustment (VA) was applied in order to evaluate the potential improvement of repeatability after compensating the influence of inspiratory level.

Results: A wide variation of inter-scan repeatability was observed among the evaluated protocols showing a difference of up to a factor of 9. Susceptibility of inspiratory level was found to be highly associated with the inter-scan repeatability of densitometric protocols. The application of VA could substantially reduce the influence of inspiratory level for all protocols, which results in an improvement of repeatability up to 51%. The order of repeatability among the protocols remained unchanged after VA. The resulting two best protocols in terms of inter-scan repeatability were RA970 and Perc1 which showed SD of 0.8% and 5.5 HU, respectively.

Conclusions: Lung densitometry protocols produce different levels of repeatability for an asymptomatic population, each being influenced by inspiratory level to a different degree. For surveillance of emphysema in a lung cancer screening setting, RA970 and Perc1 may be the most suitable protocols, in which the application of VA needs to be included as a critical part.