Efficacy of model-based iterative reconstruction in cystic fibrosis assessment using CT.

Abstract

AIM\nTo evaluate the efficacy of model-based iterative reconstruction (MBIR) constructed non-enhanced ultra-low dose (ULD) computed tomography (CT) of the chest to evaluate cystic fibrosis (CF) pathology.\n\n\nMATERIALS AND METHODS\nULD-CT was compared with chest X-ray and standard adaptive statistical iterative reconstructed (ASIR) non-enhanced low-dose CT (LD-CT). The effective radiation dose was calculated from the recorded dose-length product (DLP) values and compared between the two CT methods. Identification of pathology was compared between ULD-CT and chest X-ray. It was hypothesised that ULD-CT would be superior to chest X-ray in the identification of CF pathology at lower doses than LD-CT.\n\n\nRESULTS\nThe mean effective radiation dose of ULD-CT was 0.073 mSv, comparable to one chest X-ray, which was a 94% reduction compared to LD-CT. Compared to chest X-ray, ULD-CT detected on average, 2.3 more regions of bronchiectasis per study and better delineated varicose and cystic forms of bronchiectasis (p≤0.0001). ULD-CT identified four-times more mucous plugging than chest X-ray (p<0.000001) and twice the amount of consolidation (p=0.0002).\n\n\nCONCLUSION\nULD-CT is superior to chest X-ray in quantifying CF disease and achieves remarkable radiation doses significantly lower than LD-CT, comparable to one chest radiograph. The present results suggest that MBIR-constructed ULD-CT is an effective imaging technique for CF surveillance, with potential applications in other disease settings.