Terry Robinson

Sampling density for the quantitative evaluation of air trapping.
Goris ML, Robinson TE. Pediatr Radiol. 2009; 39 (3): 221-5
BACKGROUND: Concerns have been expressed recently about the radiation burden on patient populations, especially children, undergoing serial radiological testing. To reduce the dose one can change the CT acquisition settings or decrease the sampling density. OBJECTIVE: In this study we determined the minimum desirable sampling density to ascertain the degree of air trapping in children with cystic fibrosis. MATERIALS AND METHODS: Ten children with cystic fibrosis in stable condition underwent a volumetric spiral CT scan. The degree of air trapping was determined by an automated algorithm for all slices in the volume, and then for 1/2, 1/4, to 1/128 of all slices, or a sampling density ranging from 100% to 1% of the total volume. The variation around the true value derived from 100% sampling was determined for all other sampling densities. RESULTS: The precision of the measurement remained stable down to a 10% sampling density, but decreased markedly below 3.4%. CONCLUSION: For a disease marker with the regional variability of air trapping in cystic fibrosis, regardless of observer variability, a sampling density below 10% and even more so, below 3.4%, apparently decreases the precision of the evaluation.

PubMedID: 19096836