OBJECTIVE: To investigate the effect of the radiation dose (tube current second product) and the attenuation value of nodules with ground-glass opacity (GGO) on their detectability at multidetector computed tomography (MDCT). METHODS: We scanned a chest CT phantom that included simulated GGO nodules with an MDCT scanner. The attenuation value of the simulated lung parenchyma was -900 Hounsfield units (HU); it was -800 and -650 HU for the simulated GGO nodules. We used a tube current second product of 180 mA as the standard and 21, 45, 60, and 90 mAs as the low-dose and performed receiver operating characteristic analysis to compare the performance of 5 radiologists in detecting GGO nodules at each milliampere. To assess the detectability of GGO nodules on human lung images, the observers were presented with 38 GGO nodules from 15 patients. The 5 radiologists independently reviewed chest CT images at 21 and 45 mAs. RESULTS: In the phantom study, the Az value for GGO nodules with a CT number of -800 HU was significantly lower at 21 than 180 effective mA (0.86 vs. 0.96; P < 0.01). There was no statistically significant difference in the Az value of GGO nodules with a CT number of -650 HU, irrespective of milliamperes used (P = 0.165). In the clinical study, 39.5% and 25.8% of GGO were missed at 21 and 45 mAs, respectively. CONCLUSIONS: At MDCT, GGO nodules with a CT number of -650 HU or less were difficult to detect at the lower milliampere settings (21 and 45 mAs).
OBJECTIVE: To investigate the effect of the radiation dose (tube current second product) and the attenuation value of nodules with ground-glass opacity (GGO) on their detectability at multidetector computed tomography (MDCT). METHODS: We scanned a chest CT phantom that included simulated GGO nodules with an MDCT scanner. The attenuation value of the simulated lung parenchyma was -900 Hounsfield units (HU); it was -800 and -650 HU for the simulated GGO nodules. We used a tube current second product of 180 mA as the standard and 21, 45, 60, and 90 mAs as the low-dose and performed receiver operating characteristic analysis to compare the performance of 5 radiologists in detecting GGO nodules at each milliampere. To assess the detectability of GGO nodules on human lung images, the observers were presented with 38 GGO nodules from 15 patients. The 5 radiologists independently reviewed chest CT images at 21 and 45 mAs. RESULTS: In the phantom study, the Az value for GGO nodules with a CT number of -800 HU was significantly lower at 21 than 180 effective mA (0.86 vs. 0.96; P < 0.01). There was no statistically significant difference in the Az value of GGO nodules with a CT number of -650 HU, irrespective of milliamperes used (P = 0.165). In the clinical study, 39.5% and 25.8% of GGO were missed at 21 and 45 mAs, respectively. CONCLUSIONS: At MDCT, GGO nodules with a CT number of -650 HU or less were difficult to detect at the lower milliampere settings (21 and 45 mAs).
Authors: Rolf Symons; Tyler E Cork; Pooyan Sahbaee; Matthew K Fuld; Steffen Kappler; Les R Folio; David A Bluemke; Amir Pourmorteza Journal: Phys Med Biol Date: 2016-12-17 Impact factor: 3.609