Digital phantom versus individualised radiation dosimetry in CT examination protocols for children and adults

The aim of this study was to compare the organ doses estimated through individual model-based and digital phantom-based Monte Carlo (MC) methods in radiation treatment plan-oriented computed tomography (CT) examination protocols. Six physical anthropomorphic phantoms that simulate the average paediatric individual as a neonate, 1-year-old, 5-year-old or 10-year-old child and the average male or female adult individual were used. Organ dose was assessed using two MC-powered dosimetry tools: an individual model-based tool (ImpactMC, CT Imaging GmbH, Germany) and a digital phantom-based dosimetry tool (NCICT, National Cancer Institute, USA). Normalised to volume CT dose index (CTDI_vol) organ dose (<i>n</i>OD) was assessed for primarily exposed radiosensitive organs using head, thorax and abdomen/pelvis CT examination protocols intended for radiation treatment plans.<i>n</i>OD were compared with measurements performed using thermoluminescent detectors (TLDs) and physical anthropomorphic phantoms. The average per cent differences between ImpactMC and TLD measurements across all organs and phantoms were 8%, 20% and 16% for head, thorax and abdomen/pelvis, respectively, whereas the corresponding differences between NCICT and TLD measurements were 64%, 48% and 34%. The differences between estimated and measured normalised organ dose values were higher in paediatric than adult phantoms. Compared with NCICT, organ doses estimated with ImpactMC were in closer agreement with TLD measurements: this is due to the methodology employed by ImpactMC to create phantoms that mimic the anatomical characteristics of the examined patient.