Low-dose CT in body-packers: delineation of body packs and radiation dose in a porcine model

Purpose To compare low-dose computed tomography (CT) with standard CT and conventional radiography (CR) regarding delineation of body packs and radiation dose. Methods Nine samples of illicit drugs including cocaine, heroin, and hashish were positioned in the rectum of a 121.5 kg pig cadaver. Each sample was scanned on a 64-row MDCT with 120 kV: one standard modulated pelvic protocol (STD), and without modulation at 80 mA (LD80), 30 mA (LD30), and 10 mA (LD10). Additionally, conventional abdominal anterior-posterior radiographs (77 kV and 106 +/- 13 mA) were taken. Body pack characteristics (wrapping, content, shape) were rated independently by two radiologists and summarized to a delineation score from 0 to 9 with scores >= 6 representing sufficient delineation. Mean delineation scores were calculated for CR and CT protocols. These were additionally differentiated for readings in soft tissue (S), lung (L), user defined, variable window settings (V), and in cumulative window evaluation including all the other window settings (SLV). Effective doses were calculated (mSv). Results The CR delineation score was insufficient (3.1 +/- 2.5; 2.4 +/- 0.3 mSv). For CT, the SLV window setting performed best (p<0.01). Its score significantly (p<0.01) declined with decreasing effective radiation doses: STD (8.8 +/- 0.5; 10.6 mSv), LD80 (8.2 +/- 0.7; 2.6 mSv), LD30 (6.8 +/- 1.3; 1.0 mSv), and LD10 (4.6 +/- 1.9; 0.3 mSv). Thus, LD30 was the protocol using the lowest but sufficient dose. Moreover, for LD30 further differentiation between the particular window settings resulted in scores of 6.4 +/- 1.3 (L), 6.3 +/- 1.2 (V), and 3.1 +/- 1.0 (S). Conclusions With appropriate window settings, low-dose CT at 30 mA allowed for sufficient body-pack delineation below the dose of CR, which itself performed insufficient.