Feasibility of sub-milliSievert CT of the cervical spine: Initial results in fresh human cadavers

To investigate the feasibility of sub-milliSievert CT of the cervical spine in fresh human cadavers using a standard-dose (SD) and four different reduced-dose (RD) protocols reconstructed with filtered back projection (FBP) and iterative reconstruction (IR).

The cervical spine of 29 cadavers was examined using different RDCT protocols with decreasing reference tube currents (RDCT-1:70 mAs; RDCT-2:50 mAs; RDCT-3:30 mAs; RDCT-4:10 mAs) at 140 kV. A clinical SDCT (160 mAs, 120 kV) served as reference. Raw data were reconstructed using FBP and two increasing levels of IR (IRL4&6). Images of the upper (C1-4) and lower (C5-7) cervical spine were evaluated for image quality, diagnostic acceptability and visibility of anatomical structures according to a 5-point-scale.

Image quality of the upper cervical spine was diagnostically acceptable for all protocols using FBP and IR except for RDCT-4 with FBP. Image quality of the lower cervical spine was rated as non-diagnostic in RDCT-3 with FBP and RDCT-4 with FBP and IR. RDCT-3 with IR was the most reduced dose CT protocol allowing diagnostically acceptable image quality for both upper and lower cervical spine in all cadavers. RDCT protocols achieved significantly reduced effective radiation doses (SDCT: 1.5 ± 0.7 mSv; RDCT-1:1 ± 0.6 mSv; RDCT-2:0.7 ± 0.4 mSv; RDCT-3:0.4 ± 0.2 mSv; RDCT-4:0.2 ± 0.1 mSv; p < 0.001).

Diagnostically acceptable sub-milliSievert CT of the cervical spine is feasible with a low reference tube current at 140 kV using iterative reconstruction and could be suitable for isolated cervical trauma in cooperative patients.