Cervical spinal computed tomography utilizing model-based iterative reconstruction reduces radiation to an equivalent of three cervical X-rays.

Abstract

To evaluate radiation dose and image quality of cervical spinal computed tomography scanned with low-radiation dose (LD-CT) utilizing model-based iterative reconstruction (MBIR). We retrospectively examined 14 patients (65.5 ± 13.9 years) who underwent both standard-radiation-dose CT (SD-CT) reconstructed with hybrid iterative reconstruction and LD-CT of cervical spine. The radiation dose, objective image quality indicator, which includes signal-to-noise and contrast-to-noise, and subjective image quality score of the anatomical landmarks in the SD-CT and LD-CT were statistically compared. In addition, the measurement errors of the length of C3 vertebrae (height, anteroposterior length, inner and outer pedicle diameters) between SD-CT and LD-CT were analyzed. Radiation dose of LD-CT was reduced to one-sixth of the dose of SD-CT. The objective image quality indicator of LD-CT was significantly better than that of SD-CT. The subjective image quality of LD-CT was relatively worse than that of SD-CT but generally graded as clinically accepted or higher. There was no remarkable difference between SD-CT and LD-CT in the measurement value of height and anteroposterior length. Inner pedicle diameter was significantly (0.21 ± 0.13 mm) smaller, and outer pedicle diameter was (0.24 ± 0.14 mm) larger on LD-CT than on SD-CT. Cervical spinal LD-CT that utilized MBIR enabled radical decrease in radiation dose and provided sufficient image quality for clinical use. This scanning protocol can be a good alternative for protecting patients from exposure to unnecessary radiation, especially when a patient requires multiple CT scans.