Fig. 2

Phantom- and cadaver-based computed tomography protocol development. Plain radiographic shunt series confirming appropriate positioning of the de novo inserted ventriculoperitoneal shunt (a–c). Objective image noise values for the neck, thorax, and abdomen using two kVp/mA settings (80/10 and 100/10) reconstructed using hybrid IR and pure IR (d). Data represent mean and standard deviation. Difference in means at each anatomical level was assessed using paired sample t-tests. For comparisons of hybrid IR and pure IR at 80/10 (neck: * p = 0.019; thorax: **p = 0.003; abdomen: *p = 0.013) and 100/10 (neck: ** p = 0.002; thorax: **p = 0.004; abdomen: **p = 0.007). There was no significant difference (ns) in image noise between pure IR images acquired at 80/10 or 100/10. (e) Three-dimensional reconstruction demonstrating the ventriculoperitoneal shunt at the final protocol setting 100/10