TABLE II.

SUV Errors Over the Volume Compared to CTAC

Method	SUV RMSE ($\mu$, $\sigma$, p-value)
In-plane with metal implant
BpCT-AC	0.10 SUV ($\mu =-0.014$SUV, $\sigma =0.10$SUV,p< 0.001)
Naive MLAA	0.19 SUV ($\mu =0.06$SUV, $\sigma =0.18$SUV,p < 0.001)
MLAAPET (${\beta }_{MR}={10}^3$)	0.16 SUV ($\mu =0.047$, $\sigma =0.15$, p < 0.001)
MLAAPET (${\beta }_{MR}=7\times {10}^5$)	0.09 SUV ($\mu =0.001$, $\sigma =0.09$, p < 0.001)
MLAAPET (${\beta }_{MR}=5\times {10}^6$)	0.09 SUV ($\mu =-0.008$, $\sigma =0.09$, p < 0.001)
UpCT-MLAA	0.12 SUV ($\mu =0.018$SUV, $\sigma =0.12$SUV, p< 0.001)
Out-plane with metal implant
BpCT-AC	0.086 SUV $\mu =0.009$ SUV, $\sigma =0.085$SUV, p< 0.001)
Naive MLAA	0.14 SUV $\mu =0.018$ SUV, $\sigma =0.14$SUV,p < 0.001)
MLAAPET (${\beta }_{MR}={10}^3$)	0.13 SUV $\mu =0.016$, $\sigma =0.13$, p < 0.001)
MLAAPET (${\beta }_{MR}=7\times {10}^5$)	0.09 SUV $\mu =0.012$, $\sigma =0.09$, p < 0.001)
MLAAPET (${\beta }_{MR}=5\times {10}^6$)	0.09 SUV ($\mu =0.010$, $\sigma =0.09$, p < 0.001)
UpCT-MLAA	0.086 SUV ($\mu =0.012$SUV, $\sigma =0.085$SUV, p< 0.001)