Table 2.
Normal tissue complication probability (NTCP) models. This table shows the parameter estimates for the Lyman-EUD-model, the Lyman-MED-model and the cutoff-dose logistic regression model: 95 %-confidence interval, LogLikelihood (LL), corrected Akaike information criterion (AICc)
| Model | Parameter estimates | 95 % CI | LL | AICc | |
|---|---|---|---|---|---|
| Lyman-EUD | D50 = 44.9 | 19.4 | 75.1 | −38.43 | 83.25 |
| m = 0.34 | 0.19 | 1.06 | |||
| n = 0.34 | 0.02 | 19.9 | |||
| Lyman-MED | D50 = 32.8 | 27.7 | 52.5 | −39.08 | 82.35 |
| m = 0.48 | 0.28 | 1.29 | |||
| Cutoff-dose | Dc = 37.9 | 28.2 | 57.3 | −38.17 | 82.67 |
| β0 = -3.06 | −5.1 | −0.94 | |||
| β1 = 0.06 | 0.02 | 0.12 | |||