臺灣農藥科學 第一期

49 / 220

基準劑量推估每日可接受攝食量 47
附錄二、基準劑量分析軟體中非連續型與連續型數據模式分類
Appendix 2. Model classification of quantal and continuous data in BMDS
Model Equation for the probability of a response Parameter constraints
Quantal Models Here, 0 ≤ P(X) ≤ 1, X > 0
(1) Gamma Model P(X) = γ + (1 – γ) [Γ(α)-1 {0∫ βx tα-1 e t dt}] α ≥ 0, β > 0, 0 ≤ γ < 1
(2) Logistic Model P(X) = F{-( α + βX)} = F{-([X + (-α)(1/β)]
/ |1/β|)}
0 ≤ γ < 1, -∞ < α < +∞, β > 0
(3) Log-Logistic Model P(X; γ, β) = γ + (1 – γ ) F{-(α + β lnX)}
(4) Log-Probit Model P(X) = γ + (1 – γ) Φ{α + β lnX} 0 ≤ γ < 1, -∞ < α < +∞, β > 0
(5) Multistage Model P(X) = γ + (1 – γ) [1 – exp{- Σ βj Xj}] j = 1, ..., k, 0 ≤ γ < 1
(6) Multistage-Cancer
(7) Probit Model P(X) = P(X; γ, β) = Φ{ α + βX} 0 ≤ γ < 1, -∞ < α < +∞, β > 0
(8) Weibull Model P(X) = γ + (1 – γ) [1 – exp{- βxα}] α ≥ 0, 0 ≤ γ < 1, β > 0
(9) Quantal-Linear
Continuous Models μ(X) is the mean response at
dose X >
Polynomial Continuous
Model
μ(X) = γ + Σ βj Xj j = 1, ..., n
P o w e r C o n t i n u o u s
Model
μ(X) = γ + β Xα α > 0, β > 0
Hill Continuous Model μ(X) = γ + ν X n / (kn + Xn)
Exponential Continuous
Models, a set of nested
models
Model 2: μ(X) = γ exp{sign k X} Model 3:
μ(X) = γ exp{sign (k X)d}
Model 4: μ(X) = γ (c – (c – 1) exp{-1 k X})
Model 5: μ(X) = γ (c – (c – 1) exp{-1 (k X)
d})
Source: 25).