Some aspects of maximum likelihood estimation of multistage cancer dose-response odels with application |
| |
Authors: | HR Al-Khalidi AR Parsa |
| |
Institution: | 1. Department of Epidemiology and Biostatistics , University of Illinois at Chicago , Chicago, Illinois, 60612, U.S.A;2. Department of Statistics and Actuarial Sciences , Drake University , Des Moines, Iowa, 50311, U.S.A |
| |
Abstract: | Quantitative cancer dose-response models play an important role in cancer risk assessment. They also play a role in regulatory processes associated with potential occupational or environmental exposures. The multistage model is currently the most widely used cancer dose-response model. This paper describes the construction of the likelihood function in the special case of the multistage cancer dose-response models. The concavity of the likelihood function is also established. A criterion is developed to determine the degree of the polynomial portion of the multistage model. Finally, the restricted and unrestricted maximum likelihood estimators are considered and applied to some experimental data sets. |
| |
Keywords: | Aic cancer dose-response models likelihood response surface multistage model |
|
|