Computation of population and physiological risk parameters from cancer data

Abstract

A biological hypothesis known as the "two-stage cancer model" conceptualizes carcinogenesis in mathematical terms. Prof. W. Thilly (MIT) and Prof. S. Morgenthaler (ETH,) and their students, especially Dr. Pablo Herrero-Jimenez have organized age- and birth cohort-specific cancer mortality data from the United States (1900-1997). Unfortunately, no computer program exists that permits cancer researchers to develop and explore quantitative hypotheses about the values of biological parameters through the analysis of this large and comprehensive data set. In this thesis, initial efforts of the Morgenthaler and Thilly groups have been extended by improving and transporting a preliminary Fortran program incorporating an approximate model into a Java-based program. This program, CancerFit, can be used by cancer researchers without backgrounds in mathematics or computer programming. CancerFit is the first of its kind that deals with the carcinogenic process as a whole, rather than in smaller, individual components.