Premium
The assessment of tumour response to treatment
Author(s) -
Demidenko Eugene
Publication year - 2006
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.1111/j.1467-9876.2006.00541.x
Subject(s) - mathematics , linear model , generalized linear model , estimation theory , feature (linguistics) , statistics , least squares function approximation , variance (accounting) , continuation , iterated function , fraction (chemistry) , computer science , mathematical analysis , philosophy , linguistics , accounting , estimator , business , programming language , chemistry , organic chemistry
Summary. A parsimonious model for treated tumours is developed as a continuation of our previous work on regrowth curve theory. The statistical model belongs to the family of marginal non‐linear models since the only linear parameters of the model are tumour specific and random facilitating parameter estimation. An important feature of the model is that it enables the estimation of the fraction of cancer cells surviving the treatment in vivo having easy‐to‐obtain longitudinal measurements of tumour volume. We compare several methods of estimation, including Lindstrom–Bates, iterated reweighted least squares and maximum likelihood. The last two methods are computed via the total estimating equations approach and variance least squares. The theory is illustrated with a photodynamic tumour therapy example.