A spatial statistical-analysis of tumor-growth
Growth models for tumors commonly are developed for a one-dimensional summary (e.g., number of cells, volume). At the supracellular level, however, ignoring tumor shape leads to oversimplification of the growth mechanism. For example, an oncologist would view the discovery of a compact tumor of volume v differently from the discovery of two osculatory compact tumors each of volume v/2. This article presents a growth model that uses shape information at a previous time to describe the tumor at the present time. An analysis is given of three successive two-dimensional images of cell islands, which are obtained from in vitro growth of human breast cancer cells.