Computational Modelling of Possible Temporal Growth Curves of Pre-Metastic Primary Tumours by Linear Induction

In this paper we aim to devise a simple pseudo-bootstrapping
algorithm from which we can form an educated opinion based
on limited clinically observed data, regarding whether a benign
tumour can subsequently become malignant. The strength of
our approach lies in its simplicity—our model does not involve
any physiological variables but is entirely based on time series
observations of growth in tumour volume from the point of first
detection to the onset of metastases. The basic premise under¬
lying our methodology is that a primary tumour which initially
starts off as benign hyperplasia cannot increase in volume
beyond some critical limit without showing metastasis. We
propose a practicable statistical method to extrapolate any such
critical limit to primary tumour volume.