## Optimum parameters in a model for tumour control probability, including interpatient heterogeneity: evaluation of the log-normal distribution

Keall, P. J., Webb, S.
(2007)
*Optimum parameters in a model for tumour control probability, including interpatient heterogeneity: evaluation of the log-normal distribution.*
PHYSICS IN MEDICINE AND BIOLOGY, 52 (1).
pp. 291-302.
ISSN 0031-9155

Full text not available from this repository.

## Abstract

Optimum parameters in a model for tumour control probability, including interpatient heterogeneity: evaluation of the log-normal distribution The heterogeneity of human tumour radiation response is well known. Researchers have used the normal distribution to describe interpatient tumour radiosensitivity. However, many natural phenomena show a log-normal distribution. Log-normal distributions are common when mean values are low, variances are large and values cannot be negative. These conditions apply to radiosensitivity. The aim of this work was to evaluate the log-normal distribution to predict clinical tumour control probability ( TCP) data and to compare the results with the homogeneous ( d-function with single alpha-value) and normal distributions. The clinically derived TCP data for four tumour types-melanoma, breast, squamous cell carcinoma and nodes-were used to fit the TCP models. Three forms of interpatient tumour radiosensitivity were considered: the log-normal, normal and d-function. The free parameters in the models were the radiosensitivity mean, standard deviation and clonogenic cell density. The evaluation metric was the deviance of the maximum likelihood estimation of the fit of the TCP calculated using the predicted parameters to the clinical data. We conclude that ( 1) the log-normal and normal distributions of interpatient tumour radiosensitivity heterogeneity more closely describe clinical TCP data than a single radiosensitivity value and ( 2) the log-normal distribution has some theoretical and practical advantages over the normal distribution. Further work is needed to test these models on higher quality clinical outcome datasets.

Item Type: | Article |
---|---|

Authors (ICR Faculty only): | Webb, Steve |

All Authors: | Keall, P. J., Webb, S. |

Uncontrolled Keywords: | Radiation-therapy; radiotherapy; carcinoma; logarithm; biology; cancer |

Research teams: | ICR divisions > Radiotherapy and Imaging > Radiotherapy Physics Modelling |

Date Deposited: | 10 Aug 2007 20:58 |

Last Modified: | 10 Feb 2010 11:47 |

URI: | http://publications.icr.ac.uk/id/eprint/3026 |

### Actions (login required)

View Item |