Sophie F. Riches1, Geoffrey S. Payne1, Charlie Jameson2, Christopher Ogden3, Mike Partridge4, Veronica A. Morgan1, Sharon L. Giles1, Nandita M. deSouza1
1CR-UK and EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom; 2Department of Histopathology, University College London Hospitals, London, United Kingdom; 3Department of Surgery, Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom; 4Gray Institute for Radiation Oncology and Biology,Department of Oncology, University of Oxford, Oxford, Oxfordshire, United Kingdom
Currently the use of a boosted radiation dose to tumour in localised prostate cancer is limited by the accuracy of tumour localisation within the prostate on anatomical imaging; functional imaging is therefore being explored. This study shows that a model combining parameters from multiple functional MR techniques gives greater accuracy in discriminating between tumour and normal prostate tissues than individual parameters. The average accuracy of prediction of tissue type for individual patients is lower than that suggested by the population ROC curve due to high inter-patient variability within the population.