Castration Resistant Prostate Cancers – an Expert Derived Grading System Augmented by Machine Learning Image Analysis Methods
A collaboration with Prof Johann DeBono (ICR)
Clare Verrill, Johann DeBono, Jens Rittscher and others
This builds on collaboration within the 100,000 Genomes Project Prostate Genomic Clinical Interpretation Partnership (GeCIP) (DeBono – lead, Verrill – member)
Utilising a unique cohort of castrate resistant (hormone resistant) prostate cancers based at ICR, which have undergone extensive molecular testing, we will seek to develop a urological pathologist expert panel based grading system which will seek to predict time to adverse outcome (progression, death). This exercise will then be undertaken by machine based learning methods, in order to either augment or improve upon the expert based grading system.