Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Prof Clare Verrill, Dr Anne Warren, Dr Jon Oxley, the ProtecT Pathology Group, Prof Freddie Hamdy, Prof David Neal, Prof Jens Rittscher

Gleason grading is a well-established system for pathologist assessment of prostatic adenocarcinoma.  It attempts to represent tumour differentiation by stating 2 numerical grades which are added together to make a sum Gleason Score.  This grading system is firmly embedded within clinical practice and has stood the test of time well as an assessment tool, forming (with staging and biochemical data), the basis for patient management, stratification and prognosis.  Despite this, there remain inherent problems.  Inter-pathologist variation can be significant and kappa agreement between pathologists can be extremely poor, reported in the literature to be between 0.15 and 0.7 [Waliszewski P, 2015].  Also, gland formation is a continuum and Gleason grading forces arbitrary boundaries to be drawn. 

Image analysis has the potential to improve on Gleason grading by better representing the spectrum of gland formation, removal of subjective pathologist assessment and capturing far more detailed information than can be appreciated by a human being.  Image analysis can capture spatial information, structure and context with features such as colour, texture and shape being able to be assessed in great detail.  Feature extraction, such as degree of gland formation can be quantitatively assessed by machine learning algorithms with the ultimate aim of identifying new tumour subgroups and prognostic models.

This project aims to develop detailed tissue maps for use in prostatic adenocarcinoma with annotation and segmentation of epithelium (glands), stroma, nerves, blood vessels and inflammatory cells and to develop an unbiased and systematic assessment of tissue images stained with H&E that can be tested to answer the following:

Can image analysis better predict biochemical relapse, development of metastatic disease or death than the original pathologist Gleason Scoring*

Could image analysis have added value to predicting whether patients would be able to remain in the active monitoring group, or would be rendered disease progression free by radiotherapy or radical prostatectomy.