In silico Pathology - Improving diagnosis by utilizing big data and software-driven automation of pathology

The main project is all about researching and develop ICT solutions to supplement or replace methods in pathology to increase productivity and identify computer based prognostic biomarkers based on Big Data produced by digital pathology. It is generally accepted that most prognostic studies suffers from severe undersampling. Increasing the number of patients and number of samples necessary to account for the heterogeneous nature of tumors and the individual nature of patients, most studies would have to increase the workload by a factor of 10-100. With limited resources and strongly limited access to pathologists, this cannot be achieved unless we manage to digitalize and largely automate both the preparation and analysis required to render a diagnosis and identify prognostic biomarkers. This is what this project is designed to do, and we have put together a group of international leaders in the different fields involved, from robotics to digital image analysis and from tumor pathology to cancer surgery and oncology. We intend to focus on two major cancer forms, i.e. colorectal cancer and prostate cancer and focus on generic markers involved in control of cell division, gene expression and genomic instability. We will work with large retrospective clinical materials with a known clinical outcome and the methods will hence be applied to archival paraffin imbedded material from routine pathology. There are three main challenges; automation of lab procedures, an efficient pipeline for digitalization of microscopy and analysis, and utilizing the Big Data produced to identify and establish robust generic biomarkers for cancer prognosis and prediction. This may at first sight seem like 3 different projects, but the latter cannot be solved without the first two, and all three are required to establish a more efficient and objective cancer diagnostics that is equally available to all patients.