Entry, discharge, radiology and pathology reports and other clinical documents are a valuable resource to be harvested for precision medicine. They are typically stored in a free text format, only little structure is imposed and terminology is heterogeneous. We will apply natural language processing (NLP), machine and statistical learning methods for automated information extraction for dititalized medical reports, and apply these technologies on a concrete example, namely extracting standardized information from radiology and pathology reports for tumour classification. Our goal is to develop computerized methods such that Systematised Nomenclature of Medicine Clinical Terms (SNOMED-CT) concepts and the tumour classification according to the TNM system can be derived for a large collection of radiology and pathology reports automatically.