From the editor: In a very interesting study published in last weeks issue of Science Translational Medicine Andrew Beck and colleages demonstrate that modern bioinformatics has the potential to transform pathology. The authors use sophisticated image analysis for a systematic analysis of breast cancer morphology and correlate over 6000 features extracted from the tissue samples with patient outcome. They could show that a subset of these  – including both epithelial and stromal features – is highly associated with survival. David Rimm puts this landmark work succintly in focus in a complementary comment: “Computer-based quantification of tumor morphology has arguably solved the problem of standardized tumor grading.” His point is not that Beck et al. solved all challenges already. But they indicate the direction modern pathology should take.

I couldn’t agree more. We at Definiens are convinced that this is the way to go and hope to make going it a little easier with the introduction of the Definiens Image Miner. While not available yet at the time Beck et al. performed their breaking work, it has been designed to make similar things happen. And it builds on the same  sophisticated image analysis technology Becks C-Path (computational pathologists) relies on – Definiens Cognition Network Technology.

Both the article itself and the insightful introduction by David Rimm are recommended reading for anyone interested in how quantitative image analyis can be combined with powerful data mining to transform pathology. Open-minded pathologists will embrace their computational colleague and should keep on teaching him. The rewards may well include better tools to fight cancer.