The potential of machine learning in the digitalisation of pathology

Cancer diagnostics is facing immense challenges. Against the backdrop of the sharp rise in the number of cancer cases worldwide, pathology must produce more and more diagnoses while capacities remain the same or even decrease. The digitalisation of cancer diagnostics plays a decisive role in this issue. The enormous importance of machine learning and deep learning for the digitisation of pathology and thus the improvement of cancer diagnostics is described by the founder and CTO of Mindpeak Dr. Tobias Lang and Dr. Holger Hauspurg from the Institut für Hämatopathologie Hamburg in an article for the latest issue (04 / 2021) of the German Journal Onkologie (www.journalonkologie.de).

According to this article, digitalisation can offer enormous advantages for oncological therapy in a number of simplifications and improvements: With digitalisation, more efficient logistics in case distribution and the use of telemedicine become possible. Not least the SARS-CoV-2 pandemic has shown what is not only possible but necessary for this. However, the even greater benefit of digital pathology lies in the diagnostic support of the human expert – especially the pathologist – by the computer. Artificial intelligence (AI) methods, and especially machine learning (ML), have the potential not only to provide considerable support to experts in oncological diagnostics, but also to be a decisive tool on the way to precision medicine.

Just how advanced the possibilities for cancer diagnostics with the help of artificial intelligence already are is shown by the fact, that Mindpeak – a European leader in computational pathology – announced several days ago that it has received CE-IVD Mark for BreastIHC, an Artificial Intelligence software that detects and quantifies breast cancer cells, for primary diagnosis. Germany based Mindpeak is the first company to receive approval for a Deep Learning solution which can distinguish between tumorous and non-tumorous structures on a cellular level.

Click here for the article: http://userpage.fu-berlin.de/tlang/pub/hauspurg_lang_journal_onkologie_2021.pdf