Deep Learning to predict recurrence in triple negative breast cancer (TNBC)

Deep Learning to predict recurrence in triple negative breast cancer (TNBC)


Triple negative breast cancer (TNBC) accounts for 15-20% of all invasive breast cancers. TNBC is defined by the absence of expression of the hormonal receptors (HR) and the absence of overexpression of the human epidermal growth factor receptor 2 (HER2).

Research question: Is it possible to predict recurrence for TNBC based on morphological features?


Aim of the presently proposed ‘proof of concept‘ study is to develop digital pattern recognition algorithms (more specifically: deep neural networks) for the extraction of morphological features from scanned H&E stained tissue sections from TNBC which are indicative for the risk of recurrence.

Clinical relevance

Approximately 1 in 4 TNBC patients develop a recurrence within 3 years, after which the survival outlook is very poor. To date, no biomarkers are available which are indicative for response to therapy or of prognosis for TNBC. TNBC is a morphologically heterogeneous group comprising mainly high grade tumors with highly diverse morphological features. Although literature suggests that morphology could be related to prognosis, the human eye can only capture these visual characteristics to some extent and these manual observations are hampered by interobserver variation.


  • Students with a major in computer science, biomedical engineering, artificial intelligence, physics, or a related area in the final stage of master level studies are invited to apply.

  • Interest in image analysis and machine learning

  • Affinity with programming in Python is required.


  • Project duration: 6 months

  • Location: Radboud University Nijmegen Medical Centre

  • For more information please contact Francesco Ciompi


Francesco Ciompi

Francesco Ciompi

Associate Professor