Haematology meets Image Processing


Blood cells are created and mature in the bone marrow. As mature blood cells, they later perform several important tasks in our body. Diseases like leukemia change the number of certain kinds of cells in the bone marrow. In particular, more immature cells are present. Automatically detecting cell type and analysing a bone marrow image in terms of numbers and distribution of certain cell types could accelerate an objective diagnosis and enable further research.


Central research target is the detection, classification, and counting of cells. Each of these steps is challenging and has the potential for new contributions to research. Detection is made difficult by having a high density and diversity of cells. There is a need for sophisticated classification methods due to a large number of cell classes. In addition to classic image processing techniques, several modern technologies such as deep learning provide important tools to overcome these obstacles.


There are several topics for Bachelor and Master Theses. If you’re interested, feel free to send me an email.


This project is a cooperation of the LfB together with haematology experts of the Klinik für Hämatologie, Onkologie, Hämostaseologie und Stammzelltransplantation des Uniklinikums Aachen. Their medical expertise and our knowledge in digital image processing is the basis for innovative research in both fields.


Philipp Gräbel