Imaging and machine learning

Our research program is characterized by combining high throughput technologies, computational modeling and experimental cell biology to streamline the wealth of biological knowledge to real clinical applications. Using a design algorithm, we have generated numerous of different patterns, which can first be reproduced on a silicon mold and then imprinted onto polymers using microfabrication. We typically apply quantitative high content imaging and machine learning algorithms to characterize the response of cells to these thousands of different surfaces to learn more about the relation between surface topography and cell response. For this, we use a high speed and fully motorized Ti-S inverted microscope of Nikon.

EM images of the TopoChip platform, a high-throughput screening platform with 2176 unique topographies.

Quantitative high content screens are applied on ICC experiments, which are subsequently analyzed to characterize the response of the cells on the topographies.

Quantitative high content screens are applied on ICC experiments, which are subsequently analyzed to characterize the response of the cells on the topographies. Our historical focus has been on bone tissue, e.g. we were the first lab to use endochondral ossification as a strategy for bone tissue engineering, thus impinging on the intrinsic property of stem cells. However, we are also interested in the interaction between biomaterials and other cell/tissue types such as endothelial cells, macrophages and tenocytes.

Further reading

Unadkat et al., Proc Natl Acad Sci U S A, 2011
Hulsman et al., Acta Biomater, 2015
Reimer et al., Sci Rep, 2016