GROUP: intelligent imaging

Novel fluorescence microscopy techniques have extended imaging beyond simply descriptive visual inspection in life science. The intelligent imaging group focuses in particular on automation of high content screening in 3D tumor models.



 

"...automated microscopic imaging based on machine vision and learning provides new tools for high content screening in systems biology and personalized medicine."

Using such technologies, the intelligent imaging group is focusing on assays and screening of 3D cultured cancer cells to correlate subcellular phenotypes, cellular function, and tumorgenesis on tissue level.

 

As an interdisciplinary team we unify expertise in biology, physics, computer science, and aim to further develop imaging systems to predict mechanistic and diagnostic models in tumor cell biology.

Advanced fluorescence live cell imaging and image based quantification of nuclei and organelles permit unbiased automatic analysis of e.g. mitotic aberrations followed by cell death. Even the identification of rare events can be controlled through automatic recognition of specific cellular or mitotic phases.

Dual view inverted light sheet microscopy (diSPIM) allows unsurpassed 3D time lapse imaging at low photoxicity in cultured tissues (right image).

Distributed computing techniques and machine intelligence are in place to explore cells with suitable throughput also in three dimensions over time.

 

 

 

diSPIM_4

 


 

 

 

Based on minimal object information (left image, white box) we are able to reconfigure microscope systems of different origin remotely for acquisitions in higher spatial and temporal resolution (center and right image).

Following the same concept Micropilot automatically classifies objects by machine learning algorithms, feedback  positions to the controlled microscope, and executes more complex imaging or spectroscopy protocols only for a selected region.