Automated image analysis of biomedical microscopy data

Recent advances in microscopy techniques provide insights into cellular structures and mechanisms that were not accessible a few decades ago. While in the era of light microscopy we could only speculate about certain biological processes, nowadays we can study them at a high level of detail. Unfortunately, the stunning microscopic images are often used for illustrative purposes only, while statistical tests are often omitted or only performed based on a manual image analysis. In order to obtain statistically sound results, large amounts of data need to be analysed, for which automated image analysis is crucial. In this talk, I will address different steps of microscopy image analysis, such as preprocessing, segmentation and object detection, as well as various aspects of quantitative analysis: from simple object counting and size evaluation to characterization of object shape and co-localization analysis. I will demonstrate how automated image analysis routines enable objective and high-throughput evaluation of confocal, multiphoton and light-sheet microscopy data, which – in turns – allows to answer specific questions in the fields of biology and medicine.