Florian Jug is a research group leader at the Center for System Biology Dresden and the Max-Planck Institute of Molecular Cell Biology and Genetics. The overarching goal of his research is to push the boundaries of what image analyses and machine learning can do for quantifying biological data. Research in the Jug Lab is aiming at finding ways to efficiently analyze large amounts of microscopy data, while avoiding impossible amounts of manual analyses and curation tasks – often a major bottleneck in biomedical research projects. Besides finding novel algorithms and machine learning methods, the Jug Lab is also critically involved in the development of Fiji, a popular and widely applicable open software tool for biomedical image analysis and the ideal way to disseminate newly developed methods.
In recent years, fluorescent microscopy saw tremendous advances. Today we routinely image beyond the resolution limit, acquire large volumes at high temporal resolution, and capture many hours of video material showing processes of interest inside cells, in tissues, and in developing organisms. Despite these possibilities, the analysis of raw datasets is usually non-trivial and cumbersome.
In my talk, I will show how machine learning can help to tap the full potential of fluorescent microscopy by first enhancing raw data, followed by joint segmentation and tracking of objects in time-lapse data. Although being powerful in their own right, the combination of such approaches unrolls the full potential of (semi-)automated image quantification pipelines.
The top image shows a comparison of low SNR input data (RAW) with restoration results as we obtain them with CARE (Restoration) and a high-SNR acquisitions (Ground- truth). The bottom image shows a zoomed comparison, as indicated by dashed box in the top image. My talk will argue that CARE can improve/enable the analysis of microscopy data in various ways.