Unveiling the Secrets of a Seawater Drop: AI-based Microscopy Segmentation Reveals Biodiversity

Did you know that a single drop of seawater can be a place to live for a remarkable number of organisms? And which profound impact human activities have on the sea as an invaluable habitat?

Europe’s coasts and oceans are treasures of biodiversity, home to a myriad of plants and animals. These living beings play an integral role in the stability and sustainability of larger ecosystems. However, the effects of human activities on these fragile environments not only sound like threats: habitat degradation, pollution, biodiversity loss and climate change, just to name a few. These effects have raised a growing awareness of the urge to comprehend and address these adverse effects of human activities. The European Molecular Biology Laboratory (EMBL) has taken up this challenge and embarked on a mission:
 

Confocal microscopy was an essential tool on the TREC expedition: It allowed Dr. Wiegand and her colleagues to explore the subcellular architecture of living plankton and other marine species at high resolution in 3D. The technique also gave them the flexibility to visualize different natural pigments as well as specific stainings, and to perform experiments like photo-activation of convertible dyes to mark individual cells.
The TREC expedition uncovered a wealth of new information about the molecular and cellular processes of marine organisms.

Not only for molecular biologists, but for all researchers, it is critical to have the right tools to perform experiments and analyze data accurately. That is why solutions for automated detection and imaging with AI-based segmentation are both flexible and scalable. Automated detection and imaging of plankton species is just one example of an application where such an automated recognition could be of interest for. Other application examples are the automated recognition of different cell types in blood samples for cell counting or automated tagging and follow-up of stem cells – in summary, all applications in which cellular structures need to be differentiated from one another.

For some applications, automatic tagging with a fluorescent marker via photo-activation is not necessary. In this case, a confocal microscope is not even required, but widefield microscopes can be used to run the AI-based automated recognition of cell types.
 

Now that the TREC sails have been hauled in, Dr. Wiegand and her team have been reviewing the data acquired during the expedition. They have set up a high-throughput imaging pipeline to analyze more than 500 samples collected in the field and stored in their sample hub.

The TREC expedition shows how powerful advanced microscopy techniques are in exploring the diversity of coastal ecosystems. By using confocal microscopy and AI-supported segmentation, Dr. Wiegand and her colleagues were able to gain new insights into the molecular and cellular processes of marine organisms: 

They will investigate this further in the samples they collected during the expedition. The team's efforts to make their data publicly available will facilitate future studies and provide a valuable resource for researchers in the field.

Tina Wiegand's podcast about the TREC expedition "RausFinden" on YouTube (available in German)

ZEISS partnership with EBML