DIAMOND: DIgitAl MicroscOpy platform for phytoplaNkton iDentification

 

Start: Diciembre 2022

End: Septiembre 2025

Principal researcher: Gloria Bueno

Summary:

This project has advanced digital microscopy and artificial intelligence tools for the automatic identification and analysis of freshwater phytoplankton, with a focus on diatoms and cyanobacteria. Its main goal has been to develop cost-effective, reliable, and accessible solutions for water monitoring, supporting applications such as watershed management, drinking water production, and early warning systems.

An open-source platform has been developed for phytoplankton identification and analysis, integrating AI modules for the automatic classification of diatoms and cyanobacteria, as well as tools for managing digital biological image collections. Its modular architecture allows new functionalities to be added easily and supports integration with reference databases. The platform is available at: Digital ecology

A portable low-cost microscopy system with edge computing capabilities has also been designed and tested, enabling multimodal image acquisition in field conditions (see Real-Time Edge Computing vs. GPU-Accelerated Pipelines for Low-Cost Microscopy Applications).

In addition, the project has generated curated datasets of diatoms and cyanobacteria from controlled experiments and Spanish reservoirs, providing valuable resources for training and validating AI models in digital microscopy. These datasets are currently being prepared for dissemination through specialist journals and open-access repositories.

The project has also achieved significant impact through scientific publications, international conference participation, outreach activities, specialized annual training courses organized with the CSIC Postgraduate School, and further actions in academic development and technology transfer.

Digital ecology – Papers

Project partners: IO-CSIC, UAM

Funding: This project was funded by MCIN/AEI/10.13039/501100011033 and by the European Union NextGenerationEU/PRTR (Ref. TED2021-132147B-I00).