The proliferation of cyanobacteria often provides significant benefits for humans, ranging from the use of these microorganisms in various medicines and vaccines to their application in fertilizers and other biofuels. Although there are many benefits in various fields, there are also several drawbacks to using wrong microorganisms, some of which can, actually, be toxic in humans. Therefore, the identification and environmental monitoring of these cyanobacteria and diatoms is crucial. This requires analyzing the environment in which they live and their development at different stages of maturity.
Sometimes, the study of these microorganisms can be excessively complex to human eyes and knowledge due to our limitations. This is where technology comes into play, as it has been shown to improve decision-making mechanisms when we introduce deep learning models that support the decisions of the expert community.
This project was initiated with the aim of developing a mechanism to aid in the identification of different genera and species of cyanobacteria and diatoms. We have utilized images taken by experts in aquatic environments to train and integrate various artificial intelligence models into a web environment.
On this webpage, you can find an overview of the project. To request access to the page where the trained deep learning models and some examples are collected, and where you can use the tool, follow this link where you will be asked to fill out a questionnaire. You can also access this link in the Call for Demo tab. Additionally, in Open Demo tab, you can find various explanatory videos demonstrating the platform's usage.
This project has been carried out by a team of experts from the University of Castilla-La Mancha (UCLM) together with the Daza de Valdés Institute of Optics (CSIC), the Autonomous University of Madrid, University of León (UL).
Compilation of Deep Learning models:
You Only Look Once (YOLO)
Residual Network (ResNet)
ConvNeXt
AlexNet
Faster R-CNN
In-house Multimodal Network
Microscopes used for image digitization:
Olympus CX 41 photomicroscope
DP20 Olympus Digital camera
Olympus BX63 equipped with Normarski
Brunel Microscope LTD - AMA050
Openflexure microscope
On the left side, we can see various images of diatoms used in the project. On the right side, there is a compilation of different images of cyanobacteria used in the project:
We acknowledge financial support from the European Commission through the project DIAMOND TED, grant number TED2021-132147B-100, supported by the Spanish Ministry of Science, Innovation and Universities, and by the European Union NextGenerationEU/PRTR.