Plankton (also known as microplankton) are one of the most important animals in the ocean. It is the foundation of the seafood pyramid, which feeds on large plants and animals, as well as indirectly to humans through its dependent fisheries.
“Their lives have a significant impact on our lives,” says Harshith Bachimanchi, PhD student in physics at the University of Gothenburg (Sweden), noting that phytoplankton (microscopic plant-like organisms) produce half of the oxygen we produce. Breathe. The amount of Plankton in the ocean affects the entire carbon cycle of the ocean. In total, single-cell microflora consumes about three times as much carbon as humans emit from fossil fuels.
“But we know surprisingly little about them,” he added, largely because of their microscopic size. With large organisms such as animals or birds, there is a good understanding of who eats whom, how much material is lost, and so on.
Along with researchers in physics and marine science from Gothenburg, Bachimanchi is now investigating how to reduce the lack of knowledge using new microscopic techniques.
“We are preparing to come up with individual solutions for microbes to gain a better understanding of Plankton, the most important but least known animal in the ocean,” says Bachimanchi.
A microscope sensor is placed at the bottom of the well, allowing the area to be completely visualized in a single view. This way all of Planton can be visualized continuously for the duration of the experiment.
The study stems from work presented a few years ago by Giovanni Volpe, a professor of physics in Gothenburg, who co-authored a new study on holographic microscopic methods. With it, the hologram is made up of light reflected through particles. This allows the study of holograms rather than particles.
When light passes through a cell, Bachimanchi explains that it acquires a complex size that depends on the optical properties of the material through which it passes. This creates a holographic profile in the range or pattern of interference on the camera. holograms encode 3D locations of plankton as well as size and refractive index information.
In the Gothenburg study, Plato was placed under a holographic microscope in two different settings. For short-term experiments, the dry mass of microfilant – essentially the mass of the biological sample after the water content was removed – was estimated and its feeding was observed. For long-term experiments, such as cell growth and division, the atoms were stored in a closed circular well.
To analyze the holograms, the team applied an in-depth study method that used two artificial neural networks in sequence. The first – the regression network or RU-Net – captures the hologram recorded from the camera and locates the plankton along with the dry mass and the vertical distance. The form of holograms (transformation patterns) acts as a unique fingerprint of the dry mass refraction index size and the posterior position and axis of the plankton. “This information is also useful in classifying plankton by its type,” says Bachimanchi.
The second neural network – the medium-weight convolutional network, or WAC-Net – then takes a small crop of planktons and refines the information to predict more accurate values for dry mass and vertical distance. The microplankton examined using the new technique is no larger than a few hundred millimeters. Traditionally for ecologically related microorganisms such as Plankton, this small size means that researchers have to rely on bulk and average measurements. Viewing and analyzing each microphone with existing microscopes and imaging techniques is not possible.
Bachimanchi says another existing challenge with holographic data analysis is that it is expensive to calculate, especially for holographic microscopes, less than the lenses that his team uses in their studies. He added that it provides a large view image that can capture more than 700 Plankton at a time, but “the computational process becomes enormous,” he added. “And extracting dry mass information from holograms in a row is not straightforward.”
In-depth learning algorithms avoid computational processes that rely heavily on holographic data, study notes, and allow for rapid measurement over time.
The combination of holographic microscopes with in-depth learning makes the technology more diverse and the order of magnitude faster, which is the key to tracking and characterizing Plankton throughout their lives. Bachimanchi says the new approach also provides a robust complement to marine microbial ecology, Bachimanchi says, because it allows for non-destructive and minimal 3D positioning and dry mass of individual microbes.
“By combining holographic microscopy with artificial intelligence, we can imitate many micro-organisms, such as plankton and diatoms, at an individual level throughout their lives,” says Bachimanchi. “And we can gradually see what they are, depending on what they eat, how fast they grow and “Swimming and so on.”
In addition to working with microbes (for example, microscopic plates and bacteria) – tracking their movements and hiding places for a continuous period of time – new microscopes / in-depth study techniques serve as a quick method for sizing and Weigh the cells in suspension. Bachimanchi says the hardware of the technique is also cheap, and researchers can share their code with examples and demonstrations to facilitate use by other teams.
“Since microscopic organisms in the oceans form the basis of the seafood network and drive large-scale processes on a global scale, the new approach allows us to better understand the effects of micro-organisms on large-scale processes such as global carbonation. Cycle, ”Bachimanchi said. “Our approach enables us to follow individual organisms in the soup of this life.”