Imaging Flow Cytometry and Microscopy Combine to Identify Pollen

Facebook X LinkedIn Email
Collaborating scientists from the Helmholtz Centre for Environmental Research (UFZ) and the German Centre for Integrative Biodiversity Research (iDiv) developed an image-based particle analysis system, reliant upon laser light and AI, to automate the process of pollen analysis. The method combines flow cytometry and deep learning to create a tool capable of correctly identifying species and quantifying pollen grains in a sample.

Pollen analysis delivers important ecological and economical information, which is a useful component in many areas of research, including that which aims to study climate change and the loss of species by examining the relationship between plants and pollinators. Though microscopy remains the standard technique for pollen analysis, it requires instrumentational and technical expertise and is a time-consuming method.

In the new method, scientists led by Susanne Dunker, head of the working group on imaging flow cytometry at the Department for Physiological Diversity at UFZ and iDiv, added a pollen sample to a carrier liquid. The liquid containing the sample then flows through a channel that increasingly narrows, causing the individual pollen grains to separate and assemble linearly, as if attached on a string of pearls. Each grain individually passes through the built-in microscope element.

The system is ultimately able to capture images of up to 2000 individual pollen grains per second; two microscopic images are taken, plus an additional 10 fluorescent microscopic images per grain.

Microscopic images from pollen, which are important for pollinators, obtained by image-based particle analysis. Each row shows a single pollen grain of a specific plant with a normal microscopic image (first image on the left) and fluorescence images for different spectral ranges (colored images on the right). Courtesy of Susanne Dunker.
Microscopic images from pollen, obtained by image-based particle analysis. Each row shows a single pollen grain of a specific plant with a normal microscopic image (first image on the left) and fluorescence images for different spectral ranges (colored images on the right). Courtesy of Susanne Dunker.
Using laser light at varying wavelengths, the scientists were able to systematically excite the pollen grains to themselves emit light.

“The area of the color spectrum in which the pollen fluoresces, and at which precise location, is sometimes very specific. This information provides us with additional traits that can help identify the individual plant species," Dunker said.

In the deep learning component of the method, an algorithm works in successive steps to abstract an image’s original pixels to an increasingly greater degree, eventually succeeding in extracting and presenting species-specific characteristics.

Where purely microscopic analysis of even a straightforward sample may take four hours, the particle analysis system achieves the same result in approximately 20 minutes.

Microscopic imaging, fluorescence characteristics, and imaging flow cytometry’s high throughput have not previously combined in a single analytical procedure for the detailed study of pollen, Dunker said.

In testing, researchers examined 35 plant species and made 430,000 unique images to establish a data set. They reached 96% accuracy, reliably identifying species for which microscopic imaging alone may not consistently deliver identifying characteristics.

“We are now able to evaluate pollen samples on a large scale, both qualitatively and, at the same time, quantitatively. We are constantly expanding our pollen data set of insect-pollinated plants for that purpose," Dunker said. The team aims to expand its data set to include at least 500 plant species whose pollen is a significant source of food for honeybees, to answer questions and launch research related to plant/pollinator interactions.

UFZ applied for a patent for the method, and Dunker received the UFZ Technology Transfer Award for the work last year. The research was published in New Phytologist (


Published: October 2020
flow cytometry
Flow cytometry is a powerful technique used in biology and medicine for the quantitative analysis of the physical and chemical characteristics of cells and particles suspended in a fluid. The method allows for the rapid measurement of multiple parameters simultaneously on a cell-by-cell basis. It is widely used in various fields, including immunology, microbiology, hematology, and cancer research. Here are the key components and features of flow cytometry: Sample preparation: Cells or...
deep learning
Deep learning is a subset of machine learning that involves the use of artificial neural networks to model and solve complex problems. The term "deep" in deep learning refers to the use of deep neural networks, which are neural networks with multiple layers (deep architectures). These networks, often called deep neural networks or deep neural architectures, have the ability to automatically learn hierarchical representations of data. Key concepts and components of deep learning include: ...
fluorescence microscopy
Fluorescence microscopy is a specialized optical imaging technique used in biology, chemistry, and materials science to visualize and study specimens that exhibit fluorescence. Fluorescence is the phenomenon where a substance absorbs light at one wavelength and emits light at a longer wavelength. In fluorescence microscopy, fluorescent dyes or proteins are used to label specific structures or molecules within a sample. The basic principles of fluorescence microscopy involve illuminating the...
Imagingflow cytometryHelmholtz Center for Environmental Researchimage-based analysis of plantsAIdeep learningfluorescence microscopyMicroscopyeducationResearch & TechnologyLasersagricultureenvironmentBioScan

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.