Scientists at the University of Exeter and Swansea University are developing a system that combines rapid imaging with artificial intelligence to analyse pollen more quickly and accurately. By identifying and categorizing pollen grains from sediment cores or air samples, scientists can understand which plant species were present at different points in history, potentially dating back millions of years. Traditionally, counting pollen types under a microscope has been a time-consuming task. The new system uses imaging flow cytometry to capture pollen images and artificial intelligence based on deep learning to identify different types of pollen, even when the sample is imperfect. The researchers have already used the system to rapidly classify over a thousand pollen grains from a 5,500-year-old sediment core, completing the task in under an hour. The team hopes that this technology will not only provide a more comprehensive picture of past flora and climate change but also improve pollen readings in today’s environment, potentially helping hayfever sufferers mitigate symptoms. The system is also able to deal with poor quality images and categorize pollen that is not included in training libraries, making it more versatile and effective. The research is supported by the National Environment Research Council (NERC) and the US National Institutes of Health.
>>Join our Facebook Group be part of community. <<