AI-Powered Microscope Revolutionizes Soil Health Testing for Farmers

This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

Validation image outputs from the fungi-detection neural network (images not used for training); (left: microscope image, middle: expert human label, right: neural network prediction). Credit: UTSA

The classic microscope is getting a modern twist—US researchers are developing an AI-powered microscope system that could make soil health testing faster, cheaper, and more accessible to farmers and land managers around the world.

Researchers at The University of Texas at San Antonio, U.S., have successfully combined low-cost optical microscopy with machine learning to measure the presence and quantity of fungi in soil samples. Their early-stage proof-of-concept technology is presented at the Goldschmidt Conference in Prague on Wednesday 9 July.

Determining the abundance and diversity of soil fungi can provide valuable insights into soil health and fertility, as fungi play essential roles in the biogeochemical cycling of nutrients, water retention, and plant growth. With this knowledge, farmers can optimize crop production and sustainability by making informed decisions about soil management, including fertilizer application, irrigation, and tillage.

Optical microscopes are the oldest design of microscope and have long been used to discover and identify tiny organisms in the soil. Other forms of soil testing use techniques like phospholipid fatty acid testing and DNA analysis to detect organisms, or to measure the presence of chemicals such as nitrogen, phosphorus and potassium. While powerful, these modern methods tend to be costly or just emphasize chemical composition, often overlooking the full biological complexity of soil ecosystems.

Alec Graves from The University of Texas at San Antonio College of Sciences, U.S., is presenting the research at the Goldschmidt Conference this week. He said, \"Current forms of biological soil analysis are limited, requiring either expensive laboratory equipment to measure molecular composition or an expert to identify organisms by sight using laboratory microscopes. Comprehensive soil testing isn't widely accessible to farmers and land managers, who need to understand how agricultural practices impact soil health.

\"Using machine learning algorithms and an optical microscope, we're creating a low-cost solution for soil testing that reduces the labor and expertise required, while providing a more complete picture of soil biology.\"

In their early-stage design, the researchers built and tested a machine learning algorithm to detect fungal biomass in soil samples, incorporating this into custom software for labeling microscope images. This was created using a dataset of several thousand images of fungi from soils across South Central Texas. The software works with just 100x and 400x total microscope magnification, available in many affordable off-the-shelf microscopes, including those found in school laboratories.

\"Our technique analyzes a video of a soil sample, breaking this into images, and uses a neural network to identify and quantify fungi,\" says Graves. \"Our proof-of-concept can already detect fungal strands in diluted samples and estimate fungal biomass.\"

The team is now working to integrate their technique into a mobile robotic platform for detecting fungi in the soil. The system will combine sample collection, microphotography and analysis into a single device. They aim to have a fully developed, deployable device ready for testing within the next two years.

The research is led by Professor Saugata Datta, Director of Institute of Water Research Sustainability and Policy at UTSA. Details of the machine learning algorithm are due to be published in a peer-reviewed journal later this year.

More information: Machine Learning Microscopy Analysis for Rapid Sample Biogeochemical Assessment: Applications from Agricultural Soils to Exobiology. Goldschmidt Conference.