Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Molecular Plant Pathology Laboratory » Research » Publications at this Location » Publication #420626
Research Project: Omics-Based Approach to Detection, Identification, and Systematics of Plant Pathogenic Phytoplasmas and Spiroplasmas

Location: Molecular Plant Pathology Laboratory

Title: Leveraging artificial intelligence and big data to advance Phytoplasma disease detection and crop health management

Author
item Wei, Wei
item Shao, Jonathan
item ZHAO, YAN - Retired ARS Employee

Submitted to: Phytopathogenic Mollicutes
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/15/2024
Publication Date: 3/5/2025
Citation: Wei, W., Shao, J.Y., Zhao, Y. 2025. Leveraging artificial intelligence and big data to advance Phytoplasma disease detection and crop health management. Phytopathogenic Mollicutes. 15(1):15-16. https://doi.org/10.5958/2249-4677.2025.00007.6.
DOI: https://doi.org/10.5958/2249-4677.2025.00007.6

Interpretive Summary: Phytoplasmas are small, unculturable bacteria that infect a wide range of economically significant crops, causing substantial global agricultural losses. Traditional diagnostic methods are slow, require specialized expertise, and often hinder timely intervention, highlighting the need for more efficient, accessible solutions for farmers and growers. To address these challenges, ARS scientists in Beltsville, Maryland, have developed two complementary approaches: an AI-based diagnostic system using Convolutional Neural Networks (CNNs) and a comprehensive online image and symptom database. The AI diagnostic system, trained on extensive image datasets, provides rapid, accurate phytoplasma infection detection. Meanwhile, the image and symptom database was constructed to offer an accessible tool for farmers to visually compare crop symptoms for early disease management while the AI model undergoes further optimization. Ultimately, these tools aim to be integrated to provide real-time, in-field support, empowering farmers and agricultural professionals with advanced detection methods and proactive crop management solutions. This study equips farmers and growers with accessible disease management tools and serves as a valuable resource for researchers, students, and agricultural specialists focused on advancing plant pathology and precision agriculture.

Technical Abstract: Phytoplasmas are minute, cell wall-less bacteria that infect various economically important crops worldwide, leading to significant agricultural losses. Traditional diagnostic methods are often time-consuming, require specialized expertise, and thus limit timely intervention. This study addressed these challenges by combining two complementary approaches. The first is devising an AI-based diagnostic system utilizing Convolutional Neural Networks (CNNs) trained on extensive image datasets, enabling rapid, accurate detection of phytoplasma infections with promising accuracy. The second approach involves the construction of a comprehensive image and symptom database that allows farmers to compare crop symptoms for early disease identification. The image database serves as an initial online-accessible tool for proactive crop management while the AI model is being further optimized. Eventually, the AI diagnostic model and image database will be integrated, forming a scalable, powerful solution aimed at advancing phytoplasma disease management and enhancing crop protection strategies.