University of Surrey and APHA Researchers Develop World's First AI System for Rapid Rabies Testing

Phys.org Science•Published: Oct 12, 2026 05:30 PM IST

Researchers at the University of Surrey and the Animal and Plant Health Agency have created the world's first artificial intelligence system designed to detect the rabies virus in cell cultures and measure antibody concentrations. Published in Scientific Reports, the breakthrough deep learning model achieved a 99.6% accuracy rate in classifying samples while eliminating human interpretation errors.

AI Model Performance: Trained on more than 2,300 fluorescence images featuring over 36,000 annotated sections, the top-performing deep learning model accurately distinguished infected cells from uninfected ones without false alarms.
Streamlining Diagnostics: The new automated workflow removes the slow, labor-intensive process of specialists manually examining thousands of microscope slides in endemic regions.
Broader Pathogen Framework: Researchers noted that the underlying technology can be adapted to analyze other infectious diseases and is currently being expanded to detect the virus in brain tissue.

The Context & Background

Despite being nearly 100% preventable via vaccination, rabies continues to claim approximately 59,000 human lives annually across the globe. Traditional diagnosis and antibody quantification heavily rely on scarce specialist expertise, making timely surveillance difficult in nations where the disease remains endemic.

Why it matters:This breakthrough provides laboratories in resource-limited and endemic countries with rapid, consistent diagnostic tools to curb a disease that kills tens of thousands annually. By automating complex cellular analysis, the technology could drastically improve global surveillance and pet travel testing while establishing a scalable framework for diagnosing other deadly pathogens.

Enjoyed this 2-minute breakdown?

Share it with your colleagues and friends!