AI System Achieves 99.6% Accuracy in Rabies Virus Detection, Transforming Global Testing

AI System Transforms Rabies Virus Detection in Cell Cultures

Researchers at the University of Surrey's Institute for People-Centred AI, working alongside the Animal and Plant Health Agency (APHA), have developed the first AI system capable of detecting rabies virus in cell cultures with 99.6% accuracy1

1

2

2

. Published in Nature Scientific Reports, this DEFRA-funded breakthrough addresses critical gaps in diagnosing one of the world's deadliest infectious diseases, which claims approximately 59,000 human lives annually despite being almost entirely preventable through vaccination1

1

.

The AI-based workflow for detecting rabies replaces the current labor-intensive process where specialists manually examine thousands of microscope images. This manual microscopy approach is slow, requires significant expertise, and often leads to inconsistent interpretations across different laboratories2

2

. Joan Amaya-Cuesta, PhD researcher at the University of Surrey and Rabies and Viral Zoonoses Research Scientist at APHA, emphasized the urgency: "Once clinical signs appear, it is almost always fatal, so prevention and early diagnosis are essential"1

1

.

Deep Learning Models Analyze Fluorescence Images at Scale

The research team trained several deep learning models using more than 2,300 fluorescence images containing over 36,000 annotated image sections1

1

. The best-performing model not only achieved 99.6% classification accuracy but also distinguished infected from uninfected cells with near-perfect precision, effectively eliminating false alarms that could compromise diagnostic reliability2

2

.

This AI system for rabies testing can accurately measure vaccine-induced antibody levels, a critical function for assessing protective immunity and meeting pet travel requirements across international borders1

1

. Professor Miroslaw Bober, Professor of Video Processing and Machine Intelligence at the University of Surrey, noted that the system performs "complex image analysis quickly, consistently and at scale"2

2

.

Accelerating Diagnosis in Endemic Countries

The technology holds particular promise for diagnostic laboratories in countries where rabies remains endemic and specialist expertise is scarce. By providing faster and more consistent sample analysis, the AI-driven diagnostic tools could accelerate rabies virus detection where timely diagnosis remains challenging1

1

. The workflow reduces reliance on manual interpretation, making advanced diagnostic capabilities accessible to laboratories that previously lacked the resources or trained personnel to conduct comprehensive testing2

2

.

Dr. Lorraine M. McElhinney, Head of the National Reference Laboratory for Human and Animal Rabies at APHA, revealed the next phase: "This technology is now being further developed for use to detect the virus in infected brain tissue which we hope will transform international rabies diagnostic laboratories"1

1

. This expansion could enable post-mortem diagnosis directly from brain samples, a standard requirement for confirming rabies cases.

Framework Extends Beyond Rabies to Other Infectious Diseases

The research establishes a framework adaptable for detecting and analyzing other pathogens that rely on similar laboratory testing methods2

2

. Professor Bober highlighted this broader potential: "The same framework could be applied to other infectious diseases that rely on similar laboratory tests, helping researchers and diagnostic laboratories make better use of AI in the future"1

1

.

This adaptability positions the AI system as a template for addressing diagnostic challenges across multiple infectious diseases, potentially accelerating global health responses beyond rabies. The team hopes this innovation will strengthen rabies surveillance and support global rabies elimination efforts, moving closer to the goal of eliminating human deaths from this preventable disease2

2

.