This Technique Can Predict Life-Threatening Bacterial Disease In Dogs: Study

Scientists at the University of California, Davis, have discovered a technique which can predict Leptospirosis in dogs using artificial intelligence.

Representational Image | The lead author on the paper said traditional testing for Leptospira lacks sensitivity early in the disease process, and detection also can take more than two weeks. The AI model eliminates the two roadblocks to a swift and accurate diagnosis.

Source : Getty

New Delhi: Leptospirosis is a disease that dogs can get from drinking water contaminated with Leptospira bacteria, and causes kidney failure, liver disease, and severe bleeding into the lungs of the infected dog. Since Leptospirosis is life-threatening, early detection of the disease is very important.

Scientists at the University of California, Davis, have discovered a technique which can predict Leptospirosis in dogs using artificial intelligence. The scientists, after several months of testing different models, developed one which outperformed traditional testing methods and provided accurate early detection of Leptospirosis. 

The study, describing the technique, was recently published in the Journal of Veterinary Diagnostic Investigation.

What Makes The New Technique Special?

In a statement released by University of California, Davis, Krystle Reagan, the lead author on the paper, said traditional testing for Leptospira lacks sensitivity early in the disease process, and detection also can take more than two weeks because of the need to demonstrate a rise in the level of antibodies in a blood sample. The AI model developed by the team eliminates the two roadblocks to a swift and accurate diagnosis.

Historical data of patients at University of California Davis Veterinary Medical Teaching Hospital that had been tested for Leptospirosis were analysed as part of the study. The AI prediction model was trained with the help of routinely collected blood from 413 dogs. 

An additional 53 dogs with suspected Leptospirosis were treated over the next year. As many as nine dogs that were positive for Leptospirosis were correctly identified with the model. In other words, the AI model was found to have 100 per cent sensitivity.

According to the study, the model also correctly identified approximately 90 per cent of the 44 dogs that were ultimately Leptospirosis negative.

The researchers want the model to become an online resource for veterinarians to enter patient data and receive a timely prediction. 

In the same release, School of Veterinary Medicine Dean Mark Stetter said that AI-based, clinical decision making is going to be the future for many aspects of veterinary medicine. 

How The Detection Model Could Help Humans

Leptospirosis is a life-threatening zoonotic disease, which means it can spread from animals to humans. Since it is difficult to  diagnose the disease in people, Reagan hopes that the technology behind this groundbreaking detection model has translational ability into human medicine.

She said her hope is that the technology will be able to recognise cases of Leptospirosis in real time, giving clinicians and owners important information about the disease process and prognosis. She added that the researchers hope to apply AI methods to improve the ability to quickly diagnose other types of infections.