New AI-based diagnostic tool predicts feline kidney disease—two years in advance

2019-9-11 ox_655x325 RenalTech - blog.jpg

Chronic kidney disease (CKD) is the number-one cause of death in cats older than five, and it’s estimated that it affects around 30% of cats over 12.

It’s a highly complex disease that’s difficult to diagnose—current detection methods find disease only once significant kidney damage has occurred, making intervention and treatment challenging.

Now imagine being able to predict feline CKD in a client’s cat two years before it happens. And doing it by doing what you’re already doing.

“You’re already doing all the lab work,” David Aucoin, DVM, told NEWStat. Aucoin is chief technology officer of Antech Diagnostic and Imaging, and he led the research team that developed the new diagnostic test, called RenalTech.

RenalTech is a new kind of diagnostic tool that blends artificial intelligence (AI), machine learning, and data collected from more than 150,000 feline patients seen in VCA animal hospitals over 20 years to predict CKD in cats two years before onset. Aucoin calls it a “predictive diagnostic tool,” and it’s an incredibly complex algorithm that can predict whether a cat will develop CKD—or not—two years before onset with 95% accuracy.

To create the algorithm, researchers whittled down 35 possible parameters, such as weight, sex, breed, and various lab values, and came up with 6 laboratory parameters—blood urea nitrogen, creatinine, urine specific gravity, urine protein, urine pH, white blood cell count, and age.

And, Aucoin says, veterinarians are already doing the lab work during annual exams, so all you have to do is send the lab results to Antech, which applies the RenalTech algorithm (it’s free), then sends you the results: yes, this cat will get CKD within two years, or no, this cat won’t get CKD within two years.

Aucoin said that’s revolutionary. “We’ve never had the ability to do predictions of either positive or negative. And we’ve certainly never had the confidence in any of these tests that this would actually mean disease versus just an abnormal lab result.”

Aucoin called it a first for veterinary medicine: “RenalTech represents a new paradigm in diagnostics where we are actually looking at the patient’s future and being able to come up with a really, really highly qualified answer as to whether or not this patient will or will not get this disease.”

Aucoin said knowing that a cat is going to get CKD means veterinarians will be able to develop a personalized care plan that will not only help mitigate symptoms and slow the progress of the disease, it will inspire a new level of owner engagement and improve compliance.

Aucoin said RenalTech is the first of several predictive tools that Antech is developing that will use AI to predict disease before it happens.

“We truly believe that there is much more to come in AI and deep learning in veterinary medicine,” he said.  

AI has the potential to transform veterinary care by helping veterinarians shift from disease detection to disease prediction.

Photo credit: © iStock/absoluteimages