Next generation of artificial intelligence for veterinarians showcased

Imagine a robotic employee who can observe, interpret, evaluate, and offer a set of possibilities from a large body of veterinary-industry data for decision-making? Not only that, but he/she can do it at the speed of light.

Meet Watson, a cognitive computing tool being showcased at IBM’s World of Watson conference, May 5-6, in New York City. But Watson is not alone. LifeLearn’s Sofie application will also be there, and its specialty is the veterinary industry. (James Carroll, president and CEO of LifeLearn, will be sharing the application at the conference.)

And there’s reason to pay attention.

Eighty percent of data today is unstructured, derived from literature, art, research, blogs, posts, and tweets. Generated by humans for humans, it is, well, messy.

But Sofie can handle that because Sofie is powered by Watson’s cognitive computing algorithms, and trained by veterinarians for veterinarians. Like Watson, Sofie thinks differently.

Unlike traditional computing that looks for keywords and matches, Sofie looks at language. It breaks it down—its grammar, context and culture—to surface the author’s real intent. And then it offers veterinarians a set of possibilities for overcoming their patients’ challenges.

In other words, Sofie cognitively processes information like an expert.

And Sofie’s knowledge base is broad and deep. That knowledge base includes data from across the veterinary industry, including publishers Wiley and Elsevier, as well as LifeLearn, Merck, VetStream, VetFolio, the American Animal Hospital Association (AAHA), Western Veterinary Conference (WVC), and North American Veterinary Conference (NAVC).

But that’s not all. Like humans, Sofie continues to learn, adapt, and get smarter. Its machine learning enables it to, for example, answer questions it wasn’t initially trained on, such as, “Are there acupuncture points to treat coughing?” 

How does Sofie work? Veterinarians simply access the “Ask Sofie” tab and type in a question based on initial patient consultations. That’s when Sofie goes to work. In seconds, it generates focused, evidence-based treatment options.

According to LifeLearn, Sofie functions differently than regular search engines. It has been trained to answer direct questions rather than open-ended ones, and full sentences rather than search terms.

So rather than inputting “epinephrine dose,” for example, asking “What is the epinephrine dose for a patient in shock?” will return more relevant results.  

The app also features support tabs for case overviews, differentials, diagnostics, treatment plans, and overall prognosis to help veterinarians manage each patient case.

Needless to say, Sofie is getting job offers.

Thanks to LifeLearn’s early adopter program, veterinarians are signing Sofie on to work for them. To-date, 210 veterinarians and staff members across 57 practices are using Sofie.

Heidi L Fritz, DVM, CVA, owner of Aberdeen Veterinary Clinic, was one of the first veterinarians to use Sofie and noted its ease. “Instead of pulling out a few different texts and having them on my desk, I just have a few windows open on my desktop.

“Sofie will go a long way toward strengthening the collective intelligence of the veterinary industry with evidence-based medicine accessible at our fingertips within seconds."

“During the time I’ve used Sofie,” said Troy Bearden, DVM, of AAHA-accredited practice Shallowford Animal Hospital, who joined Carroll in New York, “I have gained new insights on diagnostics and treatments of common diseases of dogs and cats.

“This new technology is unlike anything I've seen before and I believe it will soon become my most valuable tool toward improving patient care.”