Harnessing the Power of Artificial Intelligence in Veterinary Medicine
December 04, 2023
You likely already use artificial intelligence (AI) for everyday tasks. It is used to optimize the spam filter in your email account, help your phone predict text while you type and recommend products when online shopping. But how might AI come into play when you take your pet to the veterinarian? What if AI could give pet owners peace of mind and help predict the course of an animal’s disease?
OVC radiologist and assistant professor Dr. Ryan Appleby believes that AI could be used for so much more than those everyday tasks, specifically in veterinary diagnostic imaging.
AI is already a regular tool in human medicine, and it is now making its way into veterinary medicine. From helping the radiologist or technician position patients for more accurate radiographs, to acting as a second eye for practitioners by identifying abnormalities in radiographs through computer-aided diagnosis – Appleby believes that AI could revolutionize veterinary care as we know it.
When a patient at the OVC Health Sciences Centre (HSC) requires diagnostic imaging – radiographs (X-rays), ultrasound, CT or MRI – their tests are performed, and the resulting images are sent to the OVC Diagnostic Imaging team for interpretation. The team is comprised of board-certified radiologists – veterinarians with specialized post-doctoral training in diagnostic imaging, like Appleby– diagnostic imaging residents, and student veterinarians. Radiologists oversee all activity on their service, as well as taking on a teaching role for student veterinarians and residents.
On a typical day, Appleby and his team begin with rounds at 8:00 a.m. This is when the team meets to discuss particularly interesting images from clinical cases, or review literature on important topics in diagnostic imaging. By 9:30 a.m., the team begins to see patients. Radiology technicians acquire diagnostic images and send them to a resident – veterinarians that are completing postdoctoral training to become board-certified in veterinary radiology – to interpret. Residents write a report on the case, which is then sent to the overseeing radiologist for review. The report is then finalized and sent to the requesting veterinarian to inform diagnosis and treatment. with the radiology team at the OVC HSC supports the client facing services such as internal medicine, surgery, emergency and critical care, oncology, neurology and others. In addition to his role as a clinician, Appleby is also a professor and researcher at OVC. His research program focuses on artificial intelligence (AI) – computer programs that can do tasks which usually require human intelligence – and diagnostic imaging.
As Appleby reviews one of the many reports from his resident for today’s caseload, he shares some of the potential applications AI may have in his field.
“My current research aims to better understand how artificial intelligence can be safely implemented in the field of veterinary diagnostic imaging,” Appleby says. “The benefit of using AI in the field of veterinary radiology is the potential to improve quality of care and extend the reach of high-quality diagnostics. With AI-based practices, we will be able to better diagnose, predict and treat a variety of illnesses in animals.”
For example, graduate students in Dr. Appleby’s lab are working to develop AI that can assist in capturing X-rays. This would provide technicians with reliable guidance when taking an X-ray of an animal, which could reduce the number of re-takes needed. This means that pets could require less time under sedation and decreased exposure to radiation. It could also mean that X-rays would be of higher quality, reducing the likelihood of interpretation errors by the veterinarian or radiologist.
This is just one potential application of AI in the field of diagnostic imaging; Appleby believes that AI will be implemented in a wide variety of areas across veterinary medicine in the future.
“I expect that we will see AI being used in the near future to assess medical records, make writing medical records faster, assist in drug development – and likely many other tasks that I cannot even envision at this time,” he says.
Appleby believes that a combination of research and multi-institutional collaboration is needed to better develop our understanding and trust for AI’s responsible use in practice.
“If AI is developed properly with the appropriate ethical safeguards, we could see these systems used to help veterinarians identify diseases and determine the best treatment plan – supporting veterinarians by helping them make better decisions to improve pet health,” Appleby advocates. “These benefits could even extend to the animal’s owner by providing the most accurate information on the likely course of their pet’s disease, based on a combination of diagnostic images, laboratory results and medical record data.”
As amazing as these benefits are, Appleby cautions that AI development and use in veterinary medicine should be approached thoughtfully.
“The main risk posed to patients and owners is that AI will not be developed well or ethically. If we are not thoughtful in creating these systems, then we could end up where we are not answering the correct questions, or the AI is providing the wrong answer – providing bad information to practitioners and pet owners.”
Appleby acknowledges that there are already some AI systems available for interpretation of X-rays in dogs and cats, but they are largely unproven. He expects that within three to five years, we will see AI systems with better, research-backed accuracy that have potential to be more widely adopted in veterinary medicine.
“I believe that AI may fundamentally change the way we practice, and if done correctly – also improve life for our patients,” he says.
Top three things that Dr. Appleby wants you to know about artificial intelligence in veterinary medicine:
- We are not at the point, nor will we ever be at the point where AI is doing the work without a veterinarian present. There will always be a veterinarian looking at your pet’s diagnostics and involved in their care.
- If you see charges associated with AI diagnostic interpretation on your bill for a veterinary visit, you have a right to ask your veterinarian what this means, and how it was used to assist in the diagnostic process.
- Your pet’s medical records could be used to help develop AI systems for veterinary medicine, but those records will be anonymized by veterinary hospitals and diagnostic companies. Speak to your family veterinarian if you have any questions or concerns.
This article originally appeared in Best Friends magazine (Spring/Summer 2023)