Using artificial intelligence in healthcare and medical devices holds the promise of making advanced medicine more accessible to everyone, according to Jacob Jaremko, Amii Fellow and Canada CIFAR AI Chair.
"AI is a way we can democratize health expertise," Jaremko, who is also the founder of medical imaging startup Medo.ai, tells the Ask AI Podcast.
As both a machine learning researcher and a radiologist, Jaremko (Associate Professor, University of Alberta) has seen the ways healthcare can be difficult for people to access. Ultrasound imaging is an example of that. Ultrasounds are one of the most commonly used tools in healthcare, for everything from monitoring a pregnancy to diagnosing organ damage. But getting an ultrasound usually involves going to a hospital or specialized imaging clinic; the training needed to read an ultrasound is very specialized, and technicians are in short supply across the country.
This can lead to backlogs and waiting lists. Jaremko says it can make ultrasounds entirely inaccessible for many people, including unhoused populations and those living in rural and remote communities.
You can use AI to make sense of those fuzzy snowstorm images that the ultrasound produces, then you can take that to anyone. You can bring the hospital to the patient.
Jacob Jaremko, Amii Fellow and Canada CIFAR AI Chair
Jaremko, along with Dornoosh Zonoobi and Jeevesh Kapur, co-founded Medo.ai to use artificial intelligence to address that issue. This week, TD Bank Group (TD) announced that it awarded Jaremko $450,000 to research and test a portable ultrasound system in Alberta.
A portable scanner would take the ultrasound image, which would then be analyzed through an app using artificial intelligence within minutes. The AI app is trained on a collection of thousands of ultrasound images guided by human experts. It allows the user to identify common conditions, leading to faster treatment and freeing up ultrasound technicians to handle more complex cases.
Since the images come from a portable scanner, Jaremko says it opens up access to this vital tool in new places, like rural hospitals and family doctors. It could even be used at ski hills to detect fractures quickly.
"You can use AI to make sense of those fuzzy snowstorm images that the ultrasound produces, then you can take that to anyone. You can bring the hospital to the patient," he says.
Jaremko also shared with Ask AI how his career as a clinician has informed his work in artificial intelligence and the transition of moving from academia to entrepreneurship with the opening of Medo.ai.
Check out the full episode below. And read more about the portable ultrasound project on Folio.