How a Canadian research hospital is working with Watson to help find new treatments for Parkinson’s disease.

How a Canadian research hospital is working with Watson to help find new treatments for Parkinson’s disease.

Client:  UHN
Industry:  Healthcare


The University Health Network is using Watson to help them find new treatments for Parkinson’s disease.

There is no known cure for Parkinson’s disease. Researchers have spent years studying this devastating disease, yet, the most common treatment was brought to market in early 60s.

The average Parkinson’s patient is diagnosed in their sixties or seventies. Jonathan Rezek received his diagnosis in his fifties. The disease runs in his family. As a Business Development Executive for IBM Canada’s National Innovation Team, Mr. Rezek was used to finding ways to help solve problems using technology. Knowing that IBM Watson had already been used for medical research, Mr. Rezek broached the subject with his neurologist, Dr. Connie Marras. Dr. Marras in turn brought the idea to her research colleague Dr. Naomi Visanji at the Edmund J. Safra Program in Parkinson’s disease and Morton and Gloria Shulman Movement Disorders Centre at the University Health Network.

Enter Watson

While there is promising research on the horizon, new drugs can take decades to identify, study, develop and bring to market. For this reason, Dr. Visanji and her team were interested in evaluating whether existing drugs used for other diseases, could be repurposed to help treat those with Parkinson’s. The problem: The task of efficiently sifting through mountains of research for the thousands of drugs currently available to determine if a drug could be a potential research target is nothing short of a Herculean effort.

Enter Watson. After being trained to understand the task at hand, Watson for Drug Discovery was able to comb through and analyze millions of pages of medical abstracts and create a ranked list of potentially suitable candidate drugs for the researchers’ consideration—in a fraction of the time compared to manual searches. Now, Dr. Visanji and her team are using these insights and moving forward with lab testing for the most promising candidate drugs.

Though there is still plenty of work ahead for the team at the University Health Network, they are excited to research these candidate drugs. They believe that – aided by Watson – they will bring new hope to patients like Jonathan Rezek.

How Watson learned in 5 steps:

  1. Watson was trained to understand the problem: Parkinson’s patients use a drug treatment that helps them, but eventually causes dyskinesia.

  2. To establish the criteria for suitable drug treatments, Watson was given a list of the drugs that have shown some promise with dyskinesia in animal studies and clinical trials.

  3. Watson was given a list of over 3,500 candidate drugs and was tasked with reading more than 20 million abstracts of related scientific studies.

  4. In about 30 minutes, Watson sifted through millions of pages of scientific literature, absorbed patterns, and drew parallels between related information.

  5. Watson created a ranked list of candidate drugs, of which 5 have been prioritized for further study.

  6. Incorporating Watson into the research process

    The Team at the University Health Network received a grant from the Canadian Institutes of Health Research (CIHR) in January 2018. The grant covers the costs of testing the top five candidate drugs that Watson ranked and researchers determined were worthy of further evaluation. The team believes Watson has helped them usher in a new chapter in drug research for Parkinson's disease.

    Can AI accelerate a medical breakthrough? | TED Institute

    Dr. Visanji and her team at the University Health Network worked with Watson to find available drugs that could potentially be repurposed to help treat Parkinson’s disease. Now, Dr. Visanji and her team can focus on taking only the most promising candidate drugs to further study.

    While their work is far from done, Watson helped researchers accelerate the journey toward their goal. As Jonathan Rezek, the IBMer and Parkinson’s patient who brought Watson to the attention of Dr. Marras put it, “Parkinson’s is a really slow moving disease. It’s hard to do research on it. So, anything you can do to make research go faster is a positive.”

Related content

Watson for Drug Discovery is a cloud-based, AI-enabled solution that analyzes scientific knowledge and data to reveal known and hidden connections that can help increase the likelihood of scientific breakthroughs.

Learn more →

Start your 30 Day Free Trial today and get the best answers to your key research questions with the power of AI.

Learn more →