Ivan Telekhov and Dr. Kim Boycott speak to Prime Minister Doug Ford about the ThinkRare algorithm at his 50th birthday event (Credit: Andre Coutu, May 17, 2024).
Author: Jamie Long, CHEO
A new world-first artificial intelligence (AI) algorithm called ThinkRare has helped CHEO diagnose children and young people with rare diseases they would never have known about otherwise. And now the experts behind this groundbreaking research are ready to help children and families around the world.
CHEO researchers have developed a search algorithm that harnesses the power of AI to identify children and young people who may have undiagnosed rare genetic diseases and request genetic testing. . This brought an end to their diagnostic journey.
In late 2023, ThinkRare AI algorithms began scanning patient files dating back to 2018 to retrospectively identify patients who may have undiagnosed rare diseases.
Then, as of early 2024, the algorithm will begin running through CHEO’s current and new patient electronic medical records on a daily basis, typically identifying 5 to 10 patients each month who may require additional genetic testing. I did.
“It’s not reasonably possible for a human being to go through 260,000 medical records a day and try to find these patients. It’s like a needle in a haystack,” says CHEO. said Ivan Telekhov, Director of Research Informatics, AI and Technology at the Research Institute (RI).
“This algorithm is a game changer,” said Dr. Kim Boycott, senior research scientist at CHEO RI, chief of genetics at CHEO, and head of the ThinkRare team.
After a months-long process that included referrals from health care providers, CHEO provided eight children with a complete genetic test called genome-wide sequencing.
Six of these children, or three in four, have since been diagnosed with a rare disorder.
On average, this type of test results in a diagnosis in one in three children, demonstrating that the ThinkRare algorithm is identifying appropriate patients.
“We were really surprised by such a high success rate,” said Alexandre White-Brown, a genetic counselor at CHEO who works with patients and families identified by the algorithm.
CHEO has so far diagnosed patients between the ages of 3 and 17. There are also children as young as 2 years old who are currently waiting for genetic test results.
In one case, White-Brown said the algorithm helped diagnose a 9-year-old boy and his mother, which later led to the diagnosis of his two siblings as suffering from the same rare genetic disease. said.
Without algorithms, this family would still be searching for answers.
“Rare diseases are difficult to understand. Rare diseases can mimic non-genetic conditions, and non-genetics health care providers do not have all the tools to recognize rare diseases,” White-Brown said. There is a possibility that it is not.”
This diagnosis opens doors for families, and CHEO helps families manage all areas of child care, including advocating for more resources in schools.
The data infrastructure that made the development of ThinkRare possible was created thanks to generous donors to the CHEO Foundation. CHEO is not trying to make money from it.
The ThinkRare team hopes to provide its AI algorithms to other children’s hospitals in Canada and around the world to help children receive life-changing diagnoses sooner and live their best lives.
CHEO also continues to develop and leverage clinical expertise, research and innovation as we move forward with our historic 10-year redevelopment plan.
As part of ongoing efforts, CHEO and its RIs are working to predict which children are likely to return to the emergency department with asthma, improve organ donation and transplantation, and investigate transcription software to improve the patient experience. We are working on several AI-based innovations. visit.
AI is ubiquitous, the ethical, economic, and technological implications are significant, and the transformative potential is incredible.
CHEO, in collaboration with RI, has developed a principles-based approach to guide the responsible use of AI. In addition to considering the risks inherent in AI algorithms and what can be done to mitigate those risks, there are frameworks in place to help.
As a result, CHEO plans to effectively and responsibly implement algorithms like ThinkRare into regular clinical practice.
