AI Algorithm Identifies Children with Rare Genetic Disorders

In a groundbreaking development, researchers at the Children’s Hospital of Eastern Ontario (CHEO) have created an artificial intelligence algorithm that can identify children with rare, undiagnosed genetic disorders. By analyzing nearly 260,000 patient records from the past decade, the AI algorithm uncovered 15 cases that had previously gone undiagnosed. Dr. Kym Boycott, senior scientist at the CHEO Research Institute, described the algorithm as a way to identify children with rare genetic diseases earlier in their diagnostic journey and prevent the frustration of misdiagnosis.

Rare genetic disorders are often complex and present with symptoms that can mimic other medical conditions, leading to misdiagnosis and delayed treatment. With approximately 6,000 rare genetic diseases, some of which are found in as few as 1 in 5 million people, the ability to accurately identify and diagnose these conditions is crucial.

The CHEO researchers spent three years developing the ThinkRare algorithm, which has the ability to pinpoint likely cases of rare genetic disorders much faster than traditional methods. Dr. Boycott noted that without the algorithm, some children may never have been referred to the hospital’s genetic disease clinic for proper care.

One such case is that of Antony Wistaff, a 10-year-old boy whose rare disorder was identified by the algorithm. Antony had undergone multiple surgeries for hydrocephalus, a neurological disorder, but his developmental delays and difficulties retaining information remained unexplained. Over the course of a decade, Antony visited CHEO numerous times across various specialty clinics, with no answers to his condition. It wasn’t until August of the previous year that the ThinkRare algorithm flagged Antony as a potential case, leading to genome sequence testing. In February 2024, Antony was diagnosed with Chung-Jansen Syndrome, a rare disorder that affects only about 400 people worldwide. His parents expressed relief at finally having answers and being able to access the necessary treatment and support for their son.

The ThinkRare algorithm not only aids in clinical diagnosis and targeted treatment but also provides valuable resources for families. Dr. Boycott explained that a specific diagnosis opens pathways in both the healthcare system and the community. Without a diagnosis, families often struggle to access necessary resources, such as special education assistants in schools.

The creator of the ThinkRare algorithm, Ivan Terekhov, CHEO’s director of research informatics and technology, was deeply moved by the impact his creation had on patient care. He emphasized that AI should not be seen as a replacement for doctors but as a tool to assist them. Terekhov believes that AI has the potential to revolutionize the medical field and encourages its responsible use.

Currently, the ThinkRare algorithm has been used to identify past missed cases. However, the plan is to expand its use to sift through the records of all current and new patients at CHEO. While CHEO may be the first hospital to utilize this AI diagnostic tool, there are no intentions to monetize it. Dr. Boycott and her team aim to provide the algorithm to pediatric hospitals worldwide free of charge. Their hope is to facilitate the faster identification and diagnosis of rare genetic disorders in children, ultimately improving their prognosis and quality of life.