Face-matching technology helps diagnose intellectual disability
The Hunter Medical Research Institute (HMRI) has launched a web platform that will help unravel the mysteries behind undiagnosed intellectual disabilities in children.
Rare genetic conditions are notoriously difficult to diagnose, leaving many parents frustrated by a lack of answers. But with 30–50% of these children sharing facial similarities, clinicians are hoping to utilise these similarities to make earlier detections.
Dr Tracy Dudding-Byth, a senior geneticist at HMRI, has now developed an algorithm that matches common facial traits across ethnicity, age and gender to try to match up undiagnosed patients with intellectual disability. Combining this algorithm with facial recognition software developed by Professor Brian Lovell from The University of Queensland, Dr Dudding-Byth’s FaceMatch project applies a vital medical use for technology normally employed by policing and security agencies.
“I came up with this idea after watching an episode of ABC’s Catalyst on the use of face-matching technology in the passport office,” Dr Dudding-Byth explained. “That’s when I realised this kind of software could be used to match children with undiagnosed intellectual disabilities.”
The platform enables parents to work with their doctor and clinical geneticists to upload a photograph and description of their child, in the hope of matching with another child somewhere in the world. Parents of children with known genetic conditions can also participate to help refine the algorithm.
The Newcastle Permanent Charitable Foundation provided a grant of $143,500 to enable HMRI technicians to build the interactive FaceMatch resource. The chair of the foundation, Phil Neat, said, “This project will provide direct support to families who are currently unable to obtain an accurate diagnosis for their child’s intellectual disability.
“We are proud to invest in this groundbreaking technology, which will ultimately lead to better health outcomes for our local communities but also has the potential to positively impact communities around the world.”
Clinicians traditionally rely on manual forms of face matching by passing around patient photographs at conferences, but FaceMatch eliminates human error, fatigue and delays. Once a match is made, DNA can be collected from each patient and analysed for common traits.
“The DNA analysis of face-matched individuals will hopefully help us unlock some of the mysteries of rare genetic disorders,” Dr Dudding-Byth said. “The more people who participate in the project, the greater our chances of finding a match.”
HMRI Director Professor Michael Nilsson said a diagnosis is important for both parents and children because it can lead to treatment and long-term management.
“One of the challenges faced by researchers is understanding the genetic pathways,” he said. “This project is still in research phase, but we’re hoping it will fill in some of the knowledge gaps for geneticists.”
Parents and doctors can access FaceMatch at https://facematch.org.au/.
Known as Pattern to Knowledge (P2K), the technology uses artificial intelligence to leverage deep...
Stanford scientists have developed an algorithm, called Phrank, that automates the most...
By analysing the brain activity of tetraplegia patients, a new deep-learning algorithm can help...