FaceMatch.org.au: using computer vision to improve genomic diagnosis for people with a syndromic form of intellectual disability

Dr Tracy Dudding-Byth1,2,3, Dr Carlos Riveros4, Ms Jackie Boyle1,2, Dr Anna Hackett1,2,5, Dr Ann Baxter1,2,5, Ms Sheridan O’Donnell1, Dr Steve Brain6, Professor John Attia5

1Hunter New England Area Health Service, Newcastle, Australia, 2NSW Genetics of Learning Disability Service , Newcastle, Australia, 3University of Newcastle GrowUpWell PRC, Newcastle, Australia, 4Hunter Medical Research Institute , Newcastle, Australia, 5University of Newcastle , Newcastle, Australia, 6Imagus Technology , Brisbane, Australia

Background
Despite advanced genetic testing, 60% of children with early onset intellectual disability [ID] never receive a molecular diagnosis, and there are an estimated 2000 ID genes still to be discovered. As half of these people have facial features which may provide a clue to diagnosis, it is our hope that our FaceMatch technology will dramatically lower this statistic by becoming the first tool medical professionals reach for to fast-track an accurate diagnosis of an intellectual disability.

Method
In the past 18 months, our team has developed the initial iteration of the FaceMatch platform – the first technology of its kind to empower parents, in partnership with their doctors, to proactively participate in the diagnosis of their child by providing information, editing and updating their child’s profile page and images into a global database used for face-matching.

Findings
We will present our pilot data showing that within 10 syndromes, FaceMatch outperformed senior clinical geneticists on phenotyping tasks.

Implications
Currently, when a diagnosis of an intellectual disability cannot be made by genetic testing, clinicians rely on manual face matching techniques, like comparing images at conferences or entering written descriptions accompanied by DNA variant data into phenotyping databases, hoping to locate another individual with similar features. FaceMatch is a pioneering health care solution that not only provides a means for effectively and efficiently comparing facial images and clinical features, but also eases years of heartache and uncertainty for families and parents searching for a diagnosis.


Biography:

Dr Dudding-Byth is a clinical geneticist within the NSW Genetics of Learning Disability Service; a Hunter New England Clinical Research Fellow and a founding co-director of the National Rare Disease Advocacy group Rare Voices Australia. Tracy is leading the international FaceMatch Project which uses computer face-matching technology to help improve the rate of molecular diagnosis for people with a syndromic form of intellectual disability.