New software that analyses 3D scans of peoples' faces could speed up the diagnosis of rare genetic conditions, say UK researchers. A team based at the Institute of Child Health in London has collected scans of children with known conditions, as well as unaffected children, creating a database of 'average faces' against which new patients can be compared. The technique has a 90 per cent success rate, say the scientists, who presented their findings at the British Association Festival of Science in York.
The system can already diagnose around ten conditions, and will soon be able to detect more than 20 others. It will provide geneticists with a powerful new tool for helping diagnose conditions such as Fragile X syndrome - the most common cause of inherited learning difficulties. While individuals with Down syndrome have recognisable facial features, other conditions can alter the appearance in more subtle ways that can be difficult to spot.
Professor Peter Hammond, a computer scientist who carried out the research, told the Times newspaper: 'If you've got a child with an unknown condition, you can screen the face and reduce the number of genetic tests you need', adding 'for the NHS, the benefit would be that you are doing much more focussed genetic testing, and these tests can cost in the order of £500 to £1000 apiece'.
Professor Hammond also pointed out that earlier diagnosis means better care from an earlier age - for example, children with Costello syndrome are prone to tumours, and so need regular scans. And, he says, 'perhaps the biggest advantage is for the parents. It's very stressful not to know what your child has got'. An accurate diagnosis means that the family can be told about the prognosis, and offered genetic counselling if they want to know the risk to any future children they might have.
The system is currently only being used in London, but Professor Hammond would like to see it rolled out across the UK, once more 3D cameras are available. He also wants to compile more data, for example, by including average faces for different sexes and ethnicities affected.