Date: 11 September 2007

Facial dysmorphology research at ICH/UCL - further information

The diagnosis of a genetic condition is a complex activity involving a variety of medical professionals, scientists and laboratory staff. As part of the investigation of a patient, a clinical geneticist will undertake a thorough examination. For a child, this will include an evaluation of their physical, cognitive and behavioural development. The physical assessment includes close scrutiny of facial features because they are altered in some way in about 1 in 7 genetic conditions. Thus, the face can be an early clue when trying to determine a diagnosis. Rarity of these conditions makes it difficult to develop the skill of recognising their so-called facial gestalt.

For the past seven years or so, computer scientists at UCL have been developing software that can help in the recognition of facial features that suggest a child or an adult has one of the genetic conditions where facial features are altered.  It is important to note that the software simply recognises patterns of facial features and does not in any way constitute a diagnosis which is undertaken as summarised above and often in conjunction with genetic testing. It works by gathering 3D face images of unaffected and affected children and typically involves computing the average faces of each group and the extent to which individual subjects from the group are expected to vary from the group average. Then pattern recognition algorithms are applied to determine the average rate of success in classifying unseen faces as affected or unaffected. The simplest approach it uses is to classify a face according to its similarity to an average face of the groups being studied and taking into account the expected variation of individuals within those groups.

The software development began about seven years ago at the UCL Eastman Dental Institute in a project led by Professor Peter Hammond. When attending a lecture by the late Professor Robin Winter, a world expert on genetic conditions, he was inspired by Professor Winter’s suggestion that computer technology might be as successful as expert dysmorphologists in recognising facial features altered by an underlying genetic condition. It is continuing now at the UCL Institute of Child Health where Peter Hammond moved early in 2007.

Professor Hammond’s research assistant Dr Tim Hutton, co-supervised by Professor Bernard Buxton of the UCL Department of Computer Science, developed the techniques that form the basis of the software as part of his PhD (Hutton, 2003). Dr Hutton left UCL three years ago but continues to support the research. During the development of the early versions of the software it was tested in a number of genetic conditions in co-operation with clinical geneticists in the UK and USA (Hammond et al, 2004). Since then Professor Hammond has focussed on extending the conditions covered by travelling extensively to collect 3D face scans of unaffected children and adults as well as those with genetic conditions. More recent descriptions of applications of the software have also appeared in the literature (Hammond et al, 2005; Tassabehji et al, 2005; Bhuiyan et al, 2006; Cox-Brinkman et al, 2007). A relatively non-technical description for paediatricians of the techniques of 3D assessment of facial dysmorphology is about to be published by the journal Archives of Diseases in Childhood (Hammond, 2007).

None of this research would have been possible without collaborations with many clinicians and scientists; substantial funding from NewLife (UK), Angelman Syndrome Association (USA), NIH (USA); the co-operation of families with affected children; and, support of the following family syndrome support groups (SSG) and networks:

UK
ASSERT(Angelman SSG), Cornelia de Lange, Cri du Chat SSG, Costello Syndrome SG, Ectodermal Dysplasia Society, Fragile X Society, LMBBS (Bardet-Biedl SSG), Maxappeal (22q11del SSG), Rubinstein-Taybi SSG, Smith-Magenis Syndrome Foundation, Williams Syndrome Foundation, Wolf-Hirschhorn SSG, Worster Drought SSG

Europe
European Chromosome 11 Network (Germany); Angelman SSG (Finland); Rett SSG (Netherlands); Rubinstein-Taybi SSG (Netherlands); Generation 22/22q11del SSG (France)

USA
Angelman Syndrome Association, CFC SSG, Costello syndrome Family Network, National Foundation for Ectodermal Dysplasias, Fragile X Foundation, 11q-/Jacobsen syndrome SG, IFOPA, TNSSG (Noonan SSG), VCFSEF, Williams Syndrome Association, 4p-/Wolf-Hirschhorn SSG

Canada
ASD-CARC.

References for further reading
Hutton TJ, Buxton BF, Hammond P and Potts HWW. IEEE Trans Med Imag 2003;22:747-753.

Hammond P, Hutton TJ, Allanson JE, Campbell LE, Hennekam RC, Holden S et al. 3D analysis of facial morphology. Am J Med Gen 2004;126:339-348.

Hammond P, Hutton TJ, Allanson JE, Buxton B, Campbell L, Clayton-Smith J et al. Discriminating power of localized 3D facial morphology. Am J Hum Gen 2005;77:999-1010.

Tassabehji M, Hammond P, Karmiloff-Smith A, Metcalfe K, Thompson P, Durkin M, et al. GTF2IRD1 in craniofacial development of humans and mice. Science 2005;310:1184-1187.

Bhuiyan Z, Klein M, Hammond P, Mannens MMAM, van Berckelaer-Onnes I, Hennekam RCM. Phenotype-genotype correlations in Cornelia de Lange Syndrome: the Dutch experience. J Med Gen 2006;43:568-575.

Cox-Brinkman J, Vedder A, Hollak C, Richfield L, Mehta A, Orteu A, Wijburg F Hammond P. Three-dimensional face shape in Fabry disease E J Hum Gen 2007;15:535–542.

Hammond P. The use of 3D face shape modelling in dysmorphology. Arch Dis Child 2007; in press.

Peter Hammond
September 2007

Contact information:

GOSH-ICH Press Office: 020 7239 3130
Email: Coxs@gosh.nhs.uk
For genuine and urgent out of hours call speak to switchboard on 020 7405 9200

Notes to editors:

Great Ormond Street Hospital for Children NHS Trust is the country’s leading centre for treating sick children, with the widest range of specialists under one roof.

With the UCL Institute of Child Health, we are the largest centre for paediatric research outside the US and play a key role in training children’s health specialists for the future.

Our charity needs to raise £50 million every year to help rebuild and refurbish Great Ormond Street Hospital, buy vital equipment and fund pioneering research. With your help we provide world class care to our very ill children and their families.

This page was last reviewed on 22 October 08 12:41