This will be used to form the basis of a
separate app — NeoGest — that will allow people in the developing world
to use a smart phone to more accurately estimate the prematurity of a
baby and decide whether specialist medical care is required.
Dr Michel Valstar, of the University’s
School of Computer Sciences, has been leading the development of
BabyFace.
He said: ““What defines each gestational
age is the number of wrinkles and the depth of them in the feet. The
roundness of the eyes and how sharp they are at the corners is another
indicator. By using and storing the pictures we collect, the app will
begin to recall the shape of the eyes, ears and wrinkles in the feet.”
More
than one million babies worldwide die as a result of prematurity every
year, the majority of whom are born in the developing world. Premature
babies in continents like Africa are often at greater risk of dying as a
result of poor nutrition and infections, as well as a result of serious
diseases including TB and HIV.
In rural areas, women rarely have access
to the antenatal ultrasound scans which doctors use to determine the
gestation of the babies of their western counterparts. Traditional
methods of estimating gestation based on measurements and birth weight
are not always reliable.
However, being able to determine the
development of babies — and the potential consequences of their
premature birth — can help village elders and community leaders to
decide whether they might simply need to offer guidance to the mother on
the hygiene and nutritional needs of their infant or whether it is
vital to make a trip to a hospital which can often be hundreds of miles
away.
The computer scientists at Nottingham have
joined forces with experts in the University’s Division of Child
Health, Obstetrics and Gynaecology for the project, which has been
supported with a $100,000 grant from the Bill and Melinda Gates
Foundation.
Dr Don
Sharkey, Clinical Associate of Neonatal Medicine, said: “If a smartphone
could give us the birth gestation of a newborn baby it would allow many
low-middle income countries to give more detailed local advice to
parents that could increase the chances of survival for premature
babies.”
If successful, the NeoGest app, which will
be free to download, will combine simple measurements with elements of
the Ballard test which is used by experienced healthcare professionals
to estimate gestation and look at developmental characteristics
including the characteristics of lines developed in the skin on the foot
or the shape of features like the ear.
It will use the mobile phone’s camera to
take images of the foot, face and ear of the baby and upload it to a
huge database where it will compare it with pictures of potentially
hundreds of other babies at various known gestational ages to find a
‘match’.
As part
of the project, the clinical team has been collecting baseline
comparable data by taking pictures of babies in the maternity and
neonatal units at Nottingham University Hospitals NHS Trust.
The team also plans to explore the
potential ethnic differences in gestational development between babies
born in different parts of the world — babies born in Africa or Asia may
have different developmental characteristics compared to their western
counterparts.
The research project was one of around
just 60 successful bids out of 1,700 applications worldwide for funding
from the Gates Foundation’s Grand Challenges Explorations Grant
programme.
Grand
Challenges Explorations (GCE) funds individuals worldwide to explore
ideas that can break the mould in how we solve persistent global health
and development challenges.
At the end of the project, if the team is
able to demonstrate a successful prototype they will have the
opportunity to bid for a much larger package of funding from the Gates
Foundation to develop the technology for use in the real world.
BabyFace is available now for free
download for Apple mobile devices via the Apple App Store. In line with
the project’s privacy policy, all data collected are completely
anonymous, cannot be traced back to any of the participants and will
only be accessible to the members of the research team in charge of the
project.