It is now almost two weeks since myself and my colleague Steve Smith volunteered at Rarefest and I have to say I am still so inspired by the people we met, the stories that were told and the extraordinary performances we saw. Since then I have been reviewing the literature that we took away from the event and have been further inspired, reading about new developments in the clinical world, the support of the various charities involved and the medtech approach to rare diseases.
Whilst reading through issue 005 of Rare Revolution magazine, I came across an article about diagnosis. It was hard to read that a child with a rare disease can endure a seven year wait (on average) in order to receive an accurate diagnosis and that this would entail many unpleasant tests and consultations with seven different doctors (again an average figure). It was, however, heartening to read that advancements in technology are investigating solutions to this. Facial recognition is no longer a thing of science fiction (as anyone with the latest iPhone can testify) and a combination of this, alongside computer learning is being applied here to make a difference to patients and to help them reach a diagnosis of their rare disease much quicker.
The origins of many rare diseases begin with genes mutations and due to the majority of genes associated with the development of the head, face and brain it follows that patients suffering from the same rare diseases would have similarities in appearance. This trend has caused investigations into this route as a possible starting point for diagnosis.
Christoffer Nellaker and his team at The Big Data Institute at Oxford University are researching this and utilising computer learning in order to ‘teach’ the computer to recognise similarities. The name Big Data is no coincidence as this is exactly what the computer needs to recognise appropriate characteristics but disregard those that are not important. Through Minerva And Me they are trying to amass as many photographs as possible to build a database. Obviously, the computer needs images of people with rare diseases but it also needs those of people without. To make an accurate diagnosis it needs to be able to understand how people’s faces change as they age, how they appear different due to cosmetics or different hairstyles, the angle from which the photo was taken, but also the differences reflected in facial shape due to gender and ethnic group. This means BIG DATA, to teach and reinforce correct comparisons for diagnosis, from all over the world.
Face2Gene, produced by FDNA in America, is a similar platform achieving very positive results. So much so that it is being utilised in approximately two thousand Institutes worldwide and is even becoming part of the training for future geneticists. This technology will compare facial shape/dimensions, in addition to other tests, for a suggested diagnosis but will also suggest to Clinicians other phenotypes or genes where the mutation may have originated – prompting a targeted area for further research.
After reading this article the first thing I did was visit the Minerva and Me website. Within a couple of minutes, I had signed up and uploaded my photos to help with this research. A very simple process and it also provides options on how your image can be shared. As humans, we are suspicious creatures and sometimes when faced with something so easy to do we can ask “What is the catch?”. Quite simply there isn’t one, this is you just spending a few minutes to help others. You won’t be bombarded with spam mail from various partners or asked for financial donations (although I am sure you could if you wanted to). They simply want your photographs!!! To set up your profile you are asked to complete a few questions, including your age and gender. They will also ask if your facial appearance has been changed due to plastic/restorative surgery, accident or trauma or orthodontic treatment. Again, differences that the programme will need to learn and process to be successful. There is a participant’s information sheet that covers everything you need to know, including how your information and photos will be used and how they are protected. As you have an online profile to log in you can also link yours to other family members, to help the programme recognise familial similarities and you can also change your information.
It seems spooky that this time last year Next Phase Recruitment ran a photograph competition and we asked people within our network to guess our colleagues from their baby/toddler photographs.
I am pleased to say that this year our team have agreed to share these photographs (alongside some more recent ones) to help rare disease diagnosis and are signing up to Minerva and Me.
If you are interested in learning more and are keen to sign up to help (it literally takes less than a minute) please see the link below;
A full copy of the article in Rare Revolution, by Christoffer Nellaker, can be accessed here.