Mathematician and TED fellow Max Little has analysed the human voice since studying for his PhD at Oxford University. Among his discoveries was that our vocal nuances are indicative of our health and that Parkinson’s Disease, for example could be diagnosed through voice analysis.
Catching it at the early stages of onset could make the condition more manageable and relieve the suffering in a condition that affects 6.3 million people in the world.
But there was one big problem. Though Little is a talented mathematician and could create an algorithm for analysing patient data, he knew next to nothing about the necessary programming of communications systems. However, mobile comms systems needed to play the central role in his mass diagnostic system. The voice aspect of mobile phone networks would enable his system to analyse patients and identify those who might potentially need treatment. And the data aspect of the networks would provide the foundation for more thorough statistically significant scientific and contextual analysis of all the information gathered in his studies.
The problem with codifying communications was that it was so complicated. You need to know about the languages and syntaxes used by devices to speak to each other, such as the CODEC (coders and decoders) and signalling protocols. Or you did, until fairly recently. These days there are development platforms that promise to handle all the complicated bits of building a communications app.
Little applied his mathematical knowledge to create algorithms that could analyse the data around each person’s digitised voice. His complex mathematical formulae could identify patterns in their speech that are statistically consistent with the incidence of Parkinson’s disease. According to the US National Centre for Voice and Speech up to 90 per cent of people with Parkinson disease (PD) will have speech and voice disorders at some stage. By analysing for characteristics like reduced loudness, monopitch, hoarseness, a breathy voice quality and imprecise articulation, the likelihood of the disease being present can be calculated.
Most people with Parkinson’s aren’t aware that their speech is getting softer and more difficult to understand. Little needed a system to link these in with a mobile comms network.
In order to accurately establish the link between these symptoms and the diagnosis of the disease, Little had to have a study group of 17,000 people with a variety of different genes, phone types, behaviours and environments.
Their data collection tech (and techniques) need to be accessible and easy for the public to use while allowing for the assimilation of accurate and objectively collected data.
In this case, Twilio made it possible for Little to capture all this ‘vocal behavioural information’ by using something everybody already had, a smartphone.
Normally, a clinical trial for a new system would cost millions of Euros, but Little’s budget ran to around $1000 because he could use the minicomputers (i.e the mobile phones) that everyone carries around with them.
The app was built for the portal Patientslikeme, which enables people with similar conditions to connect with each other and share experiences and support each other.
Now, through the crowdsourced scientific research platform Sage Bionetworks, the Parkinson’s Voice Initiative is offering valuable data that will help other research organisations to further their own studies. Not bad for a self made app by a man who’d never written a programme before. This is what happens when the intelligence of the people is mobilised.
There is a disconnect in the communications systems of many large organisations and a crying need for developers to tie up all the loose ends. By the same token, the omni-channels that many companies boast about actually only produce omni-chaos. The piece that is missing is context.
There are growing numbers of people who have used mobile app development tools to create their own comms applications for social good. In the US the National Association of Guide Dog Users (a voluntary organisation with limited budgets) has created a simple app for Frequently Asked Questions. The American Red Cross has created a disaster response co-ordination system, based on SMS, to mobilise volunteers.
“We’re doing for communications what AWS did for computing,” says Twilio’s head of product management Patrick Malatack. At the moment, he says, there is a gap between those who know and those who code. By demystifying comms, there could be a massive opportunity for citizen developers in the mobile comms industry.