Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary

The first Clinically Validated AI-powered Diabetes Assistant

Article Category

Article available in the following languages:

AI powered mobile app supports daily diabetes management

According to the WHO, an estimated 1.6 million deaths were directly attributable to diabetes in 2016. DIABETESMART capitalised on a collaboration between developers, healthcare professionals and people with type 1 and 2 diabetes to develop a self-management tool.

Much research has been invested in identifying the causes of diabetes and trying to find a cure. While some light has been shone on how the condition may arise, including genes and the possibility of a viral impact, there is no cure. Current treatment remains focused on the management of blood sugar levels with insulin, diet and lifestyle. Sufferers of diabetes have to commit to a life-long care management regime. Mobile phone apps have been shown to help support the regulation of this, with the marketplace offering many options. EU support through the DIABETESMART project enabled the Danish start-up company Hedia to conduct a feasibility study into further development of its mobile platform for daily diabetes management.

Supporting self-management

Diabetes mellitus is a common chronic metabolic disorder where either the pancreas does not produce enough insulin (type 1), or the body’s cells do not respond appropriately to the insulin that is produced (type 2). Insulin is a hormone needed to allow sugar (glucose) to enter cells to produce energy. Untreated, it can result in acute complications (such as cardiovascular disease or stroke) and can be fatal. The goal of the Hedia diabetes app is to help sufferers maintain a more stable blood sugar level for longer. Users can transfer information about their current blood glucose levels wirelessly, using near field communication (NFC), to Hedia through multiple blood glucose monitoring devices. The AI in the device can forecast where blood glucose levels will be in 60 minutes and as the system logs the user’s last dosage, so it can report on how much active insulin remains. Hedia matches glucose levels against its food database, maintained by a nutritionist, to make dietary suggestions, alongside a recommended insulin dosage. The AI functionality allows the app to recognise eating patterns over time and so make suggested adjustments tailored to individuals. As physical activity also impacts blood glucose levels, information about this can also be entered for even more precise recommendations. “As someone with type 1 diabetes, I know that frequent spikes and drops in blood sugar levels can feel like a roller coaster ride. Key to helping me manage this chronic condition on a daily basis, is an increase in my time-in-range,” explains Peter Lucas, project coordinator.

Meeting a growing need

According to the WHO, in 2014 422 million people were diagnosed with diabetes, with the global prevalence almost doubling between 1980 and 2014 among adults over 18 years of age (from 4.7 % to 8.5 %). With this trend increasing in middle- and low-income countries, the race is on to find an effective means to manage the illness, minimise suffering and reduce the burden on healthcare systems. The Hedia device has been clinically validated as Class I compliant with the EU’s Medical Device Directive (MDD) and is currently undergoing the final audit to be accredited Class IIB status. “We are very proud of the fact that Hedia is outperforming current state-of-the-art competitors as a self-care, one-stop-shop. It has been very motivating to get positive feedback from people, just like me, about how we are helping them,” says Lucas. With the device already available commercially in Denmark and the UK as a download from the Apple and Android stores, the team are currently working on further clinical validation of the AI components, with a view to rolling it out to companies and individuals within 2 years.

Keywords

DIABETESMART, diabetes, blood sugar, glucose, insulin, self-care, AI, diet, hormone, healthcare

Discover other articles in the same domain of application