The Alzheimer’s Disease Prediction Service (ADPS) employs an easy smartphone test to predict whether someone is likely to develop Alzheimer’s in the next six years, using data that has shown 94% accuracy. It will be one of the first validated solutions to enter the EU market as a pre-symptomatic biomarker to predict Alzheimer’s risk for people over 50.

Origins

Current diagnostic measures of Alzheimer’s Disease are invasive, expensive and time-consuming, and thus have limited use as frontline screening tools. Although experts realise the potential of screening with virtual reality, augmented reality or activity tracking technologies, routine examination is only carried out after symptoms appear and brain damage is already done. Early screening enables prevention and could dramatically reduce future costs of dementia care.

Team

ADPS unites leading academic and industry partners with a focus on computational dementia diagnosis. It involves memory clinics specialsed in data on pre-symptomatic patients – in Dublin, Barcelona, Zurich, Lausanne and Amsterdam. Altoida AG, the business champion for this project, was spun off from ETH Zurich and founded in 2016 by Ioannis Tarnanas (PhD Neuroscience), Max Buegler (PhD Machine Learning) and Richard Fischer (healthcare entrepreneur).
The project

ADPS develops and markets a solution with a 10-minute smartphone-based test that can predict Alzheimer’s Disease before patients even show symptoms. This project builds on the team’s previous datasets, which include more than 3 500 patients and have shown 87-94% diagnostic accuracy in predicting Alzheimer’s up to six years in advance.

ADPS has a long scientific history: in 2000 it became the first solution in the world to assess cognition using virtual reality. The solution now uses augmented reality and is already deployed to neurologists and primary care physicians as a cognitive assessment tool for patients.

It is also distributed to global biopharmaceutical companies, to assist with the discovery of novel treatments for Alzheimer’s. The objective of this Innovation Project is to further evaluate the accuracy and performance of the ADPS prototype – which continuously tracks everyday cognition and combines different mobile phone activity data layers using machine learning techniques.

Compared to the more invasive, expensive Alzheimer’s tests that are currently used, ADPS will be more scalable and more independent of human error during the analysis.

Impact

According to a 2014 Alzheimer report, only 35% of dementia cases are diagnosed early enough. Predicting Alzheimer’s disease before brain damage is done makes it possible to use better preventive treatment, and patients’ lives can be greatly improved, while social costs are reduced. If this solution is applied today, it could save an estimated €500 billion by 2050 in annual dementia costs.

Why this is an EIT Health project

In keeping with a key EIT Health goal, this project promotes active ageing. It is also in line with these EIT Health focus areas:

• “Care Pathways”, by facilitating an earlier start to care.
• “Real World Data”, by using patient records to define its markers for Alzheimers.
• “Bringing Care Home”, by providing a test, and exercises, that patients can do on their smartphone.