British Unis and SMEs use Big Data and AI to combat Parkinson’s

New app could lead to improved treatment for patients, say collaborators

Posted 15 March 2017 by

A new app developed by London Universities and AI start-ups could help track the development of Parkinson’s Disease symptoms, as well as potentially lead to more personalised treatments for the 120,000 people in the UK with the condition.

The software, cloudUPDRS, combines a smartphone app and an online Big Data analytics service its developers say is smart enough to make “objective and reliable assessments” of a patient’s motor performance.

And it’s so effective, it’s been certified as a medical device not a lifestyle application, as most apps are.

cloudUPDRS can be used by a patient at home, allowing them to record movements by tapping the screen to assess things like bradykinesia (slowness of movement), or holding the phone on their knee to assess tremor.

The app securely uploads these measurements to the cloudUPDRS analytics server, which uses state-of-the-art cloud-based Big Data technology and analytics such as microservices and deep learning to calculate a score in the format of the clinical Universal Parkinson’s Disease Rating Scale (UPDRS).

Additional data is being gathered from multiple uses to enable trend analysis and patient stratification that can provide detailed information on disease progression and inform treatment strategies, says the team behind cloudUPRDS.

The software was built by researchers from two London academic institutions, the Department of Computer Science and Information Systems at Birkbeck and colleagues at UCL.

They have been joined by a specialist Big Data SME Benchmark Performance and Retechnica, an AI and machine learning startup, with all partners determined to offer a practical way to help patients.

Parkinson’s sufferers currently tend to only see a specialist nurse or doctor only once or twice a year, allowing only brief and intermittent assessment of the wide range of their motor and non-motor symptoms, says the team.

For Birkbeck’s Professor George Roussos, research lead in computing technology for the app, “The cloudUPDRS system can provide a range of benefits to both patients and clinicians.

“More regular assessments of disease progression mean that patients receive more consistent and reliable care, and detailed and automated patient analytics permit the early identification of problems such as medication side-effects.

“By collecting and analysing data ahead of appointments, clinicians and patients can focus clinic time on treatment strategies, rather than clinical assessment [while] by monitoring symptoms in real time, patients can be directly involved in efforts to improve the management of their own care and can receive tailored advice on managing their symptoms through measures such as improved nutrition and physical therapy.”

cloudUPDRS has been in development since 2012.

A clinical trial to measure how it compares to traditional methods of Parkinson’s Disease symptom tracking is currently underway.