How wearable tech, big data analytics and cloud help in Parkinson’s disease research

Parkinson’s disease charity MJFF is seeking to improve research with wearable devices, big data analytics and public cloud

This article can also be found in the Premium Editorial Download: Computer Weekly: Can national security and privacy co-exist?

Actor Michael J. Fox's charity for Parkinson's disease is seeking to improve the treatment for Parkinson's through a research project that involves wearable technologies, big data analytics and public cloud computing.

Parkinson's is a neurodegenerative brain disease that affects about five million people worldwide. While it usually affects people in their 60s, it can come earlier. There is a wide range of motor and non-motor symptoms associated with the disease.

"Nearly 200 years after Parkinson's disease was first described, we are still limited in tracking and measuring the disease,” said Todd Sherer, chief executive of the Michael J. Fox Foundation (MJFF). “The big challenges are the variability in the type of people who can get it, the wide-ranging symptoms and how it progresses. Parkinson’s can be variable, subjective and episodic.”

MJFF is collaborating with Intel for the project and the chipmaker has built an open-source big data analytics platform on the Amazon Web Services (AWS) cloud to analyse the patient data collected from wearable devices.

The charity is hoping the data from thousands of individuals on slowness of movement, tremor and sleep quality will help physicians measure how Parkinson’s progresses and make breakthroughs in drug development.

"Data science and wearable computing hold the potential to transform our ability to capture and objectively measure patients' actual experience of the disease, with implications for drug development, diagnosis and treatment," said Sherer.

How wearables will help battle Parkinson's

As part of the study, smartwatches gather and transmit data in real time, all the time. Such data will allow researchers to look at even minute data points and analyse hundreds of readings per second from thousands of patients and attain a critical mass of data to detect patterns and make new discoveries.

Data from thousands of individuals on slowness of movement, tremor and sleep quality could help physicians measure how Parkinson’s progresses and make breakthroughs in drug development

"The variability in Parkinson's symptoms creates unique challenges in monitoring progression of the disease," said Diane Bryant, senior vice-president and general manager of Intel's datacentre group. 

"Emerging technologies can create a new paradigm for measurement of Parkinson's. Everyone loves and wants wearables,” she added.

MJFF and Intel initiated the first phase study earlier this year to evaluate the usability and accuracy of the smartwatches for tracking physiological features from participants and using a big data analytics platform to collect and analyse the data. 

One participant, Bret Parker, 46, from New York, said many doctors tell their patients to track their Parkinson's symptoms. 

"I am not a compliant patient on that front," he said. "The wearables did that monitoring for me in a way I didn't even notice, and the study allowed me to take an active role in the process for developing a cure."

As the next step, Intel's data scientists will correlate the data collected to clinical observations and patient diaries to gauge the devices' accuracy. They are also developing algorithms to measure symptoms and disease progression.

Big data analytics and healthcare

According to Bryant, healthcare data is growing at a rate of 35% per year, and harnessing and mining this data is important. 

“But data in healthcare is complex," she said. "Big data analytics requires very tight engagement between three sets of people – data scientists, medical researchers and technology system experts.

“Our objective is to create a cost-effective technology to analyse the mass of data and create a cloud environment to share the data."

Intel’s big data strategy is to make analytics systems pervasive by making them easier to deploy and cost-effective. "This collaboration will help us do that,” said Bryant.

Big data analytics hit the scene and became an industry buzz in 2012, but only now is it becoming feasible for enterprises to use big data analytics, as the cost of compute and storage drops – thanks to Moore’s Law, she said.

“The biggest opportunity for big data analytics is in the healthcare sector and, with our collaboration with MJFF, we aim to create a blueprint for the solution that can be extended to other diseases and research,” said Bryant.

The platform, built by Intel, collects as much as 1GB of data per person, per day. 

“We have collected data from 10,000 participants, amounting to as much data as the Library of Congress. This data is encrypted and securely stored in Tel Aviv,” said Bryant.

The platform integrates a number of software components, including Cloudera CDH – an open-source software platform that collects, stores and manages data. Intel invested $740m in the startup Cloudera, which provides enterprises with the Hadoop open-source software system to manage big data challenges. 

Cloud infrastructure

It is deployed on a cloud infrastructure to allow scientists to focus on research rather than the underlying computing technologies. “The data analytics platform is currently hosted on Amazon Web Services,” said Bryant.

AWS offers powerful, cost-effective approaches to support and analyse big data. Typically priced on a per-use basis, AWS offers several services such as data transfer, and real-time processing of streaming big data.

Intel has also developed an analytics application to process and detect changes in the data in real time. By detecting anomalies and changes in sensor and other data, the platform can provide researchers with a way to measure the progression of the disease objectively.

The data and the platform, based on open source technologies, will be open and accessible to medical researchers around the world, Bryant said.

The Hadoop platform will be expanded in the future to include other types of data, such as genome, medical research and clinical trial data, she added.

The platform could enable other advanced techniques, such as machine learning and graph analytics, to deliver more accurate predictive models that researchers could use to detect changes in disease symptoms. These advances could provide unprecedented insights into Parkinson's disease, assisting physicians with prognostic decisions, according to the chipmaker.

Later this year, Intel and MJFF will launch a mobile application that will enable patients to report their medication intake, as well as how they are feeling. It is part of the next phase of the study to enable medical researchers to study the effects of medication on motor symptoms via changes detected in sensor data from wearable devices.

Intel’s keen interest in the MJFF project comes from the fact that its former chief executive, Andy Grove, was diagnosed with Parkinson’s and serves as a senior advisor to the charity.

Read more on Clustering for high availability and HPC