Intel and the Michael J. Fox Foundation have partnered to combine mobile technology — smartwatches and smartphones — with big data analytics to further Parkinson's research.
In 2000, Andy Grove, Intel's former CEO, was diagnosed with Parkinson's disease. Being an engineer, he wanted to approach his treatment from a data driven, mathematical mindset. Since then, he has been on a mission to find ways that data analytics could drive research to better prepare patients for disease progression. Grove's passion and dedication to the cause made it an important one for Intel to tackle as a company, too.
So last month, Intel announced a partnership with the Michael J. Fox Foundation, the world's largest nonprofit funder of Parkinson's research, to utilize wearables and big data analytics to further research about the debilitating disease.
"We've long known that there is substantial need for an objective measure of the Parkinson's disease lived experience," said Todd Sherer, CEO of MJFF. "As technologies have evolved, there are more opportunities to collect that data outside the clinic passively and continuously as people with Parkinson's go about their days."
Parkinson's is second to Alzheimer's in worldwide prevalence of neurodegenerative brain diseases. According to the Parkinson's Disease Foundation, about 60,000 new cases are diagnosed each year in America, and seven to 10 million people are living with the disease worldwide.
The study was just the first step in this advanced research. Researchers at Mount Sinai Hospital in New York and Tel Aviv Sourasky Medical Center had 16 people with Parkinson's and nine healthy people participate. The participants visited the clinic for observational exams and wore smartwatch sensors and a smartphone over three days at home, as well as completed a diaries of their activities.
The smartwatch and smartphone monitored their movements — more than 300 observations each second — over a six-month period. The two systems gathered data regarding tremors, sleep quality, and slowness of movement. The smartphone served as the base, and the watch as a secondary source, which provided a more well-rounded view of disease progression.
One of the tests for Parkinson's patients when they visit the doctor is to stand up, walk across a room, turn around, and come back.
"Now you can do that every day, every 5 minutes, and capture a more accurate signal to
start to identify micro patterns in movement," said Vin Sharma, big data analytics strategist at Intel, "and build a library of data to understand how the disease is progressing."
The way Parkinson's patients are treated is very subjective. The disease takes various patterns as well as symptoms starting at different times, making it difficult to track and individualize treatments. The potential for this approach then, Sharma said, is that eventually — when smatwatches become as pervasive as smartphones — this type of research isn't prohibitive for the majority of people living with Parkinson's, which allows for a huge expansion for research.
Sherer said he would like to eventually outfit as many people as possible with these devices as a tool to supplement patient diaries and clinical exams.
"Rather than relying solely on periodic appointments or subjective daily logs, investigators could use the devices to measure effect of a candidate drug, perhaps more efficiently and more accurately," he said.
Of course, each person in this small study generated a gig of data total, Sharma said, which quickly becomes a big problem. Intel had to design a database infrastructure that could not only handle the data collected from the study, but was also equipped to handle the addition of multiple data sources — like clinical records, DNA tests, etc. — that could be added to the system for years to come.
The platform Intel developed integrates many software components, including Cloudera CDH, an open source software platform that collects and manages data. With a cloud-based database, the research projects will have no issues as they scale.
All data is anonymized, Sharma said. Patients will always have the choice to opt-in to contribute to research. The doctor's file and data gathered from each patient is kept separate, and the software has multiple layers for added security. The database is primarily for researchers to look for micro patterns, trends, and anomalies as it grows.
Patients who have not experienced symptoms yet can prepare themselves, making them feel less victimized and more in control of the disease as researchers search for a cure.
"Our platform is open-access, meaning researchers can upload their own data for analysis, contribute that de-identified for population-wide analyses and access data from studies other than their own to further explore their hypotheses," Sherer said. "We hope more data and more people working to find clues from that data means more scientific breakthroughs that will lead us to the cure."