Intel Big Data Platform Used for Parkinson's Research

Intel and the Michael J. Fox Foundation are developing a cloud-based big data analytics platform that will leverage wearable devices to collect data.

dev ops

Intel is developing a big data analytics platform that will leverage wearable devices, an Intel-based cloud architecture and Cloudera's Hadoop distribution to help in the research around Parkinson's disease.

Intel and the Michael J. Fox Foundation for Parkinson's Research (MJFF) on Aug. 13 announced a multi-phase program that will collect data from smart wrist devices worn by Parkinson's patients and send that data to the cloud-based infrastructure, where researchers will digest and analyze the data in hopes of gleaning more information that will help them in their search for a cure.

Diane Bryant, senior vice president and general manager of Intel's Data Center Group, said during a Webcast news conference that a growing number of industries are embracing big data, which refers to the massive amounts of data that is being collected from a wide range of sources—from smartphones and tablets to sensors and connected systems—and parsing and analyzing the data so it can be used to help guide decision-making. IDC analysts see big data as a $14 billion market that will grow rapidly.

While such verticals as transportation and retail are leveraging big data analytics, "the health care industry is where we believe the best opportunity lies," Bryant said.

Many of the wearable devices on the market today are used to track the health and fitness of its wearers, from heart rates and blood pressure to steps walked and miles run. Intel and the MJFF want to use the technology to improve the data collection and analysis related to the study of Parkinson's.

MJFF CEO Todd Sherer said during the Web conference that, until now, studies of Parkinson's patients have relied on the subjects to record their experiences and periodically come in to their doctors' offices.

"It's very subjective, very variable and very episodic," Sherer said, noting that many of these research trials can cost hundreds of millions of dollars to run. "We've got to do better."

In the program with Intel, Parkinson's patients and members of a control group will wear the wrist devices, which will measure around-the-clock such metrics as how slow their movements are, how well they're sleeping, and any tremors they may feel in their bodies. That information will be sent via their smartphones to the cloud-based analytics platform that includes Cloudera's Hadoop technology, where algorithms will be developed that will enable researchers to get as much information from the data as possible.

The amount of data collected will be enormous, Bryant said. The wearable devices will make 300 observations every second. Intel and the MJFF early this year ran a trial of the platform using 25 participants—16 patients and nine control volunteers—that collected data continuously over four days. The goal is to give researchers much more data from thousands of patients to give them better insights into such areas as how the disease affects individual people, how it progresses and how different patients react to different treatments.

The cloud platform also will make it easier for researchers from different institutions to collaborate, Bryant said. The platform also could be used to store other data from patients and clinical trials, and could result in other research techniques such as machine learning and graph analytics.

In addition, it could be a model for other platforms that could be leveraged for research into other health issues, Bryant said.

Later this year, Intel and the MJFF will roll out a mobile application that will enable patients to report the medicine they are consuming and how they are feeling, the organizations said. That will add even more context to the data the wearable devices are collecting.

Mount Sinai Hospital in New York City and the Tel Aviv Sourasky Medical Center in Israel are participating in the program.

Along with helping push research into Parkinson's disease, the platform will give Intel a way of demonstrating how its data center infrastructure products can be used in such big data environments. In addition, it also will help highlight the Hadoop distribution of Cloudera, a company Intel invested about $740 million in earlier this year.