Nearly 200 years after Dr. James Parkinson first described the disease, doctors "are still subjectively measuring Parkinson's disease," says Todd Sherer, CEO of The Michael J. Fox Foundation, which is collaborating with Intel on the project.
More Rigorous Monitoring of Parkinson's Patients
Grove wanted a better, more rigorous way of measuring what was happening to his body and millions of other patients like him. About 15 months ago, he asked Intel for help in doing that, Kasabian says, noting that it took the company about four months to understand the different options and how to start.
Together with researchers, Intel recognized that the problem was getting objective data, and lots of it. The solution was surprisingly simple: A wearable with an accelerometer — the same accelerometer that can be found in every smartphone, smartwatch or other wearable device.
[ Related: Intel Developing Sensor Chips for Wearables (and Robots) ]
For the clinical trials, Intel and the researchers opted for a simple, off-the-shelf smartwatch, but Kasabian notes it could be a pendant or any other type of wearable with an accelerometer. The unobtrusive device measures slowness of movement, tremor and sleep quality 24 hours a day, seven days a week. The device takes more than 300 observations per second from each patient — about 1 GB of data per patient per day.
Within the next year, researchers hope to be able to capture EKG, blood pressure and, especially, heart rate data. "They don't know whether there's a correlation between heart rate and disease progression," Kasabian says. Researchers also want to see the effect of medication in real-time.
To analyze all that data, Intel developed a big data analytics platform built on the Cloudera CDH distribution of Apache Hadoop. It deployed the platform on cloud infrastructure. Intel developed an analytics application to process and detect real-time changes. The idea is to create a baseline and then detect anomalies and changes in sensor and other data to provide researchers with a way to objectively measure disease progression.
"Given where they're coming from, a simple time-series analysis is incredibly insightful for them. They were thrilled," Kasabian says. "Working with the researchers, we can look at specific patterns in those time series. Eventually, the system will identify certain patients where it is progressing at an accelerated rate over normal, but we have to collect a lot of data first to understand what normal is."
"Data science and wearable computing hold the potential to transform our ability to capture and objectively measure patients' actual experience of disease, with unprecedented implications for Parkinson's drug development, diagnosis and treatment," adds Sherer:
Sign up for CIO Asia eNewsletters.