Parkinson’s disease affects more than five million people around the world and is second only to Alzheimer’s in prevalence. The disease is a debilitating neurological disorder. Its symptoms—tremors, stiffness, difficulty walking—start imperceptibly, but worsen over time.
And despite the billions of dollars spent each year on research, there has been very little innovation in how the disease is monitored and treated. Doctors continue to rely on traditional methods such as periodic clinical assessments and patient diaries to track symptoms and the progression of the disease. Even former Intel CEO Andy Grove, who was diagnosed with Parkinson’s in 2000, called for a fresh look at the problem, stating that, “What is needed is a cultural revolution that values curiosity, follow-through and a problem-solving orientation and also puts the data being generated in full view, scrutinizable by all.”
But now, Intel and the Michael J. Fox Foundation for Parkinson’s Research are teaming up to advance the search for breakthroughs in the treatment of Parkinson’s disease. The project combines the patient data collected from wearable devices with an analytics platform to more accurately monitor patient symptoms over time. We realized that by harnessing the power of big data analytics we could help researchers study Parkinson’s more effectively with the ultimate goal of improving the lives of patients.
Intel and the Michael J. Fox Foundation already have a multiphase study underway. During initial patient trials, participants were outfitted with wearable devices that could monitor symptoms 24 hours a day, seven days a week. The devices generated a staggering amount of data—more than 300 observations per second, amounting to about 1 GB of data per day. With 10,000 patients monitored in one month, this would equal the entire amount of data in the Library of Congress—around 10 terabytes!
This round-the-clock monitoring of symptoms and vitals can provide the raw information that may lead to groundbreaking insights, completely transforming the way we approach treatment. We had to put some serious compute power behind that work: a big data analytics platform that combined Cloudera CDH, a cloud infrastructure running on Intel Xeon processors, and customized analytics software for detecting anomalies in real time.
So what’s next?
Now, it’s up to data scientists and researchers to compare the data collected to clinical observations and patient diaries to determine the accuracy of the devices, and then develop algorithms to measure symptoms and disease progression. Long-term, the platform could incorporate genetic and clinical trial data, eventually incorporating machine learning and graph analytics for more accurate predictive models.
The Michael J. Fox Foundation has just one mission: eliminate Parkinson’s disease in our lifetime. With this project, we hope we are one step closer to that goal.