Winners for BEAT-PD DREAM Contest to Better Track Parkinson’s Named
The four winners of the Michael J. Fox Foundation (MJFF) and Sage Bionetworks Parkinson’s disease (PD) challenge to create new ways of using everyday technology to benchmark and predict disease progression remotely — including at home — have been announced.
Called the Biomarker and Endpoint Assessment to Track Parkinson’s Disease (BEAT-PD) Dream Challenge, the contest sought to determine whether Parkinson’s severity and progression can be assessed via sensor data collected as a person goes about daily life.
This year’s contest drew 43 teams, with winners sharing a $25,000 prize.
“We congratulate all the winners,” said Mark Frasier, PhD, senior vice president, Research Programs at MJFF, in a press release. “The Foundation has supported research into sensors and other digital tools for Parkinson’s for many years.”
BEAT-PD “projects are unlocking the potential of data collected by digital devices to help people with Parkinson’s, their physicians, and researchers. Now, more than ever, we understand the critical importance of remote monitoring for the safe and effective delivery of healthcare and the progress of clinical research.”
Winners of the BEAT-PD Challenge are:
- Team dbmi members Yidi Huang, Brett Beaulieu-Jones, Mark Keller, and Mohammed Saqib with Harvard Medical School
- Team ROC BEATPD members Alex Page, Monica Javidnia, Greta Smith, Robbie Zielinski, and Charles Venuto with the University of Rochester Medical Center
- Yuanfang Guan at the University of Michigan
- Team HaProzdor members Ayala Matzner, Yuval El-Hanany, and Izhar Bar-Gad at the Gonda Brain Research Center at Bar-Ilan University
The BEAT-PD Challenge built upon a previous data challenge, in which researchers showed that disease status and symptom severity could be predicted using data collected during the physician-monitored completion of certain tasks.
In the BEAT-PD Challenge, scientists sought to learn whether disease severity can be measured via passive sensor data gathered with consumer electronics and collected not during specific tasks, but day-to-day life. The ultimate goal is to be able to monitor disease progression at home.
In their efforts, teams ROC BEATPD, dbmi, and HaProzdor applied signal processing tactics to smartphone sensor data. Allowing for patient-specific characteristics, the results were subsequently used in machine learning models. Guan applied a deep-learning model using spatial and temporal sensor data augmentation.
The winners may now collaborate to optimize their models, and gauge them against clinician-validated symptom severity ratings. They also will co-author a publication on their findings.