Developing a Cloud-Based Platform to Standardize Data Storage from Wearable Brain Sensing Devices
Originally published by Rafik B. Hariri Institute for Computing and Computational Science & Engineering
By Gina Mantica
To study how the brain works in the real world rather than the lab, researchers are creating wearable devices that obtain a complete picture of the brain’s activity in real time. But the complex data collected by researchers needs to be stored, organized, and shared in a way that allows for more meaningful collaboration and analysis. Researchers at Boston University are developing a web platform that not only automates the storage of neurological data collected from wearable devices, but also standardizes how the data are stored.
David Boas, Director of the Neurophotonics Center, Professor of Biomedical Engineering and Electrical and Computer Engineering, and Research Fellow at the Hariri Institute, received a supplementary award from the National Institutes of Health to team up with the Software & Application Innovation Lab (SAIL) and create a platform that improves quality assurance and enables easy data analysis.
Data collected from wearable neurological devices can help researchers understand how a person’s environment affects their behavior and in turn, the activity of their brain. Functional near infrared spectroscopy (fNIRS) is a neuroimaging technique that tracks blood flow in the brain to understand how activity across the brain changes over time. The researchers combine fNIRS with electroencephalograms (EEG) that measure the brain’s electrical activity and eye-tracking software to time lock the brain’s activity to data on what people see and interact with in their environments. This provides researchers with a complete picture of the brain’s activity in the everyday world.
Researchers store data collected from these devices, as well as analysis methods for this rapidly growing field of research, on GitHub and other databases. But storage on these platforms is not standardized or organized in a way that facilitates data sharing. SAIL engineers will build a cloud-based web platform that helps users create and modify data folders in standardized formats. “We don’t aim to replace those servers, but to provide features that go beyond storing and sharing,” said Boas. The interface will remind users to include all necessary information and alert the user if information is missing or uploaded incorrectly. And, the platform will assess the signal to noise ratio of data files for quality assurance. Users will also be encouraged to upload their data immediately after collection to get immediate feedback of data quality issues that help improve future measurements.
Automating and standardizing the organization and analysis of complex neurological data will advance our understanding of how the brain works, and could help inform future treatments for patients with strokes, concussions, Parkinson’s disease, or Alzheimer’s disease. “It will accelerate the adoption of good data organization and data analysis methods, particularly for those new in this rapidly growing field,” said Boas, “Thus researchers will make fewer mistakes and be more likely to make meaningful interpretations of their datasets to contribute to our body of knowledge of how the human brain functions.”