David Friedenberg is taming big data for healthcare and medical device innovation.
A Principal Research Statistician at Battelle, David applies state-of-the-art machine learning methods to big data problems in neuroscience, analytical chemistry and other domains. For the last several years, he has been working primarily on Battelle NeuroLife®
, an innovative neural bridging technology that has given a quadriplegic man conscious control over his hand and fingers.
NeuroLife uses a brain implant and a specially designed sleeve to restore conscious control to muscles while bypassing the damaged part of the spinal cord. Brain signals are picked up from a chip implanted in the motor cortex and sent to the sleeve, which uses electrical signals to stimulate specific muscles. David’s part of the project focuses on what happens in between: interpreting the volumes of data collected by the brain chip and translating it into usable signals that can be sent to the sleeve. He led the algorithm and data science team responsible for developing the algorithms to make this possible.
“NeuroLife is using data science to push the boundaries of medical technology,” David explains. “The brain produces incredibly large volumes of data, which includes a lot of noise. We had to develop new algorithms and data processing methods to make this raw data usable and pull out meaningful signals.” David and his team developed the initial algorithms through pre-clinical work and have refined them over the last two years in the first human trial of the technology.
David holds a B.S. in Mathematics and Statistics from Miami University in Oxford, Ohio and a Ph.D. in Statistics from Carnegie Mellon University in Pittsburgh. His graduate work involved analyzing digital images from telescopes. “Today’s telescopes capture terabytes of data every night—more than any human can analyze,” David explains. “We developed algorithms that can scan these images automatically and identify and categorize objects of interest for researchers.” The methods that David and his team developed for telescope data also have potential applications for medical imaging.
Since joining Battelle seven years ago, David has worked on big data projects spanning national security, analytical chemistry, tobacco regulatory science and healthcare. His prior work includes development of an algorithm to analyze Electronic Health Record (EHR) data and identify patients at risk for developing complications. He also led algorithm work for two threat detection sensors for the U.S. government and developed statistical software for analyzing data from mass spectrometry instruments. Another project, for DARPA, resulted in a software tool that allows laboratory researchers to analyze outputs from a variety of different laboratory instruments on a single platform. Since 2012, most of his work has been devoted to neuroscience, including NeuroLife.
All of his projects share a common theme: the need to extract meaningful information from large, messy datasets. “From an analytical perspective, the problem is the same no matter whether the data is generated by a telescope or from an FMRI,” he says. His varied background allows him to translate methods developed in different domains to find novel solutions for healthcare and medicine.
For the near term, much of David’s work will remain focused on refining the algorithms for NeuroLife. David and his team are working on improvements that will reduce the need for daily recalibration and training of the algorithms. These improvements are necessary for the technology to be usable for patients outside of a laboratory environment. David is excited by the possibilities that data analytics brings to neuroscience. “We have the opportunity to make a real difference in these patients’ day-to-day lives,” he says.
David is also continuing work in other areas of medicine and healthcare. In addition to his work in neuroscience, he sees a lot of opportunity in clinical predictive analytics. “We’re seeing a proliferation in sensors and other medical hardware. The algorithms are just starting to catch up,” he says. “Advances in artificial intelligence and machine learning will revolutionize how we think about medical devices and what they are capable of.” Over the next few years, David and his team will continue to lead the way in the big data revolution.