Spotlight on CNS Summit 2020: AiCure Talks Digital Biomarker Development with Merck and Kent State University


An Open-Science Framework to Digital Biomarker Development and Validation

Featuring:

  • Bryan Hansen PhD, Associate Principal Scientist, Global Digital Analytics & Technologies at Merck
  • John Gunstad PhD, Professor of Psychological Sciences and Associate Director of the Brain Health Research at Kent State University
  •  Isaac Galatzer-Levy PhD, Chief Scientific Officer at AiCure
  • Anzar Abbas PhD, Director of Research & Development at AiCure

Sensor-based digital measurements, known as digital biomarkers, have rapidly emerged as an essential tool for next-generation clinical research. Digital biomarkers hold promise of greater objectivity, sensitivity, and frequency of assessment all at lower cost and lower friction. Despite their promise, digital biomarker models and methodologies are often shrouded in mystery as proprietary machine learning models are not made accessible to scientists to assess or independently validate.

This spotlight presentation will focus on opening the black box. The panelists will discuss the need and emergent approaches to evaluate, standardize, and validate digital biomarker methodologies. First, Isaac Galatzer-Levy PhD and Anzar Abbas PhD will discuss the launch of AiCure’s open-source computer vision and vocal digital biomarker platform, OpenDBM, meant to encourage scientific scrutiny and collaboration. John Gunstad PhD, Professor of Psychological Sciences and Associate Director of the Brain Health Research Institute at Kent State University, will discuss the application of computer vision and vocal digital biomarkers in his funded research into prodromal symptoms of Alzheimer’s Disease. Finally, Bryan Hansen PhD, Associate Principal Scientist, Global Digital Analytics & Technologies at Merck, will discuss his work to evaluate, validate, and recommend digital biomarker approaches to advance clinical trial research.

OpenDBM is available to download, at no cost, on Github, along with instructional tutorials and walkthroughs on how it can be applied on any video or audio dataset. Click here to learn more.