October 18, 2018 (12:00 – 1:30 p.m. ET)
Obstetric and Pediatric Pharmacology and Therapeutics Branch (OPPTB), Best Pharmaceuticals for Children Act (BPCA), NICHD
6701 Rockledge Drive, Room 9100/9104, Bethesda, MD
The World Health Organization defines a biomarker as "any substance, structure or process that can be measured in the body or its products and influence or predict the incidence of outcome or disease.” To date, most biomarkers have been substances measured in the blood, tissues or exhaled breath. The development of electronic systems that measure and collect patient data has given rise to an extension of the range of potential biomarkers into the digital space. In recent years, the availability of electronically available phenotypic data, as well as high resolution physiologic data, has allowed the development of digital biomarkers that can be combined with traditional biologic biomarkers to produce better performing predictors of disease trajectory and outcome than is achievable with either biomarker type alone.
Digital biomarker discovery is often undertaken through the use of large data sets combined with artificial intelligence or machine learning techniques. Data is the lifeblood of this type of machine learning and artificial intelligence, and thus large data sets are essential to the process of digital biomarker discovery. Digital biomarkers and predictive algorithms derived from large data sets offer clinicians the potential to catch a glimpse into the future trajectory of a patient’s physiology or disease state.
In this presentation, Dr. Frassica will explore the collection and dissemination of high-resolution patient data sets that support these discoveries and will briefly review developing global trends with regard to the use and dissemination of such large-scale patient data. Examples from work with both pediatric biomarker discovery and adult data will be shared.
Dr. Frassica will also present examples of digital biomarkers and predictive algorithms discovered through the use of extremely large data sets of high-resolution physiologic and clinical data. He will review the opportunities and potential pitfalls presented by these types of data through a review of a failed attempt at creating a predictive algorithm and a subsequent successful re-attempt to solve the same problem.
The speaker will present in person, but a webinar option is available.
To register for in person attendance or for the webinar, please visit http://www.cvent.com/d/8bq8sq/4W .
George Giacoia, OPPTB, NICHD
Phone: (301) 496-5589