Novel DHT-centric statistical methods for subject-level fingerprinting and handling missingness
CERSI Collaborators: Vadim Zipunnikov, PhD; Pratim Guha Niyogi, PhD
FDA Collaborators: Maria Matilde Kam, PhD; Paul Schuette, PhD; Andrew Potter, PhD; Lili Garrard, PhD
Project Start: September 1, 2023
Regulatory Science Framework
Primary Charge: I. Modernize development and evaluation of FDA-regulated products: C. Analytical and Computational Methods
Regulatory Science Challenge
Digital Health Technologies (DHTs) are now being used in many clinical studies to track physical activity and sleep patterns frequently. The frequent monitoring provides a wealth of data that can help enhance understanding of an individual’s health and improve treatments. However, there are major challenges in making the full use of DHT data, mainly due to its complexity and the need for advanced statistical methods to analyze it appropriately.
Project Description and Goals
Project investigators propose to develop novel statistical methods to verify that the DHT data has been generated by the same individual based on their physical activity and sleep patterns, and fill in missing data gaps in DHT data by combining tools from time series, functional, and distributional data analyses.
Anticipated Outcomes/Impact
The anticipated outcome of this project include the following:
- Technical report on developed methods, description, simulations, tutorials demonstrating the use and applications of developed methods.
- Open-source R software implementing methods with built-in public domain DHTs datasets.