NSF/FDA SIR: Towards the Establishment of a Validation Framework for Wearable Motion Analysis Systems: Development and Evaluation of an Open-Design Sync Platform

点击翻译按钮获取中文摘要

Consumer-grade wearable sensors are increasingly used in biomedical and clinical research. The ability to “measure what it is supposed to measured”, is a basic requirement for any sensor used in research. Conventional validation procedures involve comparing the outputs of a wearable device with those of gold-standard laboratory equipment. For devices designed to measure human motion, this process requires millisecond-level alignment of the data measured by the two systems. As a first step toward the development of a standardized validation framework for wearable devices, this two-year NSF/FDA Scholar-in-Residence project will establish novel data-driven method to enable accurate time alignment between consumer-grade wearable sensors and stationary laboratory equipment. The models will be implemented in an open-source, affordable hardware/software Sync Platform, whose design will be disseminated for use and further development by the research community through an open repository. The project will contribute new educational materials for the PI’s graduate-level course on wearable technologies and provide a unique training opportunity for a graduate research assistant, who will learn about regulatory science by working in the collaborating FDA laboratory for part of the project.The project seeks to address the current lack of standardized validation approaches for wearable sensors used in clinical and biomedical research to quantify human motion in controlled and natural settings. The team at Stevens Institute of Technology will develop the Sync Platform consisting of a Sync Module (hardware) and a companion Sync App (mobile application) featuring novel probabilistic and learning-based data-driven delay compensation models (D3CMs) to enable accurate on-line time alignment between wearable sensors and stationary laboratory equipment. In collaboration with researchers at the FDA Center for Devices and Radiological Health (CDRH), the team will jointly evaluate the usability of the Sync Platform and its reliability under different levels of electromagnetic interference with a representative set of wearable sensors. The project will advance the state of the art on wearable technologies and biomedical research, by addressing the challenging problem of precisely synchronizing data streams from consumer-grade wearables and stationary reference instrumentation, thereby providing the foundations for the development of a standardized procedure to characterize validity and reliability of wearable devices with enhanced rigor and granularity. The methods developed in the project will also serve the broad research community by providing a tool to enable the precise alignment of data streams from heterogenous wired and wireless devices designed to capture human motion. Ultimately, these methods will promote understanding of the validity and reliability landscape afforded by commercially available wearables and assist in the design of future clinical and biomedical studies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.