Management Framework for Biosignals in Biomedical Studies: From Study Design to Data Statistics

Abstract

A centralized and structured management framework for biosignals is proposed. This framework is dedicated to design biomedical studies, acquire and store biosignals and the associated metadata, as well as validate, synchronize, view, and evaluate these signals from an individual study-related perspective. We introduce a five-layer model as the basis of the framework to facilitate data management, which includes the acquisition, validation, postprocessing, and statistics layers, as well as the vertical presentation layer. The strategic interaction between external data (i.e., metadata) and internal data (i.e., recorded biosignals) is defined to favor high-quality results and sustainability of recorded data. For the first time, a systematic software solution was designed to mimic the proposed layer model and to reasonably combine internal and external data management using microservices within a private cloud. Researchers can individually access and manipulate all data using a client implementation or a Web browser in order to address research-related questions, whereas developers can easily implement modular functions and algorithms. A simplified implementation of the proposed framework shows its strong potential in the unification and structuring of biomedical studies as well as in the associated acquisition, storage, and analysis of biosignals.