The e-Science Central Study Data Platform

Abstract:

This paper describes a novel study management platform that is being used to collect, process and analyse data gathered from a large-scale pan-European digital healthcare study. The platform consists of two main components. Firstly a secure, scalable, cloud-based platform to ingest and process data uploaded from body-worn sensors, as well as from clinical evaluation forms. Features extracted from this data are then loaded into a Data Warehouse with a novel schema designed specifically for study data. This allows scientists to explore, analyse and visualise this data in a variety of different ways. A key aspect of the warehouse design is that it also stores metadata describing the types and format of the data. This enables automatic report generation, exploratory data analysis and error checking. The overall result is a flexible, general purpose system that is open-source and uses the cloud for scalability. This paper describes the design of the integrated study data platform and its use in the large Mobilise-D study that has collected and analysed both sensor and clinical data from over 3.000 participants.